Constraint is a high leverage growth move.
Constraint almost always makes us better.

We all love to jump to the question, “What do we need to do to get people to use our product?”. We pull up whiteboards and Google Docs and get on Slack huddles writing down the things our product is missing. The gaps to close, the features that help us win. Claude helps format them into roadmaps with pillars of work we bring to our team, and we get building. Adding things to fix our problems.

We might speak to the marketing and sales people too, getting ideas of new ways to motivate people. We come up with informational content, hooks to test, etc.

Our Q4 is about to start, and I’d bet almost all of us have plans set in motion that—summed up—are for More.

But the close-cousin of that question that we ask less is, “Why are people not using what we have the way we expect them too?”.

It can feel like the same question, but it’s not. The first question is a leading question that frames the answer as we need to do this and that thing. And doing a new this and that feels intrinsically additive.

The second question is more probing, and today’s idea is that I want you to probe the answer from the perspective of constraint, limitation, and subtraction.

For Less. Because the honest answer to why people are not using your product is often not that "it wasn't good enough." It's that you buried the good thing under a choice. You made people opt in, discover it, remember it exists, build new habits, learn new things, or compare it to other things you give them.

Again, constraint is the highest leverage growth move—for your product and business, and also for you personally.

While the evidence for this growth lever is overwhelming, our roadmaps and plans seldom reflect it. Adding more is sexy and fun to report on and lives nicely on product marketing pages and resumes. Adding constraint is mostly the opposite.

But Anthropic, arguably one of the most performative product teams operating right now, have given us a great demonstration of picking the right constraint in their product. In turn, hand-plating us a fundamentally true lesson on how humans and users behave:

Simply, the smartest thing you can do for yourself and your users is to find the right constraint—a boundary that makes the good decision the easy one.

Ok, so what did Anthropic do? 🤔

Attio picks up leads for you at 2am. Catches renewals before they slip. Hands you the answer before you ask.

If you want to build pipeline, convert leads, run a sales motion, forecast revenue, or retain and expand your customers…Attio is the agentic CRM I use for my business. So do TaskRabbit, Granola, Modal, and Wispr.

🥷 Stop adding, start subtracting, consolidating, and finding creative constraints.

Ok, Anthropic. As you know, they’ve been on a growth and addition rampage tearing through markets and verticals.

Claude → Claude Code → Claude Cowork → Claude Design → Docs, Presentations

What they do is ever-growing. But that truth of “humans have limited working memory” is as unfaltering as E = mc².

Our attention when pushed to its finite limits breaks and something has to give. You already know this. It's why we turn the radio down when we’re hunting for an address or looking for parking. The music isn't blocking our eyes, it's competing for the exact capacity we need to read street signs.

I don’t know if you ever found that when using Claude, there was some uncertainty about where you should be working—do I do this in regular Claude? Is this for Cowork, or Code? Which product should I use for this task? I have chats and work sprawling, unconnected, across desktop chat apps and cloud chats and every one of their tabbed products.

I doubt that was just me because the data told Anthropic this was becoming a problem. So, they constrained their surfaces down to just two tabs. They turned the music volume down so we could get capacity back and focus.

One surface for everything, where the smartest choice was removing the choice users had to make about what tool to use for the job. Claude decides now.

Cowork is in Chat. And so is Design, and their latest Document and Presentation features.

The capability of Claude didn't get smaller, just the number of things you had to think about before using it did.

And I’d bet every nickel I have that adoption of all those products will go through the roof. Meaning users will get more value for less thinking, and Anthropic will make more money when tokens are consumed for those heavier-compute products.

There’s another great lesson here. Because in reading what they just did it sounds so obvious…why did they not just do that at the start?

Even a brilliant product-first company follows the inertia of how delivery is done in org structures, and the default idea of segmenting new value into its own thing.

I imagine every one of those expansionary products by Anthropic is its own team. Meaning on the roadmap, each of these things was likely shown as a new bet being built and led by a unique team. Different teams like to have their products under different tabs or pages in a product—differentiating what they do from what another team does. This is just easier to organize around and maintain ownership of, as multiple products on one surface means more dependencies and teams working together.

But your customers obviously don’t care what is easier for you. What’s easier for them is almost always constraint.

And constraint also makes it easier to sell what you do. Hammers hammer things, they don’t hammer and paint walls and clean sinks. So when you have a nail that needs hammering the choice is always obvious what to buy.

I found this incredibly true with my own app.

My personal app project right now is Little Moments , a daily photo-and-video journaling app. First and foremost it’s for me, because my days and weeks tend to blur together, journaling has never really stuck, and I take a lot of photos of life that just sit in my camera roll.

In under 60s a day, captioning your camera roll to capture a moment is fun and easy, and I love watching the movies and chapters of my life get stitched together.

When usage wasn't there, my instinct was the obvious one: make it more enticing. Add more. At first, I had multiple prompt types giving several ways to capture a moment, a rewind feature to race through the camera roll, story coaches walking you through how to tell better stories. The logic was more doors in means more people find their way in, and it’s never been easier and more fun to make more doors.

