Captive User Signal

Product measurement assumes a choice. Build something better, people choose it, usage rises. In an internal tool that people are required to use, the choice step is missing, so usage measures compliance and nothing else. A mandated tool with 100% adoption and 100% frustration produces exactly the chart a beloved tool produces.

Satisfaction scores degrade in a related way. Colleagues are polite in surveys and honest in their behaviour, and a mandatory tool is rated against the irritation it caused last week rather than against an alternative the respondent could have picked.

Why it matters

Internal platform teams routinely report adoption as success and are then surprised when leadership questions their value, or when a re-org gives teams a choice and they leave. The measurement told them nothing, and it took a change in circumstances to reveal that.

The stakes are practical. Time and money are being spent on the wrong improvements, because the feedback loop that would direct them has been replaced by a compliance statistic.

Implementation patterns

• Count workarounds. Spreadsheets, scripts, shadow processes and copies of your data built to avoid part of your tool. A workaround is someone paying real effort to escape, which is what choice looks like when choice is forbidden, and it is the strongest available signal.
• Measure time to first success: from a new user arriving to completing the task they came for. A mandate cannot fake it, and it is comparable over time.
• Track task abandonment. In an internal tool, a started-and-unfinished flow usually means the person went to find a human.
• Count repeated support questions per active user, which locate where the tool fails to explain itself.
• Treat unsolicited feature requests as engagement, because people only ask for improvements to tools they expect to keep using. Silence is usually resignation, not satisfaction.
• Make one part genuinely optional for one team for a month. Take-up is the only real choice signal available and beats a year of survey data.

Industry example

Platform engineering practice has converged on this correction. Internal developer platform teams increasingly report time to first success, voluntary adoption among teams that had a choice, and the number of teams building their own alternative, rather than percentage of services onboarded. Team Topologies and the platform-as-a-product literature since 2019 argue the same point from first principles: an internal platform with captive users must manufacture the feedback that a market would otherwise supply.

Failure scenarios

• Adoption reported as success, followed by mass departure the moment usage becomes optional.
• Surveys driving the roadmap, so effort goes to the most recently irritating thing rather than the most costly one.
• Shadow systems growing unnoticed, until a security review discovers customer data in a spreadsheet.
• A team declaring victory on onboarding, while time to first success is three days and every new engineer needs a person to hold their hand.

Trade-offs

These signals are harder to collect than usage numbers and harder to put on a slide. Workaround counts require asking people and looking, which costs time and some social capital, and making a feature optional to test it risks a visible rejection. The compensation is that the measurements are real: a roadmap built on them survives contact with an executive asking whether the platform is worth its cost.

When not to use it

When users genuinely have a choice, ordinary product measurement is better and cheaper: adoption, retention and churn mean what they normally mean. Do not import the captive-user apparatus into a team whose tool competes with alternatives; it will over-weight the complaints of the people who stayed.

Interview question

Q: Your internal deployment tool has 100% adoption because it is mandatory. Leadership asks whether it is any good. How do you answer in two weeks?

What a strong answer covers: rejecting adoption as evidence, without being defensive · counting workarounds and shadow tooling as the primary signal · measuring time to first success for the last ten new joiners · pulling repeated support questions to find where the tool fails to explain itself · proposing to make one flow optional for one team as a controlled test · and presenting the result with a mechanism, not a score: what specifically people avoid, and what it would cost to fix it.

Quick check

Quiz: Why is adoption meaningless for a mandated internal tool? Because it measures the mandate rather than the product; 100% usage is compatible with universal dissatisfaction.

Flashcard: What is the strongest signal from captive users? — The workaround count, because building one is real effort spent to escape, which is the only choice available to them.