Nir Eyal is an entrepreneur, angel investor, and New York Times best-selling author of ‘Beyond Belief: The Science-Backed Way to Stop Limiting Yourself and Achieve Breakthrough,’ ‘Hooked: How to Build Habit-Forming Products,’ and ‘Indistractable: How to Control Your Attention and Choose Your Life.’
Website / LinkedIn

EC: In a world shaped by AI, abundance, and rising concerns about attention and trust, how should product leaders apply the ideas from Hooked today?

NE: The core advice hasn’t changed: build something people use because it improves their lives, not because it’s engineered to be hard to leave. What’s changed is that AI makes the Investment phase of the Hooked Model dramatically more powerful. Every interaction a user has with your product can now inform the next trigger, sharpen the next reward, and deepen the loop in ways that were previously only available to the largest platforms in the world. A small team today can build a product that personalizes itself to each user with a sophistication that would have required a data science department five years ago.

That’s an enormous opportunity, and it raises the stakes on the foundational question I introduced in Hooked: Are you building something you’d be proud to have your users hooked on? In a world of abundance, attention is the scarce resource, and users are increasingly aware of who is spending it wisely on their behalf and who isn’t. The companies that will earn lasting trust are the ones that use these new capabilities to make their products more genuinely useful, not just more sticky.

Subscribe now

EC: As products become more personalized and predictive, what responsibility do product leaders have to help users act in their own long-term interests?

NE: I’d push back slightly on the framing. The question implies that product leaders are the primary line of defense between users and bad outcomes, and I don’t think that’s right, or healthy. When we outsource all responsibility for our attention and behavior to the companies competing for it, we’ve already lost. That’s a large part of what I wrote about in Indistractable. Users have more agency than the current cultural conversation gives them credit for, and treating them as helpless victims of design doesn’t serve them.

That said, product leaders do have a real obligation, and it’s simpler than it’s often made out to be. Ask whether you’d want someone you love to use your product the way your heaviest users use it. If the answer is no, that’s your answer. Companies now have access to behavioral data that can tell them clearly when a user has crossed from engaged to dependent. Choosing to act on that data, or not, is a genuine ethical decision. The ones who ignore it aren’t just making a moral mistake. As we’ve seen in court, they’re making a business one.

Share

EC: What separates products that build meaningful, lasting habits from those that simply capture short-term attention?

NE: The variable reward. Not whether a product has one, but what it’s in service of. Every compelling product creates some form of anticipation, the email that might matter, the feed that might surprise you, the next level that might unlock something new. That mechanism is neutral. What determines whether the habit is worth forming is whether the reward, when it arrives, actually delivers something the user values in retrospect, not just in the moment.

Products that capture short-term attention are optimized for the wanting. Products that create lasting habits are optimized for the getting. Users are not as irrational as the engagement-maximizing model assumes. They notice, over time, whether a product is making their lives better. They may not quit immediately, habits are sticky by design, but they do eventually. The products people use for years, and recommend to people they care about, are the ones where the investment compounds into something real: a skill, a relationship, a measurable improvement in health or productivity or knowledge. That’s not a soft aspiration. It’s the only business model that survives long enough to matter.

“Ask whether you'd want someone you love to use your product the way your heaviest users use it. If the answer is no, that's your answer.”

— Nir Eyal

Here’s what got my attention this month — a short list of things I read and watched recently that felt worth pausing on, as a product leader.

This month’s main takeaway: The boundaries are blurring. Between PM and builder. Between software and labor. Between venture appetite and market adoption.

If you work in product, this doesn’t feel temporary. It feels like a new operating environment taking shape.

1/ AI-Native PM on the AI Exponential

One of the more interesting things I saw this month came from Cat Wu, Head of Product for Claude Code at Anthropic: A thoughtful look at how PM teams are adapting their workflows and roadmaps in the face of rapidly evolving LLMs.

There were two ideas in particular that stuck with me.

The first was a familiar-looking Venn diagram that, on the surface, could be easy to dismiss as just another version of the classic product management model. I think that would be a mistake.

What it actually reflects is the real thing happening inside product teams right now: The lines are blurring very quickly.

Tools like Claude Code, Codex, Cursor, OpenClaw, v0, Lovable, and Replit are making it easier for more people to move across the product dev lifecycle. Code is more accessible. Prototyping is faster. Distribution is being reshaped from multiple angles. Tiny teams can do the work that once required much bigger ones.

That doesn’t mean product management disappears, but it is changing. It’s almost shapeshifting.

“The new product management rhythm is rapid experimentation, consistent shipping, and doubling down on what works.”

— Cat Wu, Anthropic

We now have solo founders building real SaaS businesses. We have teams compressing in size while accelerating and expanding in output. We have people spending more time learning agentic tools — and then producing much more, faster.

The second idea was Cat’s framing around ‘side quests.’

“A side quest is a short self-directed experiment you run outside your official roadmap—an afternoon spent prototyping an idea, testing a capability you assumed was out of reach, or just seeing what happens when you push the model harder than you expect to.”

At first glance, that might sound like a modern rebrand of Google’s infamous ‘20% time.’ But what got my attention wasn’t the existence of side projects. It was the implication that, in the right environment, they may matter more than a rigid long-range roadmap.

In fast-moving environments, a side quest is not a distraction. It’s not a once-a-year hackathon either. It can be a discovery engine.

And done well — which requires the right team, and the right culture — that kind of aligned autonomy may be part of the recipe for continuously shipping actual value. It also helps explain how products or features that initially look peripheral can become strategically important faster than anyone expected.

In some ways, the PM role is changing as fast as the tools. New learning curves are forming in real time.

In other ways, the role has not changed at all.

Judgment. Taste. Decision-making. Influence. Vision. Strategy. Insight. Adaptability. Impact.
That mix is still the core of product management.

The tools are changing. The focus is changing. The lines are blurring.
But the need for people who can see clearly and bet well is not.

As Ben Horowitz put it… “Your only job is right product, right time.”

Subscribe now

2/ An Unprecedented Technology Investment Cycle

ARK’s Big Ideas 2026 report includes one of the more striking charts I saw this month: AI software investment, as a share of GDP, starting to resemble the steepest technology investment cycles in history, with the railroad era as one of the few historical comparisons.

ARK also argues that consumers are adopting AI faster than they adopted the internet. That stood out to me because we often talk about AI as if it were a replay of the internet or mobile. It isn’t. This feels so much bigger and faster, with greater intensity.

“As foundation models become a new layer of the internet stack, consumers are interacting less with apps and more through Al agents. That structural shift is activating digital experiences that delight consumers. As a result, consumers are adopting Al at a rate much faster than they did the Internet.”

— ARK Investment Management LLC

And the weird thing is: AI is not just scaling products. It is increasingly helping scale the very work that creates products.

AI is scaling… itself?

Jeremiah Owyang made that point well in his framing around self-scaling AI: agents generating knowledge, coordinating tasks, improving workflows, and reducing the degree to which innovation remains bottlenecked by purely human-paced execution.

“GenAI broke the dependency on human-paced scaling. Agents can generate knowledge, coordinate tasks, and improve workflows continuously, without waiting for us. Growth becomes less constrained by labor, and more constrained by compute, energy, and capital.”
— Jeremiah Owyang

Whether or not we agree with the most aggressive version of that thesis, the direction is hard to ignore: Growth is becoming less constrained by labor alone, and more constrained by compute, energy, and capital.

AI is eating VC

According to Carta, AI startups accounted for 41% of the $128 billion in venture dollars raised by companies on Carta last year — a record high annual share.

“This shift in the fundraising market is being driven by changes on both sides of the negotiating table: Venture capitalists are increasingly eager to back AI startups that they believe could one day deliver exponential returns, and startups are increasingly eager to incorporate new AI tools and systems into their business models.”

— eShares aka Carta

That’s venture capital reorganizing itself around AI. Investors are piling into AI so aggressively that a small fraction of startups are capturing an outsized share of funding. And then there are the headline rounds. Billions in funding by OpenAI, Anthropic and xAI is evidence of where the dollars are flowing.

AI is eating software. It’s eating venture. And when that much capital starts clustering around one technological shift, it usually tells us one of two things: Either this is a bubble that will eventually pop, or we are still early in something so significant that we’re struggling to grasp it in real time.

Subscribe now

3/ State of the Product Job Market — and Labor Market Impacts of AI

Lenny Rachitsky’s latest State of the Product Job Market showcased some encouraging news:

“There are over 7,300 open PM roles at tech companies globally, and trending up. This is 75% above the low we saw in early 2023, and already up nearly 20% since the start of this year. Today we have the most open PM roles we’ve seen since 2022.”

— Lenny Rachitsky

That does not mean we are back to the post-lockdown surge, when everything in tech seemed to be expanding at once. But it does suggest something important: For all the fear around AI and labor disruption, product hiring has not been frozen. That’s a meaningful sign of resilience.

And when it comes to ‘AI jobs,’ we’re seeing hockey-stick growth:

At the same time, Anthropic’s March 5 report on the Labor market impacts of AI added a helpful layer of nuance. Their researchers introduced a new measure of AI displacement risk based not just on theoretical model capability, but on real-world usage patterns.

One of the clearest takeaways: Actual usage is still well below what is theoretically feasible. In other words, AI may be powerful, but the labor market effects are still in relatively early innings.

A few points from that report stood out to me:

• ‘AI is far from reaching its theoretical capability: actual coverage remains a fraction of what’s feasible’

• ‘Occupations with higher observed exposure are projected by the BLS (US Bureau of Labor Statistics) to grow less through 2034’

• ‘Workers in the most exposed professions are more likely to be older, female, more educated, and higher-paid’

• ‘We find no systematic increase in unemployment for highly exposed workers since late 2022, though we find suggestive evidence that hiring of younger workers has slowed in exposed occupations’

This figure shows the top ten most exposed occupations using their task coverage measure:

“Programmers are at the top, with 75% coverage, followed by Customer Service Representatives, whose main tasks we increasingly see in first-party API traffic. Finally, Data Entry Keyers, whose primary task of reading source documents and entering data sees significant automation, are 67% covered.

At the bottom end, 30% of workers have zero coverage, as their tasks appeared too infrequently in our data to meet the minimum threshold. This group includes, for example, Cooks, Motorcycle Mechanics, Lifeguards, Bartenders, Dishwashers, and Dressing Room Attendants.”

So the job market picture right now is more paradoxical than apocalyptic. There’s demand and movement. There’s also clear signal that the shape of work is changing underneath our feet.

The question for product people is not simply whether these jobs will continue to exist. It’s whether they are learning, adapting, and evolving fast enough to remain unusually useful.

Here’s to the thinkers and builders. Stay resilient.

