5. Does the first version do what the evidence said

§The question: does the product, live, actually deliver the value the prototypes and conversations promised?

§Before launch, strangers said the value was real and behaved as if it were, on a prototype that did nothing behind the screen. Now the screen does something, and the first loop checks whether the promise survives contact with a real product and a real user who has no one sitting beside them explaining it.

§ 5.1Activation is the first proof#

§Activation is the share of new users who reach the first important step: the moment the product's value becomes real to them for the first time. Not sign-up, which is a promise. Not the tutorial, which is throat-clearing. The first time they do the thing the product is for and get something back.

§Defining that moment is the work. Sean Ellis calls it the aha moment, and it is specific to the product: the first repository connected, the first image generated and saved, the first question answered with a source the user clicks. For the hospitality group, act one's activation was a completed profile with the tier visible, because that was the moment recognition became real to the guest.

§ 5.2The activation insight#

§The famous examples are famous because a team found the specific early behavior that predicted long-term retention, and then made everything about getting new users to it. Facebook's early growth team is widely reported to have found that users who reached a certain number of friends in their first days were the ones who stayed, and pointed the whole onboarding at that number. Whether or not the exact figure is remembered correctly, the move is the lesson: find the early action that separates the users who stay from those who leave, and define activation as that action.

§You find it by comparing cohorts: what did the users who were still here in month three do in their first week that the users who left did not? The answer becomes the activation metric, and the first validation loop is whether new users reach it.

§ 5.3Cohorts, not snapshots#

§A snapshot says "62% of our users have completed a profile." It is nearly meaningless, because it blends everyone who ever signed up. A cohort says "of the users who signed up in the first week of March, 62% completed a profile within a day, and of the April cohort, 71% did." That comparison tells you the product is getting better at activating people, which is the only thing the number is for.

§Amplitude's playbook makes activation an input metric to the North Star: not the headline number, but one of the few things that reliably drive it. At this stage, before there is a North Star worth naming, activation by cohort is the headline.

§ 5.4When activation is low#

§Low activation has two causes and they need different responses. Either people cannot reach the value, a usability problem, or they can reach it and it is not valuable, a value problem. The two look identical in the number and opposite in the fix.

§You tell them apart by watching. If users get stuck at a specific step, it is usability, and chapter 12's onboarding experiments apply. If users glide through and simply do not come back, the value was not there, and no amount of onboarding will fix it; the honest move is back to the fourth whitepaper, because the direction was wrong. The field sprint's discipline returns here: watch a few real users reach or fail to reach activation, because the number tells you that it is low and only observation tells you why.

§ 5.5What you leave with#

§A defined activation moment, chosen because the users who cross it behave differently. Activation measured by cohort, week over week. And, when it is low, a qualitative read of real users to say whether it is a usability problem to experiment on or a value problem to take back to strategy. The next chapter is the assumption underneath activation: whether the activated users come back.