Time to value: what it tells you when activation drops
The activation rate fell from 34% to 29% this month. Before anyone changes the onboarding or the campaign mix, a second number shows where to look: time to value, how long the users who did activate took to see what the product does for them.
The same drop, two causes
| Usual month | Drop A | Drop B | |
|---|---|---|---|
| Activation rate | 34% | 29% | 29% |
| Days to activate, usual sources | 2 | 2 | 4 |
| Cause | New campaign | New setup step |
The time to value row only counts users from the sources you already had. In the first case, they activate as fast as before. The product works as it did, and the drop comes from the people the new campaign brought. In the second case, the same users take twice as long. Their profile didn't change, so something in the product got harder.
Comparing within the same sources matters. A new audience that is less familiar with tools like yours can take longer to activate on a product that works fine.
What is time to value, and how do you measure it?
Time to value is how long a new user takes to see what the product does for them. That moment is often called the aha moment.
You can't record the aha moment itself. So teams measure up to the activation event instead, the first time the product does its job for the user, such as a first report built from connected data. The user may only be fully convinced a little later, but it is the first point you can measure.
Time to value (days) = activation date − sign-up date
For a group of users, take the median rather than the average. Otherwise a few users who activate weeks later pull the figure up.
The activation rate tells you how many users get there. Time to value tells you how fast.
Read both per weekly or monthly cohort, over several periods. A shift that shows in one cohort is often noise, while a shift that holds over several cohorts is a trend.
Where the problem starts
-
Is the drop bigger than the usual swings between cohorts?
No Probably noise. Wait for a few more cohorts.
Yes Next question.
-
Is it concentrated in one or two acquisition sources?
Yes The audience changed. Talk to that channel's owner, numbers in hand.
No Next question.
-
Do users from your usual sources take longer to activate than before?
Yes The product got harder. Find the step where users stall, starting with recent releases.
No The audience shifted across sources. Check what changed in positioning, pricing or targeting.
What each cause calls for
An audience change starts with the channel or campaign owner, using the segment data rather than an assumption. The cause can sit in the promise itself, since a campaign that promises something the product doesn't do brings sign-ups no onboarding will win back. A product change starts with the step where users stall, and the wording on that step is often the cheapest thing to test before rebuilding a flow. Users who left before activating need something to bring them back, such as an email pointing to the exact step where they stopped. For accounts worth a conversation, an offer of help from a person can do what an email can't.
Once the cause is clear, a dated note on that period keeps the explanation next to the number, so the same dip isn't debated again at the next review. None of this works if the activation event itself changed in the meantime, since a new definition can look exactly like a drop.
See activation and time to value by source and cohort, in one view.
Start free trialWhat's a good time to value?
There's no universal benchmark worth quoting, since it depends on the product and on the activation event chosen. A product set up in minutes and one that needs a data connection can't be compared. What matters is your own trend, cohort after cohort.
Should the trial last as long as time to value?
It should at least leave room for it: if most users need eight days to activate, a seven-day trial ends too soon. A much longer trial has its own cost, though, since it delays payment and can lose people who were ready to buy. Letting active users extend their trial, or letting early buyers start their paid plan with billing deferred to the trial's end, covers both cases.