Subscribe now

Bonus: Claude Code LinkedIn survey results

A few weeks ago, I asked my LinkedIn network which AI coding tool they’d keep if they could only keep one.

Out of 230 votes, 75% said Claude Code. It wasn’t even close:

Here’s what got my attention this week — a short list of things I read and watched recently that felt worth pausing on, as a product leader.

TL;DR What you choose to build — and how small you keep the right team doing it — may matter more than ever.

1/ What to Vibe Code (and What Not To)

“The most important question in tech right now: what should you vibe code?
Vibe code your product. Buy everything else.”

— Elias Torres

Elias Torres (Drift Co-founder) weighed in on build vs. buy in the age of AI. If you can spin up a Calendly clone in a weekend, should you?

His answer: Only if scheduling is your product.

You can vibe code a URL shortener, a lightweight CRM, an Asana-style goals tool. But you pay somewhere — not always in dollars, often in mindshare. In a world where everyone is AI-assisted, prioritization compounds.

In response to this tweet by Naval ^, Saeed Khan wrote a post reminding us what PMs need to be strong at.

“Vibe coding is not a superpower… There are three ‘superpowers’ that Product Managers (and Product Leaders) should focus on. Discovery and Market Research. Clear and Intentional Strategy. Meaningful Differentiation.”

— Saeed Khan

I think his point is that AI lowers the cost of building, but it may not lower the cost of being wrong.

My take: Anyone vibe coding is doing discovery and strategy — implicitly or explicitly. Differentiation... not so much. We could all use some superpowers in that area.

By the way, vibe coding also will have huge implications on product design. In a recent conversation with Lenny Rachitsky, Jenny Wen (who leads design at Claude / Anthropic) shared how design is changing. She also pointed out how humans will continue to make key decisions (and be accountable for them).

Subscribe now

2/ Software Engineering Jobs Are Up. So Are Tiny Teams.

“After a brutal couple of years, software engineering job postings are FINALLY ticking up again… BUT, if you zoom out, we’re still nowhere near the COVID peak.”
— Arjay McCandless

I'm not sure Indeed provides the best picture for Software Engineering hiring but assuming this is directionally correct... Hiring is recovering.

That said, the structure is clearly changing.

Leaner teams.
Higher expectations.
More output (and quality, overall leverage) expected per team member.

And increasingly: Tiny teams, by design.

“Small teams are what happens when you decide to fix the actual problem.”

— Natalia Amorim

For example… PostHog has 43 small teams shipping across products.

Their philosophy, according to this post:

• Make speed the norm.

• Give real autonomy, not the illusion of it.

• Let teams own products, not just problems.

• Optimize for long-term solutions, not short-term patches.

AI doesn’t just increase individual leverage. It makes small, empowered teams structurally advantaged.

The new question isn’t ‘How big should we scale this team?’

It’s ‘How long can we keep this team small?’

Subscribe now

3/ The Rise of the AI-Enabled Builder PM

“The PMs who thrive will be the ones who treat AI the way the best PMs treated SaaS: Not as a threat or a replacement, but as a capability expansion that lets them focus on the parts of the job that are most distinctly human: judgment, empathy, strategy, and leadership.”

— Andrew Vincent

Andrew outlines three PM archetypes:

• Order-Taker

• Strategic PM

• Builder PM

His prediction:

• The Order-Taker shrinks dramatically.

• The Strategic PM becomes more demanding.

• The Builder PM emerges as a distinct archetype.

“SaaS changed delivery; AI changes creation.”

The Builder PM doesn’t just write specs. They prototype, automate, and test ideas. They collapse the gap between thinking… and shipping.

Wealthsimple is leaning into this shift — their new ‘AI builder program’ requires applicants to submit a working AI system as part of the process.

Not a deck.
Not a resume.
A working system.

“We’re putting together a small, high-bar team of builders who think in systems, move fast in ambiguity, and care deeply about where AI should — and shouldn’t — replace human judgment.”

— Dasha Sivkov

The bar’s moving.

Judgment, empathy, decision-making, and strategy matter.

But the ability to build is now table stakes.

Typos and imperfections are OK… You better move quick, and iterate.

Subscribe now

Bonus: Stripe’s Vision for Agentic Commerce

“There’s no forecasting exactly where agentic commerce will be by the end of 2026, but it’s clear we’ve already moved well beyond pure hype into a phase of building and real-world experimentation. The pace of change will likely only accelerate from here.”

— Stripe Annual Letter

In their annual letter, Stripe’s Patrick & John Collison outlined five levels of ‘agentic commerce:’

1. Eliminating web form

2. Descriptive search

3. Persistence

4. Delegation

5. Anticipation

Some notable data points:

• $1.9T in total Stripe volume (1.6% of global GDP), up 34% YoY

• Ecommerce up 30% (inflation-adjusted) vs. 5% for US brick-and-mortar

• Stablecoin payments doubled to ~$400B, 60% B2B

• GitHub pushes up 41% between Q3 2024 and Q3 2025

• 41% increase in company formations

• Funding volume up 45% YoY

We’re not in the ‘AI is coming’ phase anymore.

We’re in the ‘AI is reorganizing how products are built, teams are structured, and commerce is executed’ phase.

Here’s what got my attention this week — a list of interesting things I read and watched, as a product leader.

1/ The Power of Saying No in an AI Gold Rush

“What if the best product decision is saying no to what everyone else is building?”

In an episode of Supra Insider, Marc Baselga and Ben Erez sat down with Alexander Danilowicz, founder and CEO of Magic Patterns.

Magic Patterns is one of my favorite products right now and I appreciate their ability to focus — in their case, it’s on frontend development, prototyping and design systems.

While other AI prototyping tools like Replit, Bolt, and v0 are rushing to add backend capabilities, Magic Patterns is deliberately refusing to do so.

The episode gets into product quality, DNA, risk, feedback, presets, integrations, dogfooding, prompt engineering, and strong leadership relationships.

If you’ve ever wrestled with feature pressure that doesn’t align with your business vision/mission/strategy — even amidst a wild AI gold rush — this one’s worth your time.

Sometimes, the best decision is saying no to what everyone else is building. That’s if you get it right, of course.

Subscribe now

2/ Where AI Dollars Are Actually Flowing

ICONIQ Capital released a 2025 State of AI report — surveying ~300 execs (CEO, Head of AI, Eng, Product, CRO, CFO) at companies shipping AI in production.

A few takeaways:

• Nearly 80% of AI-native builders are investing in agentic workflows, or autonomous systems designed to take multi-step actions on behalf of users.

• 70% of companies are building vertical AI apps.

• More than one-third (37%) plan to adjust their pricing in the next year.

• Teams are using ~3.1 model providers on average — multi-model is cleary the norm.

• AI/ML engineers take the longest to hire of any AI-specific role, with an average time-to-fill exceeding 70 days.

• Differentiation is happening at the application, workflow, data and model layers.

Subscribe now

3/ Give AI Its Own Machine

If we’re serious about AI coding agents, we may need to give them their own machine.

Not a tab or side window, but a dedicated environment. Pushing the limits of delegation and automation, beyond assistance and autocompletion.

That’s what DDH and others have been asking in response to OpenClaw.

“With OpenClaw you're giving AI its own machine, long-term memory, reminders, and persistent execution... It's a sneak peek at a future where everyone has a personal agent assistant, and it's fascinating.”

— David Heinemeier Hansson, 37Signals

The question for product leaders:

What UIs, and constraints are we designing and building for? And what do we need to align to, anticipate, and get ready to reject or rethink? Apps, systems, copilots, agents, best practices, processes, team sizes, business models, etc. It’s all going to look very different by the end of this decade.

“I don’t know what I did.”

— Peter Steinberger, OpenClaw

Subscribe now

Bonus: The myth that search is dying

Graphite partnered with Similarweb to analyze ~40,000 of the largest U.S. sites to evaluate the trend in organic search traffic. Their conclusion is that the prevailing narrative that SEO traffic has dramatically declined is false.

“The Myth: Organic traffic from Google to websites is down dramatically (between -25% to -60%). The Truth: Organic traffic from Google to websites is down slightly (-2.5% YoY).”

— Graphite

Here’s what got my attention this week — a short list of things I read and watched recently that felt worth pausing on, as a product leader.

This week had a clear theme: power is shifting between roles, tools, and capital.

1/ Debates over the future of the CPO role

With the rise of AI, a question that may have felt fringe a year or two ago is now being discussed openly on LinkedIn:

Could the CPO role disappear within five years?

“PREDICTION: THE CPO ROLE WILL VANISH IN FIVE YEARS” — Gokul Rajaram

After saying this, the comments came in — some nodding along, others pushing back hard. What matters isn’t whether the prediction comes true. It’s that the conditions making it a worthwhile questions are already here.

AI is compressing the distance between strategy and execution. Teams are getting smaller. Builders are closer to customers. And leadership roles that exist primarily to coordinate, translate, or arbitrate are under pressure.

Aakash Gupta, unlike Gokul, argues the CPO role is far from dead:

“The companies winning will have CPOs with broader scope, not narrower. The title might evolve. The power concentrates.” — Aakash Gupta

Personally, I don’t think product leadership is going away — but its shape is most definitely changing.

Lines will blur. Some teams will over-index on technical leader or business credibility. Others will over-index on judgment, taste, and decision-making. CPOs, CTOs, CPTOs, and anyone aspiring to lead product vision, strategy, and execution need to embrace the shift and get good at the balancing act. Art and science. Speed and results. Customers and stakeholders. The forrest, and the trees.

In a way, it’s nothing new under the sun. Just less room for hesitation.

Carpe diem. C'est la vie.

Subscribe now

2/ Codex, Opus, and the AI Super Bowl moment

This week felt like an AI Super Bowl.

OpenAI released a new version of Codex, positioning it as another major step forward for AI-assisted software development. The release is another escalation in the AI arms race.

“Is this a sign that artificial intelligence is becoming self-improving?” — Timothy Beck Werth, Mashable

According to Mashable, OpenAI said that the new GPT-5.3-Codex model is its ‘first model that was instrumental in creating itself.’ ChatGPT wouldn’t confirm or deny.

Meanwhile, Anthropic announced Claude Opus 4.6 at around the same time:

“Drop what you are doing. Opus 4.6 is out. And it officially puts the “ClawdBot” era to shame.” — John Peslar

Anthropic also shared a multi-ad campaign of Claude ads for the Super Bowl. Maybe its the recency and relevance, but it felt like one of the most memorable competitive tech campaigns since ‘Get a Mac.’

“Anthropic absolutely killed OpenAI in their Super Bowl ad. Their positioning in the AI race has been immaculate: Thoughtful, with a human touch, and now we can add humor to that list too.” — Willem Evers

The benchmarks and feature lists continue to impress, raise questions, and draw attention in their own right, but that’s to be expected.

