Before asking people to try harder, find out where your application process makes a reasonable task unnecessarily difficult
By Arbab Naseebullah Kasi, Chief Executive Officer - CEO, Feel Worldwide Foundation Inc.
Someone wants to attend your training. They open the application on a phone, answer the first questions, and reach a request for a document they do not have nearby. They leave to find it. When they return, their answers have disappeared.
This is an illustrative situation, not a report about a particular organization. But it captures a useful distinction: leaving a form is not necessarily losing interest in the opportunity.
For a nonprofit or public-service team, an unfinished application can look like a weak response to outreach. The temptation is to send another reminder or make the invitation more persuasive. Sometimes the better intervention is inside the process itself.
A form should help someone make a clear request and help staff respond responsibly. When it becomes a test of patience, spare time, or familiarity with institutional language, the organization needs to examine what it has built.
Start with a question, not a conclusion
A completion rate tells you that something happened. It does not explain why.
Imagine a fictional service where 100 people begin an application and 60 submit it. The resulting 60 percent completion rate is straightforward arithmetic, not a benchmark for other organizations. The remaining 40 people may include someone who discovered they were ineligible, someone returning later, someone who found another option, and someone blocked by the interface.
Those situations need different responses. Making everyone finish is not the objective. Helping appropriate applicants complete an accurate request without avoidable difficulty is.
Define what counts as a start and a completed submission before comparing results. Keep the reporting period consistent. Ask whether returning applicants are counted twice. A changing measurement method can make a redesign look successful even when the experience has not improved.
Then identify where people stop, and treat that location as a clue to investigate rather than proof of a particular cause.
Watch the task unfold
Invite prospective users, with their consent, to try a realistic task in a test version. Use invented information rather than asking them to disclose private circumstances for the sake of research. Ask them to describe what they think a question means and what they expect to happen next.
Do not explain every confusing label immediately. If a participant pauses at “supporting evidence,” that pause is useful information. Find out what evidence they think the organization wants. The answer may reveal a gap that staff familiarity has made invisible.
Include people using phones and people who use assistive technology. Someone who knows the program well may navigate around problems that a newcomer cannot.
You are looking for observable difficulties: rereading a question, searching elsewhere for a document, losing an answer, or being unable to reach a control. Record the issue without turning a person's hesitation into a judgment about their ability.
Make every question earn its place
The GOV.UK Service Manual recommends a question protocol: an internal account of why information is needed and how it will be used. It also recommends routing people to questions relevant to them, rather than making everyone navigate every possibility. These are service-design principles, not a reason to copy another institution's eligibility rules. GOV.UK guidance on structuring forms.
For your own form, ask a named staff member to explain the decision each answer supports. “We have always collected it” is not enough. Neither is “It might be useful someday.”
Consider a fictional workshop inquiry that requests a detailed employment history before revealing the available dates. The team might discover that employment history is unnecessary for an initial inquiry, while availability is essential. Moving a question to a later, justified stage can be more sensible than demanding everything at the entrance.
Do not remove a requirement casually if it supports a real funding, safeguarding, or eligibility obligation. Establish its purpose and discuss changes with the responsible person. The aim is a defensible process, not simply the shortest possible screen.
Prepare people before they begin
Tell applicants what they are about to do. Explain who the process is for, what information or documents they will need, and whether they can pause and return.
Any completion-time estimate should come from observing the actual task, not from timing a staff member who already knows every answer. If a document has to be obtained elsewhere, acknowledge that preparation separately from the time spent entering information.
Be explicit about the outcome. An expression of interest is not a confirmed place. A submitted application is not an approval. The words on the start page should match what the service can actually promise.
This is also the place to explain an available assistance route. Do not advertise telephone support that nobody can staff. A modest, dependable option is more useful than an impressive list of channels that lead nowhere.
Treat correction as part of completion
A person should not need to guess why a form refuses to continue. “Invalid input” gives almost no direction. A useful message identifies the problem and explains the next action in language suited to the field.
The GOV.UK Design System pairs an error summary with messages beside the affected answers. Its guidance links each summary item to the relevant input and directs keyboard focus to the summary. This gives someone a route to the problem rather than merely announcing that a problem exists. GOV.UK error-summary guidance.
Test mistakes deliberately. Leave a required field empty. Try a supported document near the stated size limit. Use the back button. Check that previously entered information remains where it should.
Do this in a safe test environment, not by submitting fictional applications into the live service. A fix is not complete because a developer says it was deployed. Repeat the task and observe whether the original difficulty has gone.
Accessibility is part of the main route
The World Wide Web Consortium's Forms Tutorial, updated March 27, 2026, describes accessible labels, instructions, related-field grouping, and feedback. It also warns that unnecessary or excessive requests for information can contribute to abandonment. These are practical design considerations, not a numerical prediction of how much your completion rate will rise. W3C accessible-forms tutorial.
Do not equate a clean-looking page with an accessible one. The appearance of a label does not establish that software can correctly associate it with its field. Test keyboard use and screen-reader interaction, and involve people with relevant experience.
Give accessibility problems an owner and a correction date, just as you would a broken submission button. A barrier affecting fewer applicants is still a barrier to the service you intended to offer.
Make the ending unambiguous
Before submission, offer a clear opportunity to review and change answers. The GOV.UK check-answers pattern is designed for this point in the journey. GOV.UK check-answers guidance.
After submission, confirm receipt and explain what happens next. Give a realistic response expectation only when the team can support it. If the applicant needs to take another action, distinguish that action from the application they have just completed.
Finally, review the results alongside staff experience. Are fewer people getting stuck? Are applications more accurate? Has the redesign created new follow-up work or shifted difficulty to a later stage?
Choose one observed obstruction, repair it, and test again. You do not need to begin with a complete rebuild. Sometimes the most useful change is a document requirement explained earlier, a question removed for a sound reason, or an error that finally tells someone how to proceed.
The people leaving your form are not a single category called “unmotivated.” They are people making their way through a process you can improve. Where would watching one person try it teach you more than sending another reminder?

