Time to value

Time to value is the time between a user signing up and that user first getting the outcome they came to your product for. A short time to value means people see the payoff early, before their attention fades. A long one means more people quit during setup and never reach the point where the product starts to matter to them.

How to measure it

Time to value = timestamp of first activation event minus timestamp of signup

Report the median for each weekly signup cohort, not the mean, because a handful of slow accounts distorts an average. Track it alongside your activation rate, since the two move together. Amplitude's summary of OpenView's research argues teams should continually test and tweak onboarding flows to reduce time to value, with activation rate as the indicator.

A case from benchmark data

Userpilot collected first-party data from 62 business-to-business (B2B) companies and tracked time to value plus drop-off points between signup and the activation point. It found no direct correlation between company size and time to value. Times were shortest at both the smallest and the largest companies. Userpilot's reading is that mid-sized companies in rapid growth may be giving onboarding less attention.

For context, OpenView's 2022 benchmarks found only 20 to 30% of new users reach activation at standout product-led growth (PLG) companies. Mixpanel's 2026 report, covering more than 12,000 companies, describes leading B2B companies anchoring growth strategy on in-product signals such as time to value.

Usual mistakes

• Defining value as an action your team likes, not the outcome the user wants.
• Reporting one blended number. Split by plan, acquisition channel and company size.
• Measuring the time without logging where people drop between signup and activation.
• Shipping onboarding changes without a control group. Size the test first with the sample size calculator.

When it matters

It matters most when signups come in steadily but trial conversion and early retention are weak. Pull the cohort data this week, find the step with the largest drop and put one fix for it into your experiment backlog.