What’s more interesting is that we’re also seeing a shift in tone. There’s increased competition, and the pie seems to be getting bigger, and people are taking sides. Beyond the model comparisons, I’m noticing less questions about trust for AI in general, and more questions around trust for each of the models/companies/brands. Questions like…

Which model do you want open all day?
Which one do you reach for first?
Which one fits how you actually think and work?
Which one can you see doing the right thing?

The competition isn’t just about who’s smarter and faster. It’s also about what feels dependable, and trustworthy.

Subscribe now

3/ Global VC investment is surging again (with AI leading the way)

Global VC investment surged to $55 billion at the start of 2026, with AI accounting for a significant share of total funding:

“Funding last month more than doubled from $25.5 billion a year earlier and was up over 50% from December…. A total of $40.9 billion (74% of all funding) went to rounds of $100 million and more, and $31.7 billion (57% of funding) went to AI-related companies.”
— Gené Teare, Crunchbase

Most of the dollars flowed into late-stage rounds and large, already-scaled AI companies and perceived winners. For example, mega-rounds for players like xAI.

Investors are doubling down on companies they believe can dominate foundational layers, platforms, and distribution. While the funding environment is improving, the bar for ‘fundable’ is higher, not lower. Advantage compounds fastest for companies with great momentum.

Subscribe now

Bonus: Getting paid to vibe code

Vibe coding is now a full-time professional career, and not having a technical background can actually be an advantage — as Lazar Jovanovic, from Lovable and Starter Story, can attest.

He spoke with Lenny Rachitsky about how coding is no longer the bottleneck… Clarity is. He also shared tips on writing PRDs, debugging, being deliberate with AI tools, studying elite design work, and building in public.

Here’s what got my attention this week — a short list of things I read and watched recently that felt worth pausing on, as a product leader.

1/ The Adolescence of Technology

Dario Amodei, CEO of Anthropic, and former research leader at Open AI and Google Brain, shared an essay that reads less like societal speculation and more like a warning label.

“Humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it.”

Amodei’s core idea is pretty simple, and a little unsettling. Whether we like it or not, we’re racing toward a form of ‘powerful AI’ that looks less like a tool and more like ‘a country of geniuses in a datacenter.’

He writes about surveillance, propaganda, identity, economic shocks, and power concentration.

The upside could be extraordinary. The downside isn’t inevitable, but it is civilizational in scale. Unlike the emergence of the web (or web 2.0… and web 3), AI feels different: We may be moving faster than any of our institutions—governments, companies, norms—can realistically adapt.

He doesn’t argue for panic or doomerism, but he does warn against denial, complacency, and the false comfort of ‘we’ll figure it out later.’

Now, what really caught my attention is how consistent this framing has become across all AI leaders. From Elon Musk to Sundar Pichai, Sam Altman and Jensen Huang, they’re all saying some version of the exact same thing: AI is both the threat, and the solution.

On one hand, that’s obviously self-serving.

On the other, they genuinely seem uneasy and even puzzled at times about what their combined efforts are unleashing.

Some of you might be wondering what any of this has to do with product. I have way more thoughts there than fit in this week, but this quote is the one that stuck with me:

“AI will be capable of a very wide range of human cognitive abilities—perhaps all of them.”

If that’s even directionally true, and we’re heading to a world where AI does most of the coding (and a lot of the ‘thinking’)… then software as we know it is in its awkward teenage phase.

Coding gets cheap. Teams shrink. Learning cycles compress to something we’ve never seen before. The advantage shifts from ‘who can build’ to ‘who can decide well.’ Orchestrators over implementers. Taste over labor.

More speed, leverage, and impact per person than ever.

But there’s a real flip side. If models can handle end-to-end software engineering and large portions of white-collar work, we don’t just get productivity gains — we get a massive surplus of labor. That pressure shows up everywhere: Wages, roles, expectations, org design, culture, investment math. And power concentrates quickly in the hands of the companies best positioned to deploy and control these systems.

I’m not convinced AI is our greatest threat.
I’m not convinced it’s the solution either.

But one thing is undeniable at this point: Ignoring it is no longer an option.

Subscribe now

2/ Where does the IDE go from here?

This AI Engineer Code Summit talk by Steve Yegge & Gene Kim echoes some of the same arguments as Dario’s essay — in terms of how quickly AI’s advances are reshaping software development. Vibe coding is making the process of building and shipping increasingly faster, more ambitious, and arguably more fun.

Ironically, Steve emphasized that he loves Claude Code and uses it 12 hours a day, but he also claims it’s ‘not it.’

“Claude Code is like a drill or a saw. How much damage can an untrained person do with a drill or a saw? How much damage can you do as an untrained Engineer with Claude Code? It’s very similar. You can cut your foot off. But you can also be really really skilled with it, and do very precise work — like a craftsman.”

Claude Code, Replit, Cursor and Codex are among many tools actively rethinking software engineering. Steve predicts that within the next year, AI will be able to write nearly all the code—and that the primary interface will be a UI, not a command line, and not an IDE.

Now, the IDE probably won’t just die or disappear. But there’s a compelling case that it might just dissolve into the model. Regardless of timing, a real shift is brewing here—and to Amodei’s point, complacency isn’t advised.

“This is a lot like what happened with the Swiss mechanical watch industry…”

Subscribe now

3/ From a decade of FAANG to a decade of AI

Alex Turnbull, CEO of Groove, a company behind multiple AI-powered tools including Helply and InstantDocs, posted a note on LinkedIn about compensation levels at FAANG vs AI leading companies.

“The engineers who want to build the future are going to the companies building it. The ones who want stability are staying at FAANG. Both are valid choices. FAANG had its decade and this is AI’s decade now.”

The framing isn’t perfect—FAANG no longer needs to pay the same premiums, and Alphabet in particular is clearly positioning itself as an AI leader. Still, the takeaway is hard to ignore, and especially timely… even if it’s a little uncomfortable.

Subscribe now

Bonus: Lovable Downgrade Masterclass

If you’ve read this newsletter you probably know I’m a fan of Lovable — as a product, brand, user experience, and team.

That’s why I had to fully agree with Yaakov Carno, founder of PLG agency Valubyl, when he recently wrote on LinkedIn about their approach to the downgrade flow:

“Whatever happens with these monetization experiments, one thing is clear: This is a company that’s obsessed over what their users want and need, and it shows in every part of their product.”

Ed Biden is the Founder of Hustle Badger. He was formerly CPTO at FYLD, CPO at YunoYuno, SVP of Product at Jobandtalent, CPO at FutureLearn, Director of Product at Depop, Director of Product at Rocket Internet SE, Product Lead at Wooga, and PM at Kobojo.
Website / LinkedIn

EC: How can product leaders get good at AI building?

EB: Even in the last few weeks it feels like we’ve hit a new inflection point with AI:

• Boris Cherny built Cowork entirely with Claude Code in about ten days.

• A CPO I know is hiring senior PMs with the explicit expectation that they’ll ship code themselves

• The Pragmatic Engineer is reporting about teams going all in on AI … and how some devs aren’t engaging with this.

2026 is going to be chaotic and exciting wherever you work in product. None of us have ever seen change this fast. No one knows exactly where this goes, but it’s clear we’re still very early in the AI adoption cycle and things are still accelerating.

Naturally, people are nervous.

I expect roles to blur and titles to change, but fundamentally there will always be a demand for smart people with two instincts: customer empathy and commercial sense. If you’ve got those, you’re going to be ok.

So what can product leaders actually do about this?

First, recognize that this is your job, you can’t delegate it, and it’s existential for your career. This is a proper transformation and you need to lead from the front. You’ve got to be experimenting with the tools and exploring what’s possible and right for your organization.

Then you’ve got to create the conditions for meaningful change across the organization. You can’t just tell people to ‘use AI more.’ You’ve got to do this holistically, and there are four ingredients:

1. Role modeling. Leaders need to be using AI tools regularly — and visibly. If you’re not demonstrating the commitment to changing how work gets done, why would anyone else take it seriously?

2. Tool access. Teams need access to the right tools, which in practice means someone very senior pushing things through security and procurement. This is often the hardest part, but it’s non-negotiable.

3. Upskilling. Teams need to know how to use these tools well, and that usage needs to be established as the default way of working through job descriptions, hiring processes, competency ladders, communities of practice, and so on..

4. Clear goals. Why are you doing all of this? You need a clear idea of what success looks like. The more clarity you have on the outcome, the better your chances of getting there.

Subscribe now

EC: What do PMs need to get right to become a product exec?

EB: The jump from VP to CPO is as big as going from IC to manager. It really caught me off guard when I made it.

There are roughly four levels in a PM career:

1. Individual Contributor (APM, PM, Senior PM) - working in cross-functional teams to ship features

2. Manager (Head of, Group PM) - hiring and coaching other PMs, dealing with underperformance

3. Functional lead (Director, VP, SVP) - defining what product management means at your company, setting overall strategy, allocating teams against high-value opportunities

4. Executive (CPO, CPTO, CPMO) - running multiple functions, representing them on the exec team and to the board

The things you should have mastered at the VP level become table stakes as a CPO: Setting company strategy and vision, designing and running the product operating model, and cross-functional collaboration. But there are several new challenges that are fundamentally different.

The new challenges you’ll face:

You’ll have three very different stakeholder groups to manage: The people reporting to you, your exec peers, and the board. These are all high-stakes relationships that require completely different approaches.

CEOs want execs to remove a top five anxiety from their plate entirely. You need to identify what’s keeping them up at night and own it completely.

You’ll be managing non-product teams, which means you can’t fall back on your own technical expertise. You need to be a master coach.

You won’t just be dealing with product problems—you’ll get dragged into fundraising, M&A, and major company initiatives that require robust and flexible problem-solving skills.

One thing that surprised me: You’ll have VPs reporting into you, and these are serious, accomplished people. If you ever had the mistaken notion that your reports somehow needed to be ‘parented,’ that will be immediately knocked out of you.

To get a foot in the door for a CPO role, you’ll need:

• Credibility - the right logos on your CV and appropriate seniority

• Scale of team managed - comparable to what you’re being hired to manage

• Non-product teams managed - demonstrates you can abstract your leadership principles beyond product

• Very clear commercial impact - evidence you’ve really made things happen, not just been along for the ride

Then you’ll be tested on:

• Fit with the other execs and whether you fill gaps the current team has

• Whether you can solve the specific problem the company faces right now

• Strategy and execution capabilities

• Leadership aligned with company culture

• Domain expertise (overweighted in my opinion, but common)

But here’s what I’ve observed about the most effective people: they continuously take on hard, ambiguous problems with an ‘I’ll figure it out’ attitude, and they don’t worry about titles, team sizes, or internal career ladders.

