Here's something we see constantly. A founder finally gets acquisition working. The ads convert, a post takes off, and signups climb for a few weeks straight. But growth still feels stuck.

So they reach for more of the same: more channels, more spend, more top-of-funnel. Sometimes that's the right call. But once signups are already trickling in and growth still feels stuck, the bottleneck usually isn't the top of the funnel; it's everything that happens after.

What rarely shows up on a dashboard is that a lot of those new signups are already gone. They showed up, poked around, didn't understand the product, and left. And they're not the type to come back. You spent real money acquiring people who quit before they ever saw what your product actually does.

It’s quite common for new consumer apps to lose the vast majority of new signups within the first few months. For some, that can even happen in the first few days.

The part of the funnel that gets underbuilt

It helps to be precise about what "retention" even is, because people throw the word around like it's a single lever. It isn't. It's three different jobs stacked on top of each other: getting users to their first real win (activation), getting them back often enough to form a habit (engagement), and clawing back the ones who drift away (resurrection).

Most companies pour their effort into those last two - the re-engagement nudges, the win-back campaigns, and the cancellation flows. Activation, the part that has to come first, usually gets a welcome email and maybe a tutorial or two. But neither of the other two can happen without activation. You can't build a habit around a value that a user has never felt. And you can't win back someone who never understood the point of you in the first place.

So when activation breaks, the whole back half of your funnel goes with it. You just don't feel it for three months, by which point you've blamed the product, the pricing, the market, and probably yourself.

This is why activation tends to punch above its weight. Lifting the share of signups who hit that first win by 10% will often do more for the business than a 50% jump in raw signups would. And one of those costs money every single month, while the other you fix once and mostly stop paying for.

The moment your product clicks

So what's this "first win"? Every product has a moment where a user flips from skeptical to sold - from "I hope this works" to "oh, that's actually useful." Growth folks usually call it the aha moment; we like "lightbulb moment." Getting crisp on yours is worth more than any onboarding tactic you could copy from someone else. Once you know it, the rest of onboarding is mostly logistics for getting people there faster.

And the rigorous way to find it is in the data: compare the users who stuck around with the ones who bailed, and look for the earliest action that separates them. That's your activation metric, the number you optimize against.

Here's what activation is not. It's not the signup. It's not finishing your setup wizard, and it's certainly not sitting through a boring product tour. Founders fixate on those steps because they're easy to build and easy to measure - completion rates go up and to the right, everyone feels productive. The user, meanwhile, feels nothing. Nobody has ever texted a friend about a great onboarding checklist.

That's actually the test. Your lightbulb moment is the first thing a user would mention to someone else, unprompted - the "you've got to try this" moment. If you can't name it in your own product, I promise your users can't find it either.

Build backwards, not forwards

Once you can name the moment, the move looks obvious: get people there fast. And this is where a lot of teams trip, because the natural way to build an onboarding flow works against you. They build forward.

They start at the signup screen and bolt on steps in the order they think of them. Verify your email. Fill out your profile. Set your preferences. Take the tour. Invite a teammate. Each one seems reasonable on its own. But that's five things a new user has to slog through before your product gives them one good reason to come back.

Flip it around. Start at the lightbulb moment and walk backward. For that moment to happen, what has to be true? And for that to be true, what has to happen right before it? Keep going until you bump into the signup screen. The trail you just traced is your critical path - the steps that have to exist. Everything you didn't trace is optional. And "optional," during onboarding, almost always means cut it, or reveal it later on.

Duolingo does this extremely well. They ask a simple question up front: which language do you want to learn? And then you're translating a real sentence, all before it even asks you to make an account.

Canva runs the same playbook. You’re not dropped into a blank canvas with a basic product tour. You’re immediately asked, "What will you be using Canva for?" and then presented with a template that already looks good, which you change in a couple of clicks. Canva took something that could have been a long, fiddly setup and got people to "wait, I made that?" in barely any time at all.

The moves are always a version of the same idea: swap setup for samples, swap a wall of choices for one good default, swap "assess yourself first" for "just start." Run every step in your flow through a single question - could they reach the lightbulb moment without this? You'll feel a little silly by how much you're making them do for no reason.

Get the friction off the path

Even if you lead people directly to the lightbulb moment, people will still fall off the path. The thing pulling people off it is friction. All the small moments of effort, confusion, or hesitation between signing up and getting value. It rarely shows up as one obvious problem. It shows up as a dozen tiny ones.

It typically shows up in three flavors.

• Cognitive friction is making people think too hard. The field labeled "configure your taxonomy hierarchy" when what you meant was "how do you want to sort your files."
• Emotional friction is making people nervous. "Choose your plan (this can't be changed later)" takes a casual click and makes it feel permanent and risky.
• Interaction friction is just making people grind. Six form fields and an email confirmation before they've done one useful thing, when a single "continue with Google" would've carried them straight in.

None of these kills activation on its own. It’s easy to shrug off as too small to bother with, so it never gets fixed. Do that five times over and you've lost a real share of the users you paid to bring in.

The most common version, and the one I'd fix before anything else, is the empty state. That's what your product looks like to someone who just walked in the door, before any of their data is in it. The whole dashboard is there - the menus, the panels, the charts - they're just all blank. Every number reads zero, every list says "nothing here yet," and somewhere in the middle sits a "Create your first project" button, waiting for them to go fill it all in themselves.

It doesn't look like a problem. Nothing's broken, the thing loaded, the button works. That's why it’s easy to leave alone. But clicking the button was never the hard part. The hard part is everything a new user has to figure out before they're willing to click it.

Picture it from their side. Someone just signed up, which means this is the most motivated they are ever going to be. They have a problem, they went looking, they picked you. Then the product loads, and it's a blank dashboard with a "Create your first project" button and nothing else.

Watch what happens in their head. First, they might think: what do I even make? A project for what? Is this how I'm supposed to use this? They came to you for an answer, and they were handed a blank worksheet at the exact moment they understand your product the least.

Then the feeling, which is the part that actually loses them. A blank screen makes a person feel behind, like everyone else got it and they're the slow one. The bit of hope they showed up with changes to "maybe this isn't for me." Nobody rage-quits an empty dashboard. They just tell themselves they'll set it up later. Then they forget, and later never comes.

That's why this beats every other friction fix on your list. This one costs you people at the peak of their motivation - the single best chance you had to win them. We don’t want to take your best moment with a new user and then make them feel lost.

So don't start them from nothing. Drop them into a realistic example instead - a sample project with real-sounding tasks, a dashboard with believable numbers already in it - so the first thing they see is the product doing its job. Let them poke at it, change it, and delete it once they get the idea. Instead of asking "what are you going to build?", you're showing them what it looks like already built. That's a much easier place to start than a blank page.

If you’re a Notion user, do you remember when you first opened Notion, and you were like, “Uhh, what do I do?” but then you loaded up one of their many templates, and things started to click? That’s what we’re going for too.

The cheapest growth you have

The expensive part is already done. You paid to get these people in the door - the ads, the content, the launch, whatever it took - and they showed up. Then most of them wandered off before the product ever did anything for them. Working on activation is how you collect on money you've already spent.

So before you go buy more traffic, go watch what happens to the traffic you've got. Sign up for your own product with a fresh email and a stopwatch running. Find the moment it clicks - if it clicks at all. Count the steps between the front door and that moment, then start deleting them.

For a lot of companies it's the cheapest growth on the table, and it's been sitting there the whole time, inside a funnel you already built.

If you know a founder busy pouring money into the top of the funnel while the bottom leaks out the back, do them a favor and forward this along.

Thanks y'all!