Leadership has asked for 20 per cent growth in activations next quarter. How do you decide whether that is the right target?

A target without a mechanism is a wish

The question is not whether 20 per cent is ambitious, it is whether anybody can name the things that would produce it. A number handed down without a route attached gets met in one of three ways: the underlying business genuinely improves, the definition of the metric quietly loosens, or the quarter ends with an explanation. Only the first is what anyone wanted, and the second is far more common than teams admit, because loosening a definition requires no permission and produces a green number.

So the first thing you do is not negotiate. You decompose. Growth in activations is not a thing you can push on; it is the output of a small number of drivers you can push on, and until those are on the table there is nothing to discuss.

Build the number from the drivers

Start with the baseline and be explicit about what the baseline already contains. Take a product that activated 10,000 accounts last quarter. Of that, some portion is the existing trajectory with no new work at all — say the last three quarters ran 8,400, 9,100 and 10,000, which is growth of 8 per cent and then 10 per cent, so taking the cautious end of that an unchanged quarter would land near 10,800 on trend. That matters enormously, because a 20 per cent target is 12,000, and 800 of the 2,000 increment is already arriving without you. The genuine ask is 1,200 incremental activations from new work, which is 11 per cent above trend rather than 20 per cent above last quarter. Establishing that distinction is the single highest-value thing you do in the first hour, and a startling number of teams never do it.

Now attribute the 1,200. Suppose the funnel is 100,000 sign-up starts in a quarter, 40,000 completing sign-up, and 10,000 activating, so completion is 40 per cent and activation-from-completion is 25 per cent. Three candidate drivers, each with a stated assumption:

Driver	Assumption stated	Incremental activations
Reduce sign-up form from 9 fields to 5	Completion 40% to 45%, activation rate unchanged	5,000 more completions at 25% = 1,250
Onboarding checklist for completed accounts	Activation-from-completion 25% to 27%	40,000 at 2 points = 800
Paid acquisition increase	Sign-up starts 100,000 to 115,000, rates unchanged	15,000 at 40% at 25% = 1,500

The arithmetic is checkable because every input is on the page. What it shows is that the target is reachable on paper, and that two of the three drivers are not yours: the paid acquisition line belongs to marketing's budget, and the form change belongs to whoever owns the sign-up surface. This is the point at which a target becomes a negotiation about ownership rather than about ambition.

Say what the residual is

The move that separates a senior answer is stating the gap you cannot explain. If your drivers credibly total 900 and the ask is 1,200, do not close the difference with confidence. Say that 300 activations of the increment have no mechanism behind them, name what would have to be true to find them, and put a date on when you will know. The reason this matters is not honesty for its own sake; it is that an unexplained residual is where the quarter goes wrong, and naming it in week one converts a future failure into a present decision somebody else can help with.

You should also discount the effects you have listed. Each figure in that table is a hoped-for effect, not a measured one, and the historical hit rate on such estimates in most teams is well under half. So carry the number twice: what the drivers are worth if they all land as estimated, and what they are worth at a stated haircut. Saying "we plan for 900 and commit to 600" is not sandbagging when the reasoning is visible; it is the difference between a forecast and a commitment, and being clear about which one you are giving is a question you should answer before it is asked.

Name how the number could be met badly

Every metric has a cheap route. Activations can be inflated by moving the activation event earlier in the journey, by counting trial accounts that were never going to convert, by re-activating dormant accounts through a discount that destroys the margin, or by acquiring users who activate and churn within a month. All four produce 20 per cent, and none of them produces a better business.

Volunteering this before the interviewer probes it does two useful things. It shows you understand that a target changes behaviour rather than merely measuring it, and it lets you propose the guardrails alongside the target rather than as an afterthought: the activation definition frozen in writing for the quarter, retention of the activated cohort at day 30 reported next to the headline, and cost per activation reported so a paid-acquisition route cannot masquerade as a product improvement. A target quoted without its guardrails is an invitation, and the person who accepts it is accountable for what happens next.

Where good product managers still get this wrong

The subtle error is not accepting an unreasonable number. It is accepting a reasonable number attached to drivers you do not control, and then being held to it alone. Two of the three drivers above sit outside the team, which means the target is a joint commitment or it is a trap. The conversation to have in week one is with the person who owns the acquisition budget, and the sentence to say out loud is that this target depends on 15,000 additional sign-up starts, so if that spend does not happen the number is not 12,000 and we should agree that now rather than in week ten.

The second error is treating the decomposition as a one-off document. It is a live instrument: when the form change ships and completion moves 3 points rather than 5, you know immediately how much of the quarter is gone and which lever has to absorb it. Teams that set a target and next look at it in month three have converted a plan into a hope.

Decompose before you negotiate. A target you have expressed as named drivers with stated assumptions can be argued about, reforecast and shared; a target that exists only as a percentage can only be met or missed.