Most of the churn you complain about already happened before the user ever saw your product’s value. The cancel button is just the delayed paperwork. By the time someone clicks it, the bet was already lost — usually in the first session, often in the first sixty seconds.
I’ve now run activation work across three very different products — a B2C savings app (Zepargn), a B2B logistics SaaS (ChapChap), and a fintech rail (Kkiapay) — and the pattern is the same every time. The teams that pull activation up don’t do clever things. They relentlessly attack three levers, and ignore almost everything else.
This post is the version of those notes I wish someone had handed me three years ago.
Why most activation playbooks fail
Open any “activation” deck and you’ll find the same Russian doll: define your aha moment, instrument the funnel, A/B test the onboarding. It’s not wrong. It’s just the surface.
The deeper issue is that most teams confuse activation with onboarding. Onboarding is the first ten minutes of UX. Activation is the moment a user’s expected value crosses some invisible threshold and they decide — not always consciously — that coming back is worth the effort.
The thing about that threshold:
- It is product-specific. There is no “magic number”.
- It is set by what the user thought they were getting before they signed up, not by what you think you delivered.
- It is decided fast. Most users have already made up their mind by session three.
Once you accept that, the question becomes practical: what do you actually do, every week, to push more users across that threshold? After a few hundred experiments, I’ve only found three answers that consistently work.
Lever 1 — Time-to-first-win (TTFW)
Every product has a smallest possible thing that, when it happens, makes the user say “ok, this works for me.” That is the first win. Not the full vision. Not the killer feature. The first small, real, observable proof.
At Zepargn, the first win was not “reach your savings goal.” It was seeing your first automatic transfer happen on schedule. That’s it. A single ₣500 transfer landing in the savings wallet on day four was the moment retention diverged.
At ChapChap, for a B2B logistics dashboard, the first win was getting the first invoice paid through our financing rail — not the dashboard, not the integrations, not the 12 reports we eventually built. Just one paid invoice.
Most teams optimise the wrong thing because they confuse the first win with the headline feature. Headline features are slow, complex, and gated by trust. First wins are small, fast, and gated only by clarity.
How to find your first win
Ask the question backwards:
What is the smallest event in our database that, when it fires, predicts a user will still be active in 30 days?
Run that query honestly. Not by intuition — by cohort analysis. You will be surprised how rarely the answer is the feature you put on the homepage.
How to shorten TTFW
Three rules I keep on a sticky note:
- Pre-fill everything you legally can. If you ask the user to type something they could have inferred or imported, you are subtracting from your own activation rate.
- Move money / data / objects of value as early as possible. The user trusts you the moment something they care about lands inside your product.
- Cut steps until it hurts, then cut one more. Most onboarding flows have two or three steps that exist only because someone in the company asked for them.
At Zepargn, we shaved the savings flow from 7 screens to 3. Activation went from 18% to 31% in six weeks. We didn’t add a single new feature. We removed.
Lever 2 — Cognitive load reduction
New users have a budget. Not of time — of attention. Every decision you ask them to make in the first session subtracts from that budget. When the budget hits zero, they don’t rage-quit. They quietly bookmark and never come back.
Cognitive load is invisible in your funnel charts. You can’t see it on a Mixpanel dashboard. You only see its consequence: a quiet, persistent drop in session two retention.
Three sources of cognitive load to attack
Decision load. Every choice — plan, currency, notification preference, theme — is a tax. The default should always be the right answer for the median user, with a clear escape for the rest. If your settings page has more than ten toggles, you are taxing new users for the convenience of power users.
Vocabulary load. If your product invents new words (pods, threads, blocks, channels, ledgers, projects, workspaces), each new term is a small tax. Sometimes worth it, often not. At Kkiapay, we tried “wallet” and “account” for the same thing depending on context. Users never recovered. We picked one. Activation moved.
Spatial load. The number of distinct screens, panels, and tabs a user has to navigate to do their first useful thing. Every click is a tiny test of motivation. Most products fail far more of these tests than they think.
The empty-state trap
New users almost always land in an empty product. Empty dashboards. Empty inboxes. Empty pipelines. Empty is hostile. Empty says this is your problem now.
Two things help. First, seed the product with a meaningful sample — pre-populated with the user’s actual data wherever possible (their name, their company, their first invoice). Second, replace empty states with one extremely clear next action. Not three suggestions. Not a tour. One button. One outcome.
At Kkiapay, replacing the empty merchant dashboard with a one-line state — “Send your first payment link” with a giant button — cut session-one drop-off by a third.
Lever 3 — Behavioural commitment
The first two levers reduce friction. The third works in the opposite direction: it asks the user to invest something small but real, early in the relationship, on purpose.
Why? Because invested users behave differently from observers. Once someone has named their workspace, imported their contacts, scheduled their first transfer, or written their first note — even if the objective sunk cost is zero — their psychology shifts. They are no longer evaluating; they are owning.
This is the lever everyone gets nervous about, because the obvious version of it (“ask for credit card up front”) is also the version that destroys trust. The right version is subtler.
Good investment requests
- Naming. Asking the user to name something (workspace, project, savings goal, business). Cheap to do, hard to take back, very high in psychological commitment.
- Connecting. Linking a calendar, a bank, an email inbox, a phone contact list. The user just made you part of their existing infrastructure.
- Importing. Bringing in their existing data — even a tiny amount. Once a user’s data lives in your product, it is much harder to leave.
- Inviting. Bringing one teammate, one client, one family member. Social commitment is the strongest variant of all.
Bad investment requests
Anything that asks for commitment before the user has had a first win. Asking for credit card on signup, asking for KYC up front, asking for a 14-step setup before the product does anything.
The order is non-negotiable: win first, invest second. Reverse it and your activation rate halves.
Putting it together — the activation matrix
For each new user session, three questions:
- How fast is the first win? (Lever 1: TTFW)
- How heavy is the load to get there? (Lever 2: Cognitive load)
- What did the user invest, and when? (Lever 3: Commitment)
That’s it. Every other activation tactic — emails, tooltips, progress bars, gamified checklists, AI nudges — is a delivery mechanism for one of those three levers. None of them are levers in themselves.
What I stopped doing
Some things that sound like activation work but mostly aren’t, based on enough painful experiments:
- Product tours. They make stakeholders comfortable. They do not move retention. Users skip them, and the ones who don’t skip forget.
- Welcome emails. If the product itself doesn’t activate the user, no email will. They’re fine as a backstop; they’re not the work.
- Gamification. Streaks, badges, levels. Marginal for products that already work. Useless for products that don’t. Often a smell that the team has run out of real activation ideas.
- Onboarding checklists. Useful only when each item ties to a real first win. Most don’t. They become busy-work theatre.
Instrumentation that actually matters
To work the three levers, you need three numbers in your weekly review:
- Median TTFW. Across all signups this week, what was the median time from signup to the first-win event? Track in minutes, not days.
- First-session friction. Of users who don’t reach the first win, how many screens did they touch before quitting? More screens = more cognitive load. Falling number here is a good sign.
- Investment depth. Of activated users, what percentage performed at least one commitment action (invite, import, connect, name) within the first three sessions?
These three numbers, looked at together, week over week, will tell you more about the health of your activation than any churn dashboard.
The honest closing
Activation work is unglamorous. It rarely makes the changelog. It never goes viral. The wins look like “we cut three screens and retention went up four points”, which is the kind of sentence nobody retweets.
But it is, in my experience, the highest-leverage thing a Senior PM can do for a product’s long-term health. Acquisition gets you users. Activation gets you a business.
Start with TTFW. Reduce load. Earn an investment. Repeat next week.