If you take that approach and focus on solving real problems rather than checking boxes, you’ll be rewarded regardless of whatever career progression framework your company has defined. It’s uncomfortable because as humans we crave certainty. We want to see what’s ahead. But spending energy trying to predict exactly what AI will do to your job function in three years so you can perfectly plan your next move is a waste of time.

Take that energy and deploy it to being at the frontier of using new technology in your profession. You’ll be much better positioned to navigate whatever changes come.

Share

EC: How can product leaders drive continuous delivery that leads to great outcomes?

EB: This is the whole problem you’re trying to solve as a product leader!

In this case, you need to understand that you manage a value chain. The product team takes in company objectives, converts them into customer problems to solve, solutions to build, and then finally outputs working software that delivers those results.

Your job as a product leader is to understand how this value chain works, and to optimize it as a whole.

That means having visibility at the various steps along the way (e.g. through product reviews or delivery updates), and then seeing where you optimize things for even better results.

You are a systems designer and you can’t work on everything at once. You’ll need to look at the overall software development life cycle (SDLC) and treat this as you would a product. You understand the flow, where there are big drop offs, and then you make targeted interventions to resolve these.

In practice, I try to check in on teams at critical points in the SDLC, so I have an overview of value creation from the product org as a whole:

• OKRS: What are their objectives? What customer problems are they focused on?

• Kick off: What features are they going to build?

• Design review: What will the new features look like?

• GTM review: What are the launch plans for the new features?

• Impact review: What actually happened when the feature was launched?

That said, these classic phases are starting to collapse as the speed of development increases, and product leaders are going to have to rethink what their SDLC looks like in an AI world. Where they get transparency, and how they accelerate teams is going to change.

Here’s what got my attention this week — a short, curated list of things I read and listened to that felt genuinely worth pausing on, as a product leader.

1/ Welcome to the Era of Agent-Led Growth ⚡️

On TBPN, John Coogan and Jordi Hays spoke with Sequoia Partner Sonya Huang, who argues that after moving from sales-led growth to product-led growth, we’re now entering an era of ‘agent-led growth.’

She pointed out that tools like Claude Code already exert real influence over tech stacks and infrastructure decisions. LLMs — through ‘vibe coding’ workflows — are actively selecting or recommending services like Supabase and Vercel.

She predicts this dynamic won’t stop with software tooling:

“I think this is going to happen with the entire economy. Your ChatGPT is going to tell you ‘Hey, this is the travel you should be booking…’ Your agent has infinite time so it can read all the documentation, all the user comments, and figure you what you need for your use case.”

If this plays out, distribution, differentiation, and brands will increasingly be mediated by agents. Hold on to your seats.

Subscribe now

2/ Does your SaaS Make You Smile?

There was a reality show on Netflix a few years ago that made Marie Kondo a household name, built around one simple question: Does it spark joy?

That question came rushing back as I read a SaaStr post by founder and CEO Jason Lemkin. SaaS is under pressure from multiple directions right now, but his framing was clear: you can still fight back.

The under-appreciated weapon? Making customers smile.

‘Surprise and delight’ feels like a capability the industry quietly deprioritized for nearly a decade — but that feels like a mistake now. AI-native experiences like ChatGPT, Lovable, Suno, and Cursor feel automagical in a way we haven’t seen since the early days of Uber, Spotify, Waze, and Airbnb.

I continue to believe B2B has to borrow far more from the consumer playbook.
Your SaaS users are humans first — and consumers and prosumers of beautifully designed apps that relentlessly compete for their attention.

Slack internalized this earlier than almost anyone. Tomorrow’s best SaaS products may need to find their inner Slack — even if that instinct isn’t already in their DNA.

Jason’s post also covered other important pressures in SaaS today: Shrinking seat counts, AI-driven budget reallocation, painfully slow teams, TAM traps, and more.

“When users can build functional apps in minutes (Lovable), or get meaningful output in seconds (Claude), the bar for what constitutes a ‘great product’ has permanently shifted.”

SaaStr AI: How To Sell, Scale, and Win

The 6 Threat Vectors Killing Traditional B2B Software in 2026 (And How to Fight Back)

“Can your team ship AI features that make you truly smile? If you don’t smile — your customers won’t either…

Read more

3 months ago · 3 likes · 1 comment · Jason M. Lemkin 🦄

Subscribe now

3/ Consumer is Making a Comeback

I somehow missed this conversation when it first dropped, but once I found it, I couldn’t stop listening.

Y Combinator CEO Garry Tan spoke with Michael Mignano — co-founder of Anchor (acquired by Spotify) and now a partner at Lightspeed.

They covered the resurgence of consumer apps, creative tools, distribution dynamics, and what AI unlocks — including Oboe, a particularly interesting take on AI-powered learning and self-help.

“We have AI now, and we have all these new opportunities — things we can create that we previously couldn’t… Increasingly, we find ourselves betting on people who are just great at building products.”

That last line stuck with me. Tools change. Distribution changes. But great builders still matter most.

Subscribe now

Bonus: Great free products = Growth engine

On the theme of education, Duolingo Founder & CEO (and fellow Guatemalan) Luis von Ahn shared a refreshingly clear post on LinkedIn about this focus this year.

“Making the free version of Duolingo the best it’s ever been. This is how we grow. A great free product drives word of mouth and, ultimately, subscriptions.”

It’s a simple reminder — and an increasingly rare one — that generosity in product design can still be an strong differentiation and growth strategy.

Here’s what got my attention this week — a short list of things I read and watched recently that felt worth pausing on, as a product leader. The common thread this week is that the fundamentals are critical, especially now.

1/ SaaS valuations are officially back on Earth 🌍

The latest SaaS Valuation Multiples (2015–2025) report from Aventis Advisors confirms what many of us have been sensing for some time.

The SaaS boom — launched into outer space during the pandemic — is no boom anymore.

Three signals that matter:

• Multiples have dropped sharply.
  2016 called — and it wants its valuations back. Median public SaaS EV/Revenue multiples have fallen from the 2021 fantasy zone (20–30×) to roughly ~5× today.

• Growth ≠ gravity-defying anymore.
  Median SaaS growth is settling around 11–12% YoY. The market is rewarding durable growth, not ‘growth at all costs.’ Nothing lasts forever.

• Profitability is no longer optional.
  Margins are improving, Rule of 40 is trendy (but tough), and efficiency is cool again. The narrative for a SaaS business needs to be strong — and it needs to be real.

Why this matters for product leaders:

• Build products customers need — and won’t churn from

• Design business model, monetization, and tangible differentiation very deliberately

• Craft products (and teams!) for trust, efficiency and effectiveness, not excess

In some ways, it feels like a different universe than just a few years ago when your mom and my mom were downloading Zoom and everyone was buying SaaS stocks.

I suspect it will get tougher before it gets easier for some SaaS businesses.

The good news? We needed a back-to-basics reality check back then. Today, after a much needed wake up call, we can probably say SaaS is healthier.

And the companies that get it right might come out 10× stronger.

“Bad companies are destroyed by crisis. Good companies survive them. Great companies are improved by them.”

— Andy Grove

Subscribe now

2/ LLMs don’t replace marketplaces — they pressure-test them

This essay from Dan Hock explains where LLMs are a real threat to marketplaces — and where they aren’t.

Marketplaces are most exposed when:

• Supply is easy to aggregate (standardized data).

• The product stops at search or comparison.

• The purchase is high-consideration but low-frequency.

In these cases, LLMs sit in front of the marketplace. They don’t just ‘capture value’ — they break the business model by cutting off the free repeat traffic marketplaces rely on to be profitable.

“Superior search combined with the ability to transact lets AI systems like ChatGPT push further into the marketplace stack than Google ever could — doing much of what transactional marketplaces already do.”

— Dan Hockenmaier

That’s especially bad news for Expedia-style businesses.

• Travel is high-consideration (requires research).

• Inventory is widely available.

• Differentiation lives mostly in search.

Exactly where LLMs are strongest.

Dan’s draws a clear line on how to survive the AI threat:

• Go Deep: Manage the messiness of the real world.

• Go Frequent: Build habits that bypass search entirely.

The defensible ones:

• Underwrite transactions: Manage returns, guarantees, and financial risk.

• Own the physical layer: Build logistics and fulfillment (LLMs won’t drive cars or flip burgers).

• Drive frequency: If usage is high-frequency and low-consideration (like Uber), users won’t ask a chatbot… they’ll just tap the app.

• Build AI search inside the marketplace: Invest in AI within.

If a marketplace is mostly search, AI will eat the surface. If it runs operations or owns the habit, it still has a moat.

Dan Hock's Essays

LLMs vs. Marketplaces

LLMs are on a collision course with marketplaces. The Verge calls it The DoorDash Problem…

Read more

4 months ago · 20 likes · 8 comments · Dan Hockenmaier

Subscribe now

3/ Why AI hasn’t really hit Main Street yet

On ‘Mostly Growth,’ I watched CJ Gustafson + Kyle Poyar‘s conversation with Daniel Lang, CEO of Mangomint, on why AI still hasn’t meaningfully landed for SMBs.

His point is simple: Real-world SMBs don’t care about AI. They care about what the product actually does.

"In the real world, like with salons and spas, they don't care what's on the label. As long as you solve a real business problem for them, as long as you deliver value... The real economy on Main Street America, they don't care if it's AI or something else."

Just like you instantly know what ‘Dave’s HVAC’ does from a van, software needs to be clear. It needs to be obvious, not impressive.

Lang argues that most standalone AI wrappers are dead on arrival because they treat the system of record as an afterthought.

Instead, he describes a contextual, ‘layer cake’ approach to vertical SaaS AI:

• System of record first — own the core data.

• Communication layer next — be where decisions happen (texts, calls, chats).

• Agents last — only then does AI become helpful.

Without deep access to historical data and metadata, AI agents behave like a junior employee on day one — OK at the basics, pretty bad at nuance, and incredibly risky in real customer interactions.

Mangomint earned trust before AI, without contracts, data lock-ins, and relentless focus on ‘boring’ fundamentals and high quality workflows. What worked for them might seem contrarian and may not work for everyone, but there are lots of insights and hot takes on this one:

Subscribe now

Bonus: In 2026, AI will start to prompt you more than you prompt it

That’s what Gumloop’s CEO Max Brodeur-Urbas posted on LinkedIn this week.

Ps. I’m glad this post got your attention! If you made it this far, please let me know if this is valuable to you. Feel free to send suggestions for future posts. You can find me on LinkedIn.

Here’s what got my attention this first week of the new year — a short list of things I read and watched recently that felt worth pausing on, as a product leader.

1/ The next big things in 2026, curated by NBT

This week’s ‘next big thing’ Nikhil Basu Trivedi essay reminded me how much the AI boom is shifting from raw capability to product outcomes. The piece aggregates predictions from operators and investors at the frontier of AI innovation — spanning autonomous agents that own workflows to ‘taste graphs’ reshaping personalization, and even AI companions that clearly ‘remember you’ rather than just respond to prompts.

