State governments should evaluate digital transformation as a change to a public service—not simply as a website, app, or technology project. Define the outcome the service exists to achieve, record a baseline before making the change, and track both how the service operates and what happens to the people who rely on it. Where feasible, compare results with a credible comparison group or counterfactual. Digital take-up, completion, satisfaction, and transaction cost are useful signals, but none alone proves that a project caused better public outcomes.
Start with the service outcome, not the technology
Before selecting measures, describe the problem the service is meant to address and the result residents need. Then map how the proposed change is expected to produce that result. This short logic model, also called a theory of change, connects project activity to service performance and ultimately to a public outcome.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Public Administration (The Basics) | $26.99 | Buy on Amazon |
| 2 |
|
Public Administration: A Very Short Introduction | $7.99 | Buy on Amazon |
| 3 |
|
Public Administration | $110.08 | Buy on Amazon |
| 4 |
|
Introducing Public Administration | $119.97 | Buy on Amazon |
| 5 |
|
Introducing Public Administration | $144.49 | Buy on Amazon |
As an Amazon Associate I earn from qualifying purchases.
- Public problem: What need, delay, error, or access barrier should the service address?
- Users and task: Who needs the service, and what are they trying to accomplish?
- Service change: What changes for users, staff, or operations?
- Immediate result: What should improve first—for example, a completed application or a more accurate decision?
- Public outcome: What broader result is the program responsible for supporting?
- Assumptions and risks: What must be true for the change to work, and what harm or unintended burden could it create?
A launch, download, visit, or account registration records activity; it does not establish that someone completed a task or received the intended service. The U.K. Department for Business and Trade’s Digital and Data Playbook recommends identifying indicators from a theory of change and, ideally, tracking them against a baseline and comparison group. That is a useful evaluation method for U.S. states, not a state-government requirement.
Set a baseline and decide what comparison is credible
Record pre-change performance before implementation or a major service redesign. For every measure, document its definition, numerator and denominator where applicable, population, time period, data source, and intended breakdowns. Also note context that could affect results, such as changes to policy, eligibility, staffing, or demand. Without those details, an apparent improvement may reflect a changed population or definition rather than a better service.
#1 Best Overall
A before-and-after comparison can show whether a measure moved, but it cannot by itself establish why. When feasible, compare the changed service with a similar population, location, or service that did not receive the change, or use a phased rollout to create a comparison. Choose an approach that fits the service and state what it can and cannot show. The Playbook recommends comparison groups where possible; it does not prescribe one design for every agency.
Use consistent service boundaries when comparing options or time periods. If a digital form is counted as one service but calls, in-person help, and follow-up work are excluded, the resulting figures may make a channel shift look like a reduction in cost or effort when work has merely moved elsewhere.
Rank #2
Pair service-operation measures with outcome measures
The U.K. Government Service Manual identifies four useful service-performance measures. They help show whether a digital channel is operating well, but should be paired with measures tied to the particular program’s purpose.
Recommended Free Tools
| Operational measure | What it tells you | What to define or pair it with |
|---|---|---|
| Completion rate | The share of digital transactions started that users successfully complete. | Define what counts as a start and a successful completion; pair with the accuracy, timeliness, or resolution of the completed task. |
| User satisfaction | How users report their experience with the service. | Record when and how feedback is collected, and consider who responds; satisfaction alone does not show whether decisions were correct or needs were met. |
| Cost per transaction | The government cost each time a user completes the task. | Use consistent cost boundaries, including relevant staff effort, support, rework, implementation, and ongoing operation; account for assisted and non-digital routes. |
| Digital take-up | The proportion of users choosing digital rather than other available channels. | Compare with channel availability and user outcomes across channels; increased digital use is not itself proof of improved access or results. |
Choose additional outcomes to match the service. Depending on its purpose, these could include decision accuracy, time to resolution, avoidable repeat contact, or the program result the service is intended to support. A dashboard of channel activity is not a substitute for measuring those outcomes.
Rank #3
Check access, experience, and the effects on operations
Average results can conceal people who cannot use the digital route, need substantial assistance, or receive worse outcomes. Examine relevant groups and channels rather than treating all users as one population. The appropriate breakdowns depend on the service and the data an agency can responsibly collect.
- Use user research and feedback to find friction that transaction logs cannot explain.
- Check accessibility and whether people can complete the task themselves; identify where assistance is needed.
- Track demand and results across digital, telephone, in-person, and assisted routes when those channels remain available.
- Measure material operational effects, including staff workload, errors, rework, and service reliability.
- Investigate whether a faster digital transaction shifts effort to residents, call centers, or front-line staff.
For context, Adobe’s 2026 state-portal index treats customer experience, site performance, and digital self-service as separate dimensions. Such a benchmark can help frame portal comparisons, but its score is not evidence that a particular state project caused a service improvement.
Rank #4
Compare transformation options on the same terms
When leaders must choose between options, compare them using the same population, time period, and service boundary. A synthesized set of decision axes can keep the discussion focused on public value rather than technology novelty:
| Comparison axis | Question for decision-makers |
|---|---|
| Public outcome | Which option better advances the result the service is meant to deliver? |
| Service quality and experience | How do completion, satisfaction, accuracy, timeliness, and reliability compare? |
| Access and inclusion | Can relevant groups use the service, including people who need accessibility features, language support, or assistance? |
| Whole-service cost and productivity | What are the transaction costs, staff effort, rework, support, implementation, and ongoing operating costs? |
| Risk and compliance | What security, privacy, legal, or continuity considerations matter for this service? |
| Evidence strength | How sound is the baseline, comparison, and data coverage—and how confidently can the change be attributed to the intervention? |
These axes are a practical synthesis, not a mandatory common framework for state agencies. Tailor them to the service and policy outcome.
Best Value
Make measurement part of service management
Assign an owner to each important service and outcome measure, then review the indicators throughout discovery, delivery, and live operation. A change in a dashboard should prompt investigation: check the underlying data, look for changes in context, and use user feedback to understand what happened. Performance evidence is most useful when it informs service improvements and future investment decisions.
Where findings are published, give readers the definitions, time periods, methods, limitations, and results needed to interpret them. Local Digital’s 2026 evaluation illustrates a U.K. government-programme approach that combines qualitative and quantitative evidence to consider delivery, outcomes, value for money, and lessons. It is an example of evaluation practice, not a result about U.S. state services.
Use state benchmarks as context, not proof of impact
Benchmarks can show how a service or portal compares under a particular methodology; they do not establish that an individual transformation caused a public-service outcome. The Beeck Center’s state digital-service landscape resource explicitly does not measure state service performance, so it should not be treated as a performance ranking.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Clear out junk files and repair common Windows errors3Scan for outdated or missing drivers - takes under a minuteA 2019 McKinsey & Company survey of 27 state chief administrative officers reported which measures respondents ranked among their top three digital-initiative success metrics. It found that 89% included improved quality or accuracy, 82% customer service, 53% cost reduction, 29% improved speed, and 45% compliance. These are historical survey responses from that sample—not current estimates for all state-government leaders or measured results of digital projects. Percentages may not total 100% because of rounding.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
