Digital Government Metrics Worth Measuring

A government service can post record traffic and still fail every citizen who used it. Page views, sessions, and visits measure arrival, not outcome. They rise when a form is confusing and people reload it, and they rise again when applicants give up and start over. The number goes up; the citizen still does not have their permit, their benefit, or their answer.

The metric that matters is whether the person who started the transaction finished it, and whether finishing produced the result they came for. Almost everything worth measuring in digital government follows from that single shift.

The measurement gap is also a trust gap

When citizens are surveyed alongside private-sector industries, government lands last. Independent benchmarking makes the gap concrete: in Forrester’s 2018 US Federal Customer Experience Index, 80 percent of federal agencies scored in the lowest two experience categories, against just 20 percent of private-sector brands.

This is not a perception problem to be managed with better messaging. It is an outcome problem that the standard analytics stack is built to hide. Traffic dashboards reward visibility, not resolution, so a service can look healthy while citizens churn through it.

Page views go up when a form is broken. Completion rate goes down. Only one of those numbers tells you the truth.

The reason to close the gap is that experience and trust move together. McKinsey’s public-sector research finds that satisfied citizens are nine times more likely to trust the agency serving them, and nine times more likely to believe that agency is achieving its mission. Service quality is not a soft outcome sitting next to the mission. It is a lever on the mission.

A scorecard built from outcomes, not traffic

The UK’s Government Digital Service settled this debate years ago by mandating four KPIs for every transactional service: completion rate, digital take-up, user satisfaction, and cost per transaction. It is a deliberately short list, and the discipline is in the brevity. Five numbers a minister can act on beat fifty a dashboard can display.

We extend that spine into a measurement model with five questions every digital service should answer:

1. Completion. Of the citizens who start a transaction, how many finish? Completion is the percentage of digital transactions completed against all started, and a failed transaction is anyone who abandons before the confirmation page. This is the single most honest number a service produces.
2. Effort. How long does it take, how many steps, how many contacts to a call center? Effort is the upstream cause of abandonment. When effort falls, completion and satisfaction tend to rise together.
3. Satisfaction and ease. Do users rate the experience well, and would they choose the digital channel again? Ease predicts repeat use, which is what take-up depends on.
4. Reach and equity. Who is actually using the service, and who is excluded? A 90 percent completion rate that quietly omits offline or assisted-digital populations is not success. It is a coverage gap waiting to become a headline.
5. Outcome and cost. Did the citizen get the right result in the right time, and at what cost per transaction? This is where digital investment becomes accountable rather than assumed.

Tie every metric to an outcome, or drop it

Start from the outcome the agency owns, take-up of a benefit, time to resolution, accuracy of a determination, then work backward to the digital metrics that predict it. If completion rate on an application correlates with program take-up, completion becomes a funded lever rather than a vanity figure. Report each metric split by segment and channel, because an average hides the population a service is failing. And retire anything that cannot be traced to an outcome, starting with raw visit counts.

This is the same logic we apply across our practice: measure what moves the result, not what is easy to count. To put an outcome-linked scorecard behind your services, explore our Digital Government and Voice of the Citizen work, or browse the case studies.