2026 isn’t simply about more powerful models. It’s about meaningful product leverage, where memory, proactivity, and business impact become the basis for real adoption and defensibility.

“The next big thing in 2026 will be an escalation in the battle between Systems of Record companies and AI agent companies.”
— Gokul Rajaram, Founding Partner at Marathon Management

“The next big thing in 2026 will be autonomous computing; AI that doesn’t just respond but operates.”
— John Milinovich, Head of Product, GenAI at Canva

next big thing

The next big thing in 2026 will be...

Welcome to issue #62 of next big thing…

Read more

4 months ago · 25 likes · 2 comments · Nikhil Basu Trivedi

Subscribe now

2/ Grant Lee: The emotional job-to-be-done is the product

Grant Lee told a story every product leader should re-read once a quarter: An early customer call where the ‘why’ wasn’t functional at all… It was identity. The user loved Notion because it ‘makes me feel smarter,’ and that reframed the whole product lens for Gamma: Not ‘what does it do,’ but ‘who does it help me become?’

This is one of those deceptively simple insights that upgrades how you run discovery. If you’re only capturing tasks and pain points, you’ll miss the part that creates obsession and word-of-mouth: The self-image the product reinforces.

“Products don’t just do things for people. They make people feel a certain way about themselves.”
— Grant Lee, CEO at Gamma

Subscribe now

3/ Stephanie Leue: ‘Physics of the business’

Stephanie Leue lays out a helpful and clarifying framing many product leaders need to hear: When the underlying business is sales-led, no amount of PM craft can override the operating model. If the company is fundamentally sales-led in certain ways, PM craft alone can’t will you into ‘product utopia.’ She maps business models into zones where the ‘textbook product model’ either survives or gets crushed by contractual obligations and survival-driven custom work.

The takeaway isn’t ‘sales-led bad.’

It’s this: Be honest about which game you’re in.

If you’re in a feature-factory dynamic, pretending you’re in a retention-led, outcome-driven world creates frustration for literally everyone, especially PMs who are operating under different laws of motion.

“You can’t ‘mindset’ your way out of a Sales-Led business model.”
— Stephanie Leue, CPO, Ringier

Subscribe now

Bonus: Gagan Biyani on growth ‘Coming back’ in the AI era

On this LinkedIn post, Gagan’s point is basically that Growth didn’t disappear — it collapsed into product, and a lot of it devolved into paid marketing. But AI re-opens the playground: Faster shipping, cheaper experimentation, and new surfaces (e.g. AEO, app-store-like distribution, etc.) create room for the original spirit of growth again: Weird loops, channel invention, and systems-level creativity.

The bottleneck is no longer ‘can we ship?’ — it’s ‘Do we have taste + a thesis for what’s worth testing?’

“Now in 2026, I believe growth hacking will make a strong comeback… There are tons of fresh opportunities for growth marketers to play with and innovate on. The intersection of data, product, and marketing is going to have a rebirth.
— Gagan Biyani, CEO and Co-Founder at Maven

On the final week of 2025, what got my attention as a product leader were a few videos and posts that seemed to orbit around the same underlying tension — coming from very different angles.

What kind of tension? Speed vs. depth. Tools vs. craft. Experimentation vs. commitment.

Let’s get into it.

1/ Webflow’s CPO on how she AIs

On the most recent episode of ‘How I AI,’ Claire Vo interviewed Webflow’s CPO, Rachel Wolan.

Rachel exemplified the emerging ‘IC CPO’ mindset — sharing how she built her own ‘AI Chief of Staff,’ to manage calendar, email, and event prep, as well as how Webflow runs internal ‘builder days’ to level up AI usage across the company.

What I appreciated most was that while there are plenty of practical workflows, this is not the typical ‘AI does product for you’ hype. The core thread is that being an AI‑native exec means being in the weeds, building personal software, and using it to improve your performance. Sometimes that means getting brutally honest feedback on how you spend your time.

In this new era of ‘IC CPOs,’ it's also important to lead by example. As Rachel puts it,

“I really want my team to level up. It would be inauthentic for me to tell them, ‘Hey, you need to prototype with AI’ if this isn’t something that I’m doing every single day.”

With so many people focused on what AI can generate for them, it’s a good reminder that the hard part of product is still judgment: where you show up, what you delegate, and what you say no to.

Subscribe now

2/ Descript’s CEO on building AI features

Aakash Gupta had Descript CEO Laura Burkhauser on the Product Growth podcast this week. The conversation was a good reminder that speed for the sake of speed can quietly flatten ambition. The fastest path is often ‘slap AI’ on your product, but you lose the chance to imagine the product your users actually need.

Descript has been an AI‑native product since inception, and during the AI boom the team kept asking,

‘What do we do now that LLMs are generally available to bring into Descript?’

Laura kept returning to the idea of mapping the user journey first. That discipline let the team solve the right problems, roll out valuable and differentiated features, and design the right interface — sometimes a single button, sometimes a full agent — instead of chasing whatever AI pattern was trendy that week.

Laura also got into her IC-to-CEO progression, evals, and how they measure and quantify success for AI features like tools and Underlord.

Subscribe now

3/ Kyle Poyar on pricing agility, and the coming backlash to AI credits

I’ve always appreciated Kyle’s take on all things pricing. In a recent LinkedIn post, he mentioned that while many are predicting 2026 will be the year of credit-based pricing, he thinks it may also be the year of a backlash against credits.

If 2025 was about scaling AI pricing experimentation, 2026 should be about consolidation, predictability, and trust.

Enterprises don’t hate usage-based pricing. They hate surprises. The same is true for vibe coders and solopreneurs — nobody wants unexpected fees, hidden costs, or pricing traps.

Using Salesforce as an example, Kyle points out the many flavors of AI credits: pre-purchase credits, pay-as-you-go, and pre-commitments with no upfront payment.

He also makes the point that seat-based pricing isn’t going away:

“If AI replaces people with digital labor, a seat-based model is a recipe for a declining business… But AI replacing people is still a relatively small phenomenon. Most adoption is still about AI augmenting people.”

This experimentation vs. commitment tension will continue to play out across AI products. One example I can think of is Lovable’s recent tests with ‘top up credits’ — which I suspect will be followed by their competitors, if they haven’t already.

Subscribe now

Bonus: Linear CEO Karri Saarinen on design-to-code being a trap

I couldn’t end without pointing to this post by Linear CEO Karri Saarinen, which got lots of attention on LinkedIn.

Karri argues that designing in code too early doesn’t elevate design — it compresses it. When designers become builders ‘too soon,’ they start making conservative bets, optimizing for feasibility instead of vision.

The entire post is a great read, but here are some of the highlights:

“Whenever a designer becomes more of a builder, some idealism and creativity dies. Not because building is bad, but because you start out including constraints earlier in the process than they should.”

“Once you become the architect and the builder, or the designer and the developer, you start making more conservative bets.”

“Designing in code is just a path to local maxima and ruin.”

Great design (and great product thinking) often requires a phase where reality is intentionally suspended. Not ignored forever… just postponed long enough to dream something worth building.

ps.

A good PM friend asked me if I was going to write a post about my predictions for 2026. I told him I wasn’t planning on it — the future seems harder to predict these days.

That said, the tension I mentioned above is one I think will continue in 2026.

As tools get faster, cheaper, and more automagical — and as the FOMO increases — teams will continue to optimize for speed. Lines will blur. That will do wonders for productivity and outputs.

At the same time, winning products will always need strong vision, judgment, and risk-taking. Whether it’s AI in product, code in design, or credits in pricing, the real work keeps moving one layer up.

In 2026, play your role wisely.

Happy New Year.

Here’s what got my attention this week — a short list of things I read and watched recently that felt worth pausing on, as a product leader.

1/ State of Consumer AI 2025: Hits, Misses, and What’s Next

This post by Olivia Moore and Justine Moore at Andreessen Horowitz was a nice read on consumer AI. Instead of just chasing after hype cycles and the capabilities of the latest models — a16z focused on on actual usage, repeat behavior, and where consumers seem to be spending meaningful time and money.

The core takeaway is that consumers are curating: They’re choosing fewer, more opinionated tools that solve very specific problems extremely well. Broad assistants are table stakes. Narrow, workflow-specific products are where increased signals for loyalty are showing up.

a16z frames AI maturity and differentiation as a clear product-design problem, not a model arms race. This may be a useful reminder as roadmaps keep filling up with ‘Generic AI everywhere,’ whether that makes sense for their customers, or not.

“The next generation of breakout AI products won’t be the most capable — they’ll be the most focused.”

The post also points to the potentially very large white space for founders building dedicated AI customer experiences as we go into an era of greater AI-driven distribution — referencing the success of apps like Gamma, Suno, and Eleven Labs.

”Model companies don’t have the intuition or, quite frankly, the attention and resources to be innovating outside their areas of core competency. Consumer founders may even be able to benefit from ChatGPT’s upcoming push on discovery and distribution for third-party apps.”

Subscribe now

2/ Boston Dynamics Will Debut New Atlas Humanoid Robots at CES

The last time I attended CES, I was with a little company called Samsung. The floor was filled with Smart TVs, Smart Refrigerators, Smart Miscellaneous. It seemed like it was all Internet-of-Things hype, and fun prototypes like bendy screens, accompanied by an unnecessary amount of iPhone and Android covers. I remember a couple of cutesy robots, but I don’t remember any human-like C3P0 robots doing parkour.

At this upcoming 2026 CES, Hyundai — another South Korean giant — will remind us that their Boston Dynamics division has been working non-stop on building for a future where AI robots do all kinds of wild things. They’ll be showing off their next-gen Atlas humanoid robots. Some people will see this as a dream come true for the future of manufacturing and general-purpose, manual labor. Others will see this as an unpredictable nightmare of ‘Black Mirror’ proportions. The truth is somewhere in the middle. At any rate, Boston Dynamics is not asking for anyone’s permission.

The steady progress in AI-driven robotics is shaping how people imagine what ‘intelligent systems’ should be capable of.

“The new Atlas reflects advances in AI-driven perception, balance, and task execution.”

Subscribe now

3/ Why Taste Matters in Product Management

Nickey Skarstad is a Director of Product Management at Duolingo, a company with a highly differentiated and memorable design and user experience. It’s no surprise that ‘taste’ matters at Duolingo.

“At Duolingo, we have a daily ritual that trains taste across the organization: product reviews with senior leaders (including the CEO!). Not quarterly show-and-tells. Or weekly design critiques. Daily.”

She recently joined Ben Erez and Marc Baselga on the Supra Insider podcast on an episode titled ‘Duolingo’s Secret Sauce.’

