Someone submitted an application online, saw a green tick and closed the page, believing the task was complete. A few days later, they discovered that the application had only been received: review and a response were still required. The confirmation screen reported one stage, but the person understood it as a promise that the whole service was finished.
Public-service marketing helps explain the path from need to outcome. Teams need separate measures for submission, channel choice and completion of the work that follows. For an appointment-booking service, a useful outcome combines successful bookings, completion time, errors and satisfaction. Creating an account remains an intermediate action.
Name the operation being measured
Each process has its own boundaries. A form may end with a submitted application, booking with a selected and confirmed time, and payment with a completed transaction. Visiting an information page does not yet show that someone has started the operation.
The GOV.UK Service Manual relates completion rate to completed and started digital transactions. For an application process, completion may mean submission regardless of the later decision. This is a measurement method from a British manual; a team should agree events for its own service with the service owner and delivery staff.
Metric names should preserve the chosen boundary. “Completed submission” describes the form. “Received an application response” concerns the next stage. Calling both events “success” makes it harder to locate delays and assign responsibility for improvements.
Link repeat attempts to one case under an agreed rule. Otherwise, someone who returns to document upload several times inflates the number of operations started. Separate technical tests from user applications. Before comparing percentages, the analyst defines a single operation and when its observation period ends.
Separate three calculations
Consider a service whose application can be submitted on eGov or at a TsON public service centre. All figures and expenses below are made up for the example. Of 1,000 eGov cases that started the process, 600 completed submission within a sufficient observation period. Repeat attempts are linked and technical tests excluded. TsON centres accepted another 400 unique applications over a comparable period, without duplicating the digital ones.
| Question | Calculation | Value |
|---|---|---|
| What share of those who started finished digital submission? | 600 / 1,000 | 60% |
| What share of accepted applications arrived digitally? | 600 / (600 + 400) | 60% |
| What share of digital applications reached a response? | Later processing information is needed | Not yet known |
Both figures come to 60% by coincidence: the denominators differ. The first denominator contains digital starts; the second contains accepted applications across all channels. The digital take-up method uses completed digital transactions and completed transactions across all channels over the same period. In our example, an accepted submission provides a common boundary across channels.
Suppose that by the reporting date, people had received final responses for 480 of the 600 digital applications, while 120 remained in processing. The share with a response is 480 / 600 = 80%. A response may be a refusal with an explanation, so this differs from the approval rate. Assessing delays needs each application's age and applicable turnaround time.
For journey analysis, connect events within the same cases. A calendar report shows another side of operations: how many applications arrived and how many responses were issued during the month. Some responses may concern earlier submissions. Dividing monthly responses by monthly starts does not give a cohort conversion rate without that connection.
Explain the budget without inventing savings
Suppose an information campaign for this service cost 240,000 ₸. The budget divided by all 600 digital submissions is 400 ₸; divided by 480 responses received, it is 500 ₸. These figures relate spending to a chosen volume of service activity. They do not yet measure cost per additional submission caused by the campaign: some people would have used the service anyway.
Estimating additional outcomes needs a suitable comparison and awareness of other changes. Seasonal need, office availability, the condition of the form and new guidance can affect applications simultaneously. Separate observed volume from conclusions about advertising effects in the report.
Measure time savings across the whole journey too: preparation, completion, waiting, corrections and assistance. A short time inside the form may follow a long search for the right document. Even measured hours saved need to be connected to changes in work before claiming a financial budget saving.
Investigate obstacles with the user
If a substantial share of cases stops at file upload, events identify where to investigate. Watching people complete the process and reading support enquiries help explain why. Someone may misunderstand the document name, lack the required format or fail to find an explanation of an error on their phone.
Age, language and address do not give a ready-made answer about digital skills. Observe what happened at the step and which help the person sought. Assess changes through both upload completion and the quality of subsequent applications: faster submission of incomplete documents moves work to delivery staff.
Build guidance around the task: who the service is for, what to prepare, where to start, its actual costs and timing, and what follows submission. Obtain the conditions from the service owner. An invented deadline that makes an advertisement persuasive creates an expectation the delivery team never promised.
“Application received. Check its status in the applications section” is suitable if the system really accepted it and offers that way to track it. “Service delivered” requires a different process state. The marketer agrees these words with the product team and support so everyone explains the same journey.
Preserve access to help
Someone may start independently on eGov, ask a TsON consultant for help and finish online. Distinguish unassisted completion, assisted digital completion and service at a TsON. For digital-channel share, account for assistance under an agreed rule while retaining a separate flag for staff involvement.
Assess language versions in full: instructions, fields, errors and final status. Also test the journey using a keyboard and assistive technology. A translated heading does not solve the problem if an error explanation remains unclear.
If an office visit is required after submission, people should know in advance. Removing a needed support channel to improve the digital share puts service quality at risk. Alongside channel share, track successful completions, time, errors and feedback about assistance received.
After a one-off life event, repeat use may never be needed. Measure retention where a recurring need exists. For a single application, completing the journey and obtaining a clear outcome matter more than opening the app again next week.
Exercise: explain a higher digital share
In the next period, digital submissions remain at 600 while other channels receive 200. Calculate the new share, describe the change in words and name the information needed for a campaign decision. Then define start and outcome boundaries, an assistance-counting rule and one quality measure for your own service.
The new digital share is 600 / (600 + 200) = 75%, up from 60%. Online volume stayed unchanged while other-channel volume fell. Demand, access to help or tracking may have changed; the calculation cannot identify which. You also need starts, responses received, turnaround times and reasons for non-completion.
For an appointment-booking service, a strong definition connects successful bookings to time spent, errors and satisfaction. Growth in accounts created leaves the completed task unresolved. This report helps the team improve access to the service and explain spending outcomes without substituting intermediate actions for the goal.