When shipping a new prompt type took actual time, restraint was forced on you by capacity. Now the only thing standing between you and a bloated product is you actually saying no and forcing the constraint.

So, I cut what I liked making. I cut features some people were still using. I folded features that overlapped into one, and I doubled down on the things that showed in the data to be doing the most work. I stripped Little Moments down to the photo-and-video moment, added a forced time constraint, built the 10s “TikTok-able” aha moment, cleaned up onboarding until the path was constrained and there were fewer doors to open.

CVR went up. Daily usage went up. Explaining what it was became easier.
The metrics moved because I removed things and merged things and invented a time constraint, not because I found a cleverer way to explain the things already there.

Again, people seldom need more pushing to do the thing you want them to do.

Constraint does not push, constraint pulls by elimination of options.

To see the latest iteration of my constraint, and only if the problem of life blurring together resonates and the idea of noticing better is interesting to you, try Little Moments out. Capture something, and let me know how it feels.

And beyond my app, I’ve made the mistake many times and still haven’t fully learnt the lesson, that I do much better on my personal and business goals when I force constraint on myself and say no to doing expansionary things I really want to do.

It sucks but time is finite and no matter how much we try to do all the things we want, we never will be able to. Yet I take on new projects, say yes more than I should to things, and add goals too easily.

Just this year I’ve launched this newsletter, built 2 real apps (Good Times and Little Moments), launched mini games, started YouTube and TikTok, started about 50 GitHub repos of projects (some have seen the light of deployment with their own domains and many died on localhost), and have written down many more things “I really ought to try and add to my plate”. That’s just my business projects, forget work and personal!

It’s almost a joke. And really, just one of those things actually works. The newsletter. The rest I justify as fun and playing around, but the truth is it’s just poor focus and poor exercise of constraint, and without a doubt that is costing me and the success of my newsletter. My finite attention and mental bandwidth is in constant tension between these things as I try keep all of them alive.

When we focus on 1 or 2 goals, that constraint allows for a type of depth that makes you so much better.

Fewer, more focused, surfaces will always catch more stray luck.

Clever constraints can do heavy lifting for your product being spoken about and going viral, not just because it’s easy to explain what you do, but because when most products are looking to add more and expand flexibility, an opinionated constraint stands out as novel and share-worthy. Consider some of these:

These limitations are invented and baked into the product, but create great hooks and simple aha moments that drive adoption and usage.

The product examples above are all made up and I’d argue found by forcing thinking inside the box. We’re told to invent by thinking outside the box because outside the box is infinite. But finite means you need to be more resourceful and creative. Back to those examples:

They obviously all could have leant into the More mindset, but people are simple and the answer to what they want and will talk about is often inside an established box. Less.

While working memory is the bottleneck of human cognition and nothing will change that, design failures that show a failed understanding of this are everywhere.

We build organizations, policies, products, and our own lives for this imaginary person—someone with infinite attention and perfect self-control. Then we’re surprised when people don’t behave the way the perfect one does.

A group of researchers in Brazil and the US coined a useful pair of terms for this

The ideal citizen does not exist. More tooltips, modals, powerful products in new (but competing) tabs, new educational content, hacks to motivate or increase discipline…these things work great in out head for imaginary people.

The real citizen is who you’re designing for. And the most useful thing you can give a real citizen is rarely more information, choice, or motivation. It’s the right constraint—a boundary that makes the good decision the easy one.

So, how do you steal this idea? 👇

Pull the list of things you shipped this year, ranked by adoption. Take the bottom one. Before you write new onboarding copy or send one more "have you tried…" email, ask: is this opt-in? Is there a version where it's just on for a segment of users, with an easy way out, and you watch what happens for two weeks? Cutting a setting, collapsing two surfaces into one, or flipping a feature to default-on are all the same move—removing choice.

And if you haven’t shipped anything and are working on a brand new idea, lean into the box. What’s the opposite to what others in the space are doing? What’s an almost silly way to force limit the product? If you can build only one feature or flow, how do you constrain it? Start there.

Same question, just aimed inward. What's the thing you're currently trying to resist or achieve through sheer discipline that you could just remove (or add) to your environment instead? If you want to read more, have a book on your night stand. If you want to be on Instagram less, delete the app and only visit on your computer.

If you want to succeed with your one business idea and worry less about everything else on your seemingly-necessary plate and finally free yourself, say no and cut back all the other ideas and distractions you have that you’re pretending are just as important. Constrain your goals.

Fine, I wrote that last one for myself.
Now I’m going to do it.

Will you?

› The Paradox of Choice: Barry Schwartz's TED talk on why more options make people less satisfied and never more free. → Watch it here (19 min)

› Can Defaults Save Lives?: a plain-language breakdown of the Johnson & Goldstein organ donation study—opt-out countries get roughly 60 percentage points higher donor enrollment than opt-in ones. → Read it here (8 min)

› Shape Up: Basecamp's free playbook for building with a fixed appetite instead of an open backlog. → Read it here (20 min for the Betting chapter)