She describes product reviews not just as a way to gate features or align teams, but as a way to ‘scale taste’ — an open forum where anyone can watch senior leaders articulate what ‘good’ looks like and calibrate their own judgment faster.

Duolingo pairs that with ‘hot trash’ demos, a bi-weekly space where teams proudly show rough, early work so the organization can learn from experiments long before they’re polished.

“If you spend a few years building product at Duolingo, your product sense is going to be 10x better than if you work somewhere else—and this ritual is a big part of it.”

Subscribe now

Bonus: AI Compliance and Regulation

This fortnightly monthly AI policy roundup by Simmons & Simmons shows how fast AI policy is turning into concrete expectations for product and UX teams, not just lawyers.

From federal moves in the US, to EU transparency codes and New York’s AI avatar disclosure law, ‘governance’ is quietly becoming part of the product spec.

If you’re shipping AI features, this is the kind of document you want bookmarked before you design something that collides with emerging rules on transparency, safety, and content labelling.

“The Code sets out high-level commitments and technical measures for providers and deployers of generative AI systems, aiming to enhance transparency, trust and accountability in the use of synthetic content across the EU.”

I’m starting a new series here on Product State.

It’s nicknamed ‘WGMaTW’ — short for ‘What got my attention this week.’ It’s inspired by Chamath’s ‘What I read this week,’ but with a very specific filter: What’s actually worth a product leader’s attention right now.

This is a bit of a test. We’ll see how y’all respond, and how I feel about it after a few weeks. The best case scenario is it becomes the product-leader version of ‘This Week in Tech’ 😂…

More realistically, it just quietly fades into the internet’s long tail. I’m OK with either outcome.

One thing is for sure: I will be writing this myself. ChatGPT and Perplexity and such are plenty busy right now, and this is meant to be a few minutes of real curation — things that genuinely caught my attention, and might matter to other product people.

All that said, here we go.

Subscribe now

1/ Elena Verna’s growth lessons behind Lovable’s wild $6.6B valuation

Elena Verna is a true thought leader on all things product growth. She’s also an excellent communicator. It’s no surprise she’s leading growth at Lovable, one of the most AI ‘growth-y’ companies out there today.

If you haven’t used Lovable, it’s a great AI app builder — and, in my opinion, clearly differentiated both as a product and as a brand. They’ve done a lot of things right in a very short amount of time.

In this post, Elena recaps key lessons learnt from her time at Lovable, drawing on insights from her recent appearance on Lenny’s podcast.

“Everyone keeps asking for the secret sauce behind the growth. Sorry to disappoint - there’s no magic new channel. We just do a few things differently, and we’re relentless about them.”

Elena's Growth Scoop

Growth lessons behind Lovable’s $6.6B valuation (shared on Lenny’s podcast!)

It’s been a big week for Lovable - we raised at a $6.6B valuation just 13 months after launch. Wowza…

Read more

4 months ago · 46 likes · 7 comments · Elena Verna

Subscribe now

2/ Vibe Coding Engineering

Most conference talks these days are a hard pass. Too many recycled ideas, too much shameless selling, and not enough insight.

That’s why YouTube is so great. You can usually tell within seconds whether a talk is worth your time.

Case in point: This fun gem by Kitze, the founder of Sizzy, at AI Engineer Code Summit. He walks through the past, present, and future of coding with AI — and does so with effortless sense of humor.

Watch it at 1.5x speed if you can keep up. But don’t miss this part:

“Managers have been vibe coding since forever, this is nothing new.”

Subscribe now

3/ Shreyas, Claude, and the two lives of a builder

Shreyas Doshi has worked on product teams at Yahoo!, Google, Twitter, and Stripe.

In this post, he explores — via a chat with Claude — the idea of the ‘two lives’ of a builder. This is something every product leader has experienced; both the gift and the burden of it.

“Product work isn’t just about speed. It’s about 100s of micro-decisions that quietly attach you to what you build, constrain future choices, and determine outcomes before success or failure is visible.”

Shreyas Free Newsletter

The Product Builder's True Journey

The average product builder lives 2 lives…

Read more

5 months ago · 67 likes · 15 comments · Shreyas Doshi

Bonus: Dwarkesh Patel’s conversation with Ilya Sutskever

Ilya Sustskever, of OpenAI fame and infamy, talks about SSI’s (safe and super intelligent) strategy, the problems with pre-training, how to improve the generalization of AI models, and AGI.

“From 2012 to 2020 it was the age of research. Now, from 2020 to 2025, it was the age of scaling… But now the scale is so big… So it’s back to the age of research again, just with big computers.”

Ps.

Christmas is my favorite holiday. If you celebrate, Merry Christmas! ✨ If you don’t, maybe give it a try this year. Either way, this season from Thanksgiving to New Year’s can be bittersweet for many. I hope you know that you matter — far beyond your job and life’s work. This time of year, it’s a great opportunity for all of us to disconnect and show those in our lives that they matter.

Stephanie Leue is the CPO at Ringier Medien Schweiz. She was formerly CPO at Doodle, Head of Product Ops at Eurospace AG, Director of Product & Design at Contentful, Co-Founder at 3Weine, Senior PM at PayPal, and Global Manager of Innovation at Spreadshirt.
Website / LinkedIn

EC: What separates a roadmap that gets buy-in from one that falls flat?

SL: It’s disappointing, but after seeing more roadmaps than I can count across teams and industries, one thing has become painfully obvious:

Most roadmap storytelling falls flat. Not because the plan is missing— But because the narrative is.

After coaching product leaders across continents—and being one myself across industries—I’ve become convinced of one thing:

There isn’t a single right way to tell a roadmap story.

There are at least six.

And each one earns its place at the right moment.

🧭 Vision-first (think: Tesla)

Elon Musk didn’t just pitch electric cars. He laid out a 3-step master plan to transition the world to sustainable energy—starting with a luxury roadster to fund a mass-market future. The roadmap was compelling because it promised more than features—it promised a movement.

🧩 Customer-problem-first (Intercom style)

Intercom’s product team starts every planning session with one question: ‘What problem are we solving?’ Their roadmap isn’t a list of features—it’s a story of real user pain and how they’ll fix it. When used well, this turns strategy reviews into empathy machines.

📊 KPI-led (e.g. Facebook’s ‘7 friends in 10 days’)

In Facebook’s early days, growth wasn’t abstract—it was math. They found one critical metric that predicted retention and built the entire roadmap to drive it. It wasn’t about launching new features, it was about moving numbers that mattered.

📦 Feature/prioritization-first (e.g. early Dropbox)

Dropbox used to crowdsource its roadmap via a public voting system. Popular requests like shared folders or Linux support made it onto the roadmap fast—and users loved the transparency. This approach is clear and efficient—but only works when the audience trusts the prioritization logic.

💼 Sales-aligned (e.g. Salesforce & the ‘Einstein’ pivot)

When large enterprise clients started asking for AI, Salesforce bundled new features into a bold platform play: ‘Einstein.’ It gave the sales team something big to sell—and the roadmap pivot paid off. This works—especially in B2B—when roadmap velocity aligns to revenue urgency.

🏛 Internal pitch-style (e.g. Xbox inside Microsoft)

Xbox wasn’t born from a product plan. It was born from a fear of Sony. Engineers reframed their roadmap as a strategic defence play—to keep Microsoft in the living room. That story won budget, backing, and internal belief. No roadmap spreadsheet would have.

Buy-in doesn’t come from boxes on shiny slides. It comes from urgency.

Make your roadmap the answer to a business problem your clients and stakeholders already feel.

Subscribe now

EC: What’s the hidden work behind product strategy that people don’t talk about?

SL: After 17+ years in product, one lesson keeps proving itself:

Clarity scales better than process.

Every successful team I’ve led had one thing in common:

A simple, shared narrative about the problem, the value, and the direction.

That clarity helps execs delegate confidently and enables product teams to execute without second-guessing.

When the narrative is missing, people fill the gaps with good intentions — just not the same ones.

A strong red thread fixes that fast. And sometimes, even in the absence of a company strategy, as a first step alignment on the key user problem is enough.

The easiest part about Product Strategy is creating a strategy. What makes it hard, is the fact that the strategy is never just ‘the strategy.’ There is also:

The work around the strategy

👉 alignment meetings, executive buy-in, stakeholder negotiations

The work before the strategy

👉research, competitive analysis, customer interviews, market sizing

The work to align the strategy

👉presentations, storytelling, decks, influencing, managing politics

THE STRATEGY

👉frameworks, priorities, bets, OKRs, choices

The work to share the strategy

👉clarifying meaning, localization, translating to teams & goals

The work to implement the strategy

👉projects, roadmaps, resources, coordination, tradeoffs

The work beyond the strategy

👉projects, roadmaps, resources, coordination, tradeoffs

The work after the strategy

👉tracking outcomes, adjusting KPIs, feedback loops, retros

Share

EC: Why do even great product strategies fail without people alignment?

SL: No one tells you this before you become a PM:

The hardest part of product management isn’t building features or great products.

It’s the people.

Even if your strategy is clear, your roadmap tidy, and your team works in almost perfect conditions — one thing remains:

You spend most of your time managing expectations, translating between worlds, calming politics.

Stakeholder management isn’t a ‘soft skill.’

It is product work.

Without alignment, even the best strategy will rot.

I’ve seen PMs feel guilty because they’re ‘not building.’

But sometimes, aligning humans is building — it’s creating the conditions for the product to survive, or even better, to thrive.

Tami Reiss is CEO of Leader Within, Executive Coach at Women in Product, and author of ‘What do Product Managers Do?’ Formerly, she was CPO in Residence at Insight Partners, Guest Lecturer at Northwestern University, and Sr Product Consultant at Pivotal Labs.
Website / LinkedIn

EC: What separates PMs who rise quickly from those who stall?

TR: Product Maangers who focus too much on relationships — or execution — tend to stall.

There needs to be a balance.

Strong product managers make sure their teams are meeting their goals, and are also always investing time in cultivating relationships to generate influence.

Subscribe now

EC: What’s the #1 mistake PMs make when pitching an initiative?

TR: Product Managers tend to think that they need to create a singular great pitch deck to convince everyone about how great the initiative is when, in fact, they will gain much more buy-in by personalizing their pitch to different audiences and individuals.

This allows for the other folks to understand how they will directly benefit from the intuitive. Then there is potential for them to become advocates of the initiative themselves, and generate much more momentum.

Share

EC: What single behavior shift can most accelerate a PM’s career?

TR: The single behavior shift that can most accelerate a PM’s career is to ask open ended questions, instead of making statements.

For example, rather than saying ‘I think this initiative will generate X,’ Product Managers should ask ‘How much X do we think this will generate?’

Or they may ask, ‘What do you think the downstream beneficial impacts might be?’

