As we have developed our private beta minimal viable product for the eligibility checker, the number of questions we ask users has increased. By this stage they need to answer at least 26 questions.
We needed to understand how users responded when completing the full journey, especially how the length and order of questions affected their experience.
The design problem
Childcare funding eligibility is complex and cannot be assessed without asking detailed questions. Previous research had shown that users expect a checker to ask personal questions, but not to feel like a formal application process.
We had a concern that if the journey is too long, it might no longer meet users’ expectations of a quick and easy checker. In earlier rounds of research, users suggested a longer journey could feel too much like a formal application.
We also needed to consider cognitive load, particularly for users with additional needs or limited time.
At the same time, we needed to ask enough questions to give users confident answers.
What we considered
In earlier versions of the checker, we had grouped similar questions for the user and their partner together. For example, asking about work status for the user, then their partner, then their incomes, and so on. However, following previous rounds of research, we moved away from that pattern. The user now answers all questions about themselves first and then answers all questions about their partner.
Our hypothesis was that frequent switching between who the question was about increased cognitive load and made the journey feel longer. The hope was the new approach would simplify things.
Ahead of testing the full journey, we discussed ways we could shorten or simplify the journey for users, such as:
- removing questions or allowing for ‘early exit’ if the logic identifies that further questions might not be relevant
- adding additional break pages at the start of each sub-journey
However, rather than simplifying in advance and risking removing necessary detail, we chose to test the full journey and observe where users actually struggled.
What we designed and tested
This was the first time we were testing the complete end‑to‑end journey, rather than individual sections. You can read other design histories on how we design and tested sections for:
For the full journey we were testing, users answered questions in the following order:
- About their children.
- About themselves.
- About their partner, if they live with one.
We used standard GOV.UK question patterns throughout and tested the journey with users using both mobile and desktop.
The aim was to understand perceived length and effort when users completed the full set of questions.
What we learned
The round of research on the results page took place in April 2026 and involved testing with 11 participants. The sample included users with a range of access needs, income levels and childcare entitlements, and reflected a mix of ethnic backgrounds.
This allowed us to gather feedback that reflected a variety of circumstances rather than a narrow user group.
Perceived length and effort
Participants consistently described the checker as quick and easy to use, and in line with what they expected from a checker like this.
No participants said the journey felt too long. Several explicitly said that the information they received at the end justified the effort required to answer the questions.
This feedback suggested that length is not currently a clear risk in this context.
Order of questions
Participants felt it was logical to answer questions about their child first, as they were checking what support they could get.
They also responded positively to answering questions about themselves before their partner. Completing one set first reduced context shifting and helped them become familiar with the questions which made it easier when answering about a partner. This supports grouping questions by person rather than switching between them.
Where users experienced difficulty, this was usually linked to the clarity of individual questions rather than the overall structure or order of the journey.
Accessibility and familiarity
Familiar GOV.UK patterns played an important role, particularly for users with access needs.
Participants described the service as reassuring and easy to navigate because it looked and behaved like other GOV.UK services they already trusted.
What we still need to learn
Our testing on the number and order of questions took place in controlled research sessions. Participants had time and space to focus in a way that may not reflect real‑world use, where people may be tired, distracted or short on time. We need to explore how the journey performs in more realistic conditions.
The prototype used in this research was not fully dynamic based on user answers. We need to test with a more developed prototype or live version to understand how users respond when the service is generating results based on their real circumstances.
We also need to explore responses from a wider range of user cases, including those who are not eligible for any support.