The precise instant a new user finally understands why your product is useful to them.
In practice
The most-cited example comes from Facebook: growth teams found that a new user who reached 7 friends in 10 days almost always stuck around afterward — that specific threshold became their onboarding compass for years. Finding your own aha moment means looking backward, not forward: what early action do all the users who stayed have in common? It's rarely the most prominently featured part of the interface — often a secondary action nobody pushes hard enough.
The formula
Lift of a candidate action = share of retained users who did it in week 1 ÷ share of churned users who did it in week 1
- Candidate action
Any early, observable event: created a project, invited someone, imported data, hit a threshold (7 friends, 3 reports, 1 payment). Test several; the winner is rarely the feature the team is proudest of.
- Retained vs. churned
Retained: still active after three of your natural cycles (week three for a weekly tool). Churned: gone before the end of the first one. The middle is noise; leave it out of the comparison.
- Lift
A lift of 1 means the action doesn't distinguish the two groups. A lift of 5 or more, on an action a majority of retained users did, is a strong candidate. Pick the earliest strong one you can plausibly get most new users to do.
This finds correlation, not cause. Users who invite a teammate in week one don't stay because they invited someone; they invited someone because they already intended to use the tool with their team. The action is a visible proxy for that intent — which is exactly what you need to measure and to design onboarding towards.
Worked example
The famous one, and a small one
- Facebook, around 2008: users who reached 7 friends within 10 days retained far better than those who didn't. The number was found in the data, not chosen; the whole onboarding was then bent towards "find people you know".
- A two-person invoicing app, today: 800 sign-ups over a quarter. Retained at month three: 140. Churned in month one: 480.
- Candidate "created an invoice in week 1": 78% of retained, 35% of churned — lift 2.2. Candidate "sent an invoice to a real client in week 1": 64% of retained, 6% of churned — lift 10.7.
- Aha moment: first invoice sent, not created. The onboarding stops ending on a beautifully formatted draft and starts ending on "send it". The activation rate is now something worth tracking.
Creating something is the demo; using it for real is the moment. Most products' true aha is one step later than where their onboarding stops.
Orders of magnitude
- The well-known examples are thresholds, not features: Facebook's 7 friends in 10 days, Slack's 2,000 messages sent by a team, Dropbox's one file in one folder on one device. Each is early, countable, and a proxy for "this thing is now part of how I work".
- How many users reach it is the activation rate; how long it takes them is time-to-value. Both belong on the same chart, because a moment reached by 60% of users in three weeks and one reached by 35% in ten minutes describe very different onboardings.
- A moment that fewer than a fifth of retained users ever reached is not their aha moment, however much the team wishes it were. The data has to nominate it.
How to improve it
- Find it before you optimise anything: the comparison above takes an afternoon with an analytics export. Guessing it costs months of onboarding work pointed at the wrong screen.
- Write it as one sentence with a number and a deadline — "sent one invoice to a real client within 7 days" — and put that sentence where the whole team sees it.
- Redesign onboarding backwards from it: every screen either moves the user towards the moment or gets cut. Empty states, sample data and templates exist to shorten the path to it.
- Revisit it once a year or when the product changes shape. A moment that fit a single-player tool stops fitting once teams become the customer.
In the Tour
Have you defined a specific "aha" moment for new users?
- Yes, and we measure it20 pts
- We have one, but don't measure it7 pts
- Not really0 pts
"Have you defined a specific aha moment for new users?" opens the Activation stage: 20 if it is defined and measured, 7 if it exists but isn't measured, 0 if not really. The two questions after it — what share of users reach it, has onboarding been iterated on data — cannot be answered well if this one scores 0, which is why an undefined moment usually drags the whole stage into the weak band.
Questions people ask
Is the aha moment the same as activation?
The aha moment is the event; activation is the stage, and the activation rate is the share of new users who reach that event within a window. Defining the moment is what makes activation measurable at all — without it, "activated" means whatever the last person to build a dashboard decided.
What if I don't have enough users to find it in the data?
Then find it in conversations. Ask ten users who stayed what made them decide to keep using the product, and ten who left what they were hoping for; the stayers' answers converge on an event much faster than you'd expect. Write it down as a hypothesis, instrument it, and confirm it with data once the cohorts exist.
Can the aha moment be a feeling rather than an action?
It can be experienced as one — "oh, that's what this is for" — but it has to be defined as an action to be of any use, because you can only measure and design towards things you can observe. Pick the action that most reliably accompanies the feeling. Half of the work is precisely that translation.