If it’s something negative, such as a red flag — rather than stating ‘This is a BIG problem and here’s why,’ a PM could ask ‘What happens if this scenario happens?’

And Product Managers should guide other people to get to the same answer.

This creates a situation where other people feel like they generated the idea, and therefore have ownership of it.

Malte Scholtz is the Co-founder of airfocus, and Head of Product at airfocus by Lucid. He’s also the co-founder of welevel. Previously, he was a Sr Product Manager at ePages and a Research Assistant at Otto-von-Guericke University Magdeburg.
Website / LinkedIn

EC: Why do product leaders struggle with misalignment, and how can they fix that?

MS: Misalignment is the silent killer of product success. Nearly 40% of product leaders say it’s their biggest challenge – bigger than technical debt or release velocity.

The real tax isn’t tools or processes, it’s people: Egos, shifting priorities, and competing narratives (check out the full survey).

The mistake many leaders make is thinking alignment comes from ownership or control. Telling PMs to ‘own’ the roadmap sounds empowering, but it actually creates hierarchy and silos. Alignment doesn’t happen in a big meeting – it happens upstream, in the trust you build, the context you share, and the way you tailor the story to every audience.

That means product leaders need to shift from being landlords of roadmaps to orchestrators of context. Over-communicate the ‘why’ behind decisions, invite feedback early, and speak three languages fluently: exec, engineer, and customer. When you combine that with a single source of truth for strategy, roadmaps, and feedback, chaos turns into clarity.

At airfocus, that’s exactly what we provide: a shared product truth that makes the ‘what’ and ‘why’ visible and accessible to everyone.

Subscribe now

EC: What are the most impactful skills for product teams to master in the age of AI?

MS: AI is removing a lot of the busywork – writing reports, updating stakeholders, processing feedback. That’s a gift, because it frees PMs to focus on what only humans can do: asking the hard questions, framing trade-offs, and building trust across the org.

But here’s the catch: AI fluency is now table stakes.

The real differentiator isn’t slapping ‘AI-powered’ onto a roadmap; it’s knowing when to say no, and having the discipline to place fewer, deeper bets. Data literacy, the ability to interpret, challenge, and narrate insights, is now mission-critical. PMs who can’t ‘talk data’ risk being sidelined.

So the critical skills shift from execution to influence: Storytelling with data, stakeholder negotiation, first-principles thinking.

AI helps with the ‘what,’ but the human edge is still in the ‘why’ and the ‘how.’

If you want to read a bit more about this, Marty Cagan and I covered it in this webinar.

Share

EC: What’s your take on the role and value of product ops?

MS: Every other function has an ops backbone – RevOps, Marketing Ops, People Ops – but product leaders are still expected to juggle everything themselves.

That’s why burnout is on the rise.

Product Ops is the scaffolding that lets product leaders focus on leadership, not logistics. It’s not bureaucracy; it’s enablement. The real value isn’t in dictating how teams should work, but in removing friction: making data, tools, and processes accessible, standardized, and value-driven.

Instead of hiring more PMs to play backlog Tetris, smart orgs invest in Product Ops to scale clarity, efficiency, and feedback loops across every layer. Sometimes that means hiring a dedicated role, other times it means elevating a PM with a knack for ops into a hybrid role.

We’re going to be talking more about culture in product teams next month with the VP of Product for Zendesk (sign up here).

Looking ahead, Product Ops will become table stakes for any product org beyond the early stage. It’s the infrastructure for customer-centricity at scale — and a competitive advantage for high-performing teams.

Brian Elliott is CEO at Work Forward, Senior Advisor at BCG, Co-Founder of Future Forum, and the co-author of ‘How the Future Works.’ He was formerly VP & GM at Slack, GM at Google Express, CEO at Monsoon Commerce, and Alibris.
Website / LinkedIn

EC: What are the management behaviors tech leaders should prioritize to make their product teams thrive?

BE: Most product team problems aren't product problems – they're culture problems. I've seen too many leaders try to fix low velocity and missed deadlines with more process or harsher demands, when the real issue is cultural chaos.

Here are the three behaviors that actually move the needle:

Ruthless priority clarity

• Ask each team member: ‘What are our top 3 priorities and how do we measure success?’

• You'll get different answers from everyone – that's your first problem. This is a broad phenomenon that’s easy to prove: ask people how to make toast. Getting everyone on the same page about how we’ll measure success is critical.

• Create a rank-ordered top 5 list of goals and give people permission to say no to everything else. Sounds easy, but can be really hard – we all fear that potential for not maximizing.

• Your PMs need protection from random ‘urgent’ requests. Focus matters, and so does accountability. You are the sh*t shield.

Build trust through dependability

• Start simple: Can your team count on you to do what you say? Build from there: Does everyone know what they’re responsible for delivering?

• Create team-based performance wins. Small wins build trust. Celebrate teams that work well together – point out what they’re doing to accelerate progress and deliver outcomes.

• Don't tolerate toxic high performers – they destroy team dynamics faster than you can hire

Balance performance with growth

• Look at each person's last 6 months: Who's doing the same tasks repeatedly? Growth, not just rest, prevents burnout.

• If I’m working on something that stretches me, I’ll go the extra mile – but not if I’ve done it 10 times before. Dedicate 10-20% of capacity to skill development and experimentation.

• AI experimentation is a team sport. Play with the tools in public – have AI ‘show and tell’ sessions, share what you’ve built in public channels. If you’re a leader, get up to your elbows in it.

The diagnostic: If people can't articulate the same priorities, they're working hard in different directions. Fix alignment first, everything else follows.

Subscribe now

EC: How can product leaders balance delivering near-term results with building long-term capabilities?

BE: This is a false trade-off. The fastest-shipping teams are also the ones investing most in capabilities. The key is treating every sprint as both delivery and learning lab:

Protect the split

• You can vary percentages sprint-to-spring, but average 80% execution, 20% experimentation / capability building

• Yes, this feels inefficient when everything's urgent – but teams that protect this over time consistently outperform those that don't.

• Splits should vary across teams over time; every team hits ‘focus mode’ when delivery is critical; balance it out with ‘rebuild and renew’ periods.

Invest in how teams work

• Capability building isn’t just tech or product experiments – it’s setting norms for how the team works, investing in automation. Activities that make you faster over time.

• Setting norms at the beginning of a project (ex., roles, how we run meetings, what tools we use) slows down the start but accelerates long-term performance every time.

• Invest in decision infrastructure: who decides what, and how we communicate. Use Netflix's ‘informed captain’ model: person closest to problem decides, after gathering dissent

Connect every sprint to future state

• Every feature should answer: What are we learning about users? What are we learning about how we operate?

• What new technical capabilities are we building? What better practices will we use as a team next time?

• How is this expanding our team's skills? What skills do we need to build or source?

The reality: Companies that build learning into their delivery process ship faster long-term because they're building conviction about what to build next.

Share

EC: What's the key to building empathy-driven teams who ship at high velocity?

BE: Empathy drives speed, not the opposite. The teams that ship fastest have deep user understanding and trust in each other. Being empathetic isn’t being ‘nice’ – it’s being demanding while creating environments where people can perform together at their best.

Diagnose team health first. Before you can empathize with users, fix internal empathy:

• Misalignment: Everyone working on different priorities

• Overload: Too many urgent projects simultaneously

• Growth stagnation: People doing same tasks for 6+ months

Create shared user reality

• Engineers, designers, PMs all need direct user exposure. Watch prospective customers interacting with your product for the first time together – group cringe is good.

• When your iOS engineer understands why a feature matters to a specific customer, they'll find ways to ship faster – and often offer alternative approaches to solve the underlying problem.

• Build regular rhythms where the whole team hears from users.

Psychological safety = speed

• Common goals, shared transparently, updated regularly and assessed honestly build trust and reduce the internal rumor mill churn.

• Bad news shared quickly gets solved quickly, especially when teams have common goals.

• Celebrate learning-based failures alongside successes: what did we try that didn’t work, and what did we learn?

• Model vulnerability: share your mistakes publicly. Share your personal user manual.

Quality as Empathy Broken releases aren't just operational problems – they signal you care more about dates than user value. The fastest teams are also the most disciplined about technical practices.

Bottom line: Sustainable velocity requires building quality and trust from the start, not bolting them on later.

Jon Sukarangsan is CEO at Summer Friday and Partners. He was formerly SVP at Monks, Global Executive Director at DEPT, Group Program Director at Critical Mass + HUGE, VP Program at Edelman Digital, and PM at Oracle + Dell.
Website / LinkedIn

EC: How can AI transform design systems, and what role do Product Leaders play in that transformation?

JS: This is one of the most fascinating shifts happening in product development right now, and I think most people are completely underestimating the magnitude of change coming.

Today's design systems are essentially human communication tools—shared libraries of components with style guides and usage documentation. They've largely become these isolated shared services managed by ‘component librarians’ who are somewhat removed from actual product strategy. I’m generalizing — some orgs have very proactive design systems teams very involved in product strategy, but I’ve found it’s more often the case that they are treated like a centralized shared service.

That model is about to change.

What's emerging is something fundamentally different: Design systems as strategic decision repositories. Instead of just housing visual components, they're becoming a place where micro-decisions, design intent, and strategic rationale get stored and made accessible to AI systems.

Think about it this way: Your design system won't just know that you have a card component. It'll understand that ‘this specific card module is critical to our onboarding flow because it reduces cognitive load during the signup process, increases conversion by 23%, and aligns with our strategic goal of reducing time-to-value for new users.’ That's the level of strategic context AI needs to make intelligent decisions.

This evolution means design system teams can no longer operate as isolated component managers. They're becoming the custodians of product strategy at a very atomic level. They need deep understanding of user journeys, business metrics, competitive positioning, and product strategy because they're encoding the ‘why’ behind every design decision.

For product leaders, this represents a massive organizational shift that requires proactive preparation in a few key areas:

First, you need to restructure how design systems teams report and operate. Move them from shared services into strategic product leadership. Give them direct access to user research, business analytics, strategic planning sessions, and competitive intelligence. They can't encode strategic context they don't have access to.

Second, establish new documentation and governance standards. Every component, pattern, and design decision needs to capture not just implementation details, but business rationale, user behavior insights, performance impact data, and strategic importance rankings. This becomes the context layer that AI references when making design decisions.

Third—and this is where most organizations will struggle—you need to build what I call ‘mapping capabilities.’ Your design systems team needs to understand how AI will interpret and act upon all this strategic context they're encoding. They need to think like AI architects, not just component curators.

This means developing new competencies around prompt engineering, context architecture, and AI workflow design. The design systems team becomes responsible for ensuring that when AI generates a new component or suggests a design pattern, it's aligned with your strategic objectives and brand positioning.

Fourth, create cross-functional design strategy councils that bring together design systems teams, product managers, user researchers, and business stakeholders in regular strategic alignment sessions. The goal is ensuring your design system reflects current strategic priorities and captures evolving product intelligence in real-time.

What excites me most about this evolution is that it positions design systems as the strategic memory and decision-making foundation for AI-powered product development. Done right, your design system becomes the bridge between high-level product strategy and tactical implementation decisions that AI can act upon autonomously.

The organizations that get this right will have AI that doesn't just generate components—but it can start to enable bigger ‘magic moments:’ Things like production-ready patterns / users flows, and dynamic personalization.

Subscribe now

EC: What are AI-first design principles that tech teams can start applying today?

JS: You know, I think the biggest trap teams fall into is starting too big and trying to boil the ocean. I've seen countless organizations announce these grand ‘AI transformation initiatives’ only to have them fizzle out because they never gained real traction.

The principle that actually works is what I call ‘friction-first implementation.’ Start by identifying one specific task that creates regular, painful friction for a single role in your organization. And I mean really specific—not ‘help designers work faster,’ but ‘eliminate the three hours our Senior Designer spends every week manually generating design specs from Figma files.’

Here's why this works: When you solve a real pain point completely, you get immediate buy-in and proof of value. More importantly, you learn how AI actually integrates with your specific workflows, tools, and team dynamics. That learning becomes the foundation for scaling up.

But—and this is crucial—you should not just bolt AI onto existing processes. This is where I see the biggest missed opportunities. For example, teams will use ChatGPT to help write PRDs faster, but they're still following the exact same waterfall documentation process they used five years ago. That's optimizing for efficiency, not transformation.

For this example about PRDs — the transformative approach is completely rethinking the role: What if AI generates multiple lightweight requirement variations based on user research data, competitive analysis, and business objectives? Suddenly the PM isn't a document writer—they're a curator, evaluating and choosing between AI-generated options.

This shifts the entire workflow from creation to curation, from detailed specification to rapid validation. The PM can test five different approaches with stakeholders in the time it used to take to write one comprehensive document. When the economics of how long things take changes, you can start to rethink how you allocate your time and resources.

The key principle here is what I call ‘workflow transformation over tool addition.’

Don't ask ‘how can AI help us do our current work faster?’ Ask ‘how can AI enable us to work in fundamentally better ways?’ That mindset shift is what unlocks big improvements rather than marginal efficiency gains.

Share

EC: How do Product Growth teams go beyond experimentation to drive tangible impact in a Sales-assisted SaaS business?

JS: In a PLG model, growth is often about clean funnel experiments — you tweak onboarding or pricing tiers, run an A/B test, and quickly see the impact. But in a sales-assisted SaaS business, the reality is much messier. The most valuable signals aren’t just in product analytics — they’re buried in sales call recordings, CRM notes, support tickets, things like that.

Historically, Growth teams struggled to make sense of that data at scale. You’d rely on anecdotes from Sales — or spend days manually combing through call notes. That’s where AI helps significantly. In the book (‘The Complete Guide to AI-First Product Design’), I talk about ‘context engineering’ — the discipline of taking unstructured inputs like interviews or decision logs and structuring them so AI can reason over them.

Apply that here: AI can synthesize hundreds of hours of recordings, pull patterns from CRM fields, and link those with product usage data. Suddenly the ‘messy human side’ of the Sales motion becomes structured intelligence that the growth team can actually act on.

The second piece is about roles. In the book I also talk about how the Product Manager role evolves into more of a ‘strategic orchestrator.’

Instead of spending their time managing backlogs and writing detailed PRDs, their highest value is connecting signals across systems — analytics, sales conversations, Customer Success feedback — and guiding AI systems to translate those signals into decisions. In a Sales-assisted SaaS context, that might mean ensuring that what we’re learning from the field feeds back into pricing experiments, packaging updates, or roadmap prioritization. It also means giving sales a sharper story: not just ‘here’s what we think the product does well,’ but ‘here’s what the data says actually drives wins.’

So for me, the shift is this: Experimentation still matters, but the bigger lever is turning messy, human data into structured intelligence, and making sure it flows back into both product and sales. That’s how Growth teams in Sales-assisted SaaS move from just running experiments to driving real, compounding impact.

Brian Balfour is the Founder and CEO at Reforge. He’s also a Venture Partner at Long Journey Ventures. Formerly, Brian was VP of Growth at HubSpot, EIR at Trinity Ventures, Founder at Viximo, and Product Manager at ZoomInfo.
Website / LinkedIn

EC: What’s the biggest misconception product leaders still have about growth today?

BB: The biggest misconception is that AI has changed everything. It hasn't. The same fundamental principles are still true. There are just different ways to apply them.

Some examples:

• Retention is still the king of growth. Those with the highest retention will win categories.

• To have retention, you need to build a habit. The journey to build a habit is through a ‘setup’ moment, ‘aha’ moment, ‘habit’ moment. The onboarding experiences we are able to build change, but the fundamentals are the same.

• To grow you need to build a compounding growth system out of growth loops (not funnels). AI has just changed the common bottlenecks of growth loops. For example, in UGC loops the bottleneck used to be on content creation. AI reduces that friction, but shifts the bottleneck elsewhere.

• When you create your monetization model, it still needs to pass three checks:

  • Customer View — Your customers see that Perceived Value > Perceived Price + Friction

  • Growth Loops — Monetization models enable/disable certain growth mechanisms. They work together.

  • Costs — Your monetization model needs to cover cost to serve + cost to acquire.

In AI, Costs have become much more important, which is driving people towards different monetization models.

Point is, if you know the fundamentals well, then you'll have an easier time navigating and winning in all of these change.

Subscribe now

EC: When building a new product in 2025, how should product teams think about defensibility?

BB: There are a few things:

1. Defensibility is built over time, you don't have it at inception.

2. Speed and distribution are not moats. They are preconditions to defensibility. Speed and distribution give you the opportunity to sequence to something more defensible.

3. The forms of defensibility are the same (cross side network effects, direct network effects, data network effects, etc). It's just that certain forms of defensibility like Data NFX are more viable/important with AI vs others.

This means that if you are:

• Early Stage Startup - Don't focus on defensibility, focus on getting escape velocity.

• High-Growth Startup - If you have escape velocity, then you need to start sequencing to more defensible moats.

• Late Stage Startup - If you have one form of defensibility, you need to think about what you add next. You layer them over time.

• Incumbent - You need to assess whether your historical form of defensibility has gotten stronger or weaker. (It's weaker in most cases).

Share

EC: What’s the single most underrated skill that accelerates a product leader’s career?

BB: I would say two things:

1. Being a great picker of companies and problems to work on.

2. Not caring about titles, size of team you manage, internal career ladder definitions, or having a career map.

The most effective people I see are ones that continuously take on hard, ambiguous problems with an ‘I'll figure it out’ attitude, and they don't worry about the rest.

If you do that, I guarantee you'll be rewarded regardless of all the career ladder checkboxes your team has defined internally.

This is hard because — as humans — we seek certainty. We want to be able see what is ahead. The unknown is uncomfortable.

I see this today with AI's changing landscape. Everyone wants to know ‘What will AI do to my job function in X years so that I can perfectly plan my next move."

It's a waste of energy and time. No one can predict the future.

If you take that energy and deploy it to being at the frontier of using this new technology in your profession, you'll be well positioned to navigate the changing waters.

Oji Udezue is the co-author of Building Rocketships. Formerly, he was CPO at Typeform, Head of Product at Twitter, CPO at Calendly, CPO at Parsable, Head of Product at Atlassian, Founder at Kernel Fund, Director at Spiceworks, and PMM at Microsoft.
Website / LinkedIn

EC: What does a ‘Squad’ look like for product-led organizations?

OU: The Shipyard Squad Model is a new-ish way of thinking about product-led organizations.

After years of building products at Microsoft, Atlassian, Calendly, and Twitter, I've seen how traditional product team structures fall short. Here's why we need to evolve:

Traditional product teams were built around the EPD triad:

1. Engineers
2. Product Managers
3. Designers

But these teams were never truly autonomous. They relied on external departments that became bottlenecks for shipping products customers love.

The Shipyard expands the core squad to six essential functions:

1. Engineering (typically 7 engineers)
2. Product Management (1 PM)
3. Design (1 designer)
4. Customer Research (⅓ time)
5. Data Analysis (⅓ time)
6. Product Marketing (⅓ time)

This is a complete product team that can actually ship autonomously. In addition:

- Faster Execution: Squads can deliver at 2x speed when shipping to customers
- Better Decisions: Full context and data at every step
- End-to-End Delivery: Teams own the complete customer experience

But even the Shipyard Squad needs to maintain close connections to customer support, customer success, and sales. These teams provide vital customer insight that helps squads identify and solve sharp problems.

Caveat: The Shipyard squad model will NOT work if team members are fully beholden to their department heads.

Squad objectives take priority over departmental responsibilities.

Unless there's an emergency, Shipyard squads stay focused on their team mission.

Subscribe now

EC: How important is it for product leaders to focus on problem-solving?

OU: We heard this daily at Calendly: 'Why don't you build your own calendar app?'

Our customers wanted it. Our investors suggested it. Even our team pushed for it.

But here's what we understood:

Our customers didn't actually want ANOTHER calendar. They wanted their scheduling headaches to disappear. Meeting coordination was eating up their time and leading to lost opportunities.

By staying focused on that core problem - making scheduling painless regardless of what calendar you use - we delivered exactly what customers needed.

The result? We grew from $25M to nearly $100M in annual revenue during my time as CPO, and later raised $350M at a $4B valuation.

Take customer feature requests with a big grain of salt. Keep your eye on the big picture. Our most successful strategy was focusing on being the best scheduling layer ON TOP of existing calendars, not competing with Google and Microsoft head-on.

Your job is to solve the sharp customer problem, not build every feature they ask for.

Share

EC: After leading product teams while building a few rocketships, how do you define 'product strategy?’

OU: It is inevitably about choice. A software company (or any company, for that matter) has limited resources relative to its ambitions, especially early on, so it needs to prioritize.

A good strategy outlines the best campaign possible given its resources (human, culture, and capital), its customers, and the competitive and market environment.

Strategy is as much about what to do to win, but also about what NOT to do. Every choice in strategy should have a well-articulated opportunity cost in order to really evaluate if the affirmative choices are the best that are possible given the circumstances.

Good strategy flows from four main ingredients:

1. Market environment (20% of strategy)
2. Customers (50% of strategy)
3. Culture and talent (15% of strategy)
4. Company ambition (15% of strategy)

As always, knowing thy customer is the priority. Conduct rigorous (and ongoing) customer discovery and the rest of your strategy will follow.

Oji’s book ‘Building Rocketships’ is available now: