For many childcare schemes, eligibility depends on a parent's work status and income. However, people on leave often do not fit neatly into these rules. They may:
not currently be working
be receiving statutory pay
not yet know exactly what their circumstances will be when they return to work
We wanted to understand whether users could answer questions about returning to work and whether we could collect enough information to provide useful results without adding unnecessarily complexity.
What we already knew
Previous research showed that users found childcare eligibility difficult to understand, particularly when rules depended on future events or changing circumstances.
We also knew from policy discussions that several childcare schemes contain specific provisions for parents on leave. Eligibility can depend on:
the type of leave a user is taking
whether they are receiving statutory pay
their work status before leave
their expected circumstances when they return to work
which child they are taking leave to care for
This created a more complex challenge than the standard work-status journey, where eligibility is largely based on a user's current situation.
Research in earlier rounds suggested that many parents were already thinking about childcare support while planning a return to work rather than after they had returned. This meant the service needed to support decisions about future circumstances, not simply describe a user's current situation.
The design problem
We needed to determine eligibility for users whose current circumstances may not reflect the basis on which childcare support is assessed.
For many parents on leave:
current earnings may not be relevant
future earnings may be uncertain
work status may change when leave ends
different children within the household may be affected by different rules
This created several potential problems.
Asking too few questions risked producing inaccurate eligibility results.
Asking too many questions risked overwhelming users with questions about circumstances they may not yet know.
The challenge was to collect enough information to support meaningful results while keeping the journey proportionate to a user's circumstances.
We needed to understand whether users could provide information about future circumstances that they themselves were often still deciding.
What we considered
Treating users on leave as not currently working
This would have simplified the journey and reduced the number of questions.
However, policy rules for several childcare schemes recognise leave as a distinct circumstance. Treating all users on leave as simply not working would have risked producing misleading eligibility outcomes.
We did not take this option forward.
Asking only about current circumstances
This would have reduced complexity and avoided asking users to predict future earnings or work arrangements.
However, current circumstances alone would not always provide enough information to determine eligibility. Some users may qualify for support when they return to work even if they do not currently meet work-related criteria.
We did not take this option forward.
Collecting both current and future circumstances
This approach was more complex to design, but it offered greater flexibility and aligned more closely with policy rules.
For research in May, we decided to explore this approach.
What we designed
We created a dedicated leave journey from the work-status section of the service.
Users who indicated they were on leave were directed through a series of additional questions designed to capture information relevant to childcare eligibility.
The journey was intended to accommodate a wide range of leave circumstances while avoiding unnecessary questions wherever possible.
Asking about leave type and statutory pay
We introduced questions that asked users:
what type of leave they were on
whether they were receiving the relevant form of statutory pay
This reflected policy rules where leave type and statutory payments can affect eligibility for childcare support.
Rather than assuming all users on leave had the same circumstances, we wanted to understand whether users could accurately identify these distinctions.
Identifying which child a user was on leave for
One of the more significant design decisions was introducing a question asking which child the user was taking leave to care for.
This reflected a growing requirement to provide child-specific results.
For some schemes, being on leave related to that child can impact eligibility or things like application and start dates. However, if you have another child for whom you are not on leave, they will remain unaffected.
By capturing the child that leave relates to we can provide more accurate results for each child.
Capturing circumstances before leave
Where relevant, users were asked about their work situation before taking leave.
This helped support eligibility decisions relating to children who were not the reason for the leave period.
Capturing circumstances after leave
We also added questions about users' expected circumstances when returning to work.
These included:
expected work status
expected earnings
self-employment circumstances
We recognised however that there was a risk that not all users would know the answers to these questions with certainty.
Reducing unnecessary questions
A key principle we followed was only asking questions needed to support an eligibility assessment.
Users whose circumstances were not expected to change could avoid some future-focused questions.
Not asking for a return-to-work date
We considered asking users when they expected to return to work. Knowing a return-to-work date would allow us to provide more specific application and entitlement dates for some childcare schemes.
However, this would have introduced additional complexity. We would have needed to build logic to calculate dates across multiple schemes and account for different leave scenarios. The team did not have the capacity at the time to undertake this work.
We also knew that some users would not yet know when they planned to return to work. Asking for a date could therefore create additional uncertainty and reduce confidence in the results.
As a result, we decided not to capture a return-to-work date. To avoid providing inaccurate dates, where eligibility depended on returning to work, we decided to show explanatory content on the results page.
For example, we used variations for:
when to apply of "The date you return from parental leave affects when you can apply"
the start date of "When you are returning from parental leave"
We recognised that this would make some results less specific for users on parental leave. This was a temporary trade-off, and we plan to revisit the approach after MVP.
What we learned and how we iterated
The round of research we conducted took place in-person at 2 libraries in Essex and involved 16 participants. This included 7 participants who were on maternity leave, which allowed us to test the leave journey with people currently experiencing the circumstances we were designing for.
Participants also included users with differing levels of confidence using digital services and a range of employment situations, including self-employed or not currently in work.
Most participants could complete the leave journey
Overall, participants were able to progress through the leave questions without major issues.
Most participants could complete the leave questions. However, several important areas of uncertainty emerged.
Being on leave did not always match users' mental models
One of the clearest findings related to the entry point into the leave journey.
Several participants on maternity leave initially wanted to answer "Yes" when asked whether they were in paid work because they still considered themselves employed. Some only noticed the leave option after first looking for a way to identify themselves as being in work.
This highlighted a difference between how policy categorises leave and how users understand their own circumstances.
Although most participants ultimately selected the correct option, the findings suggested there was a risk that some users could follow the wrong route through the journey if they moved quickly through the question.
Statutory pay terminology was not always understood
Several participants were unsure how to answer questions about statutory maternity pay.
Some described receiving income that was partly provided through their employer and partly through statutory pay, making it unclear which answer best reflected their situation.
This raised broader questions about whether users understand terminology commonly used within the eligibility rules and whether additional explanation may be needed.
Future circumstances were often uncertain
The most significant challenge was accounting for future circumstances.
Many participants were still deciding:
whether they would return to work
how many hours they would work
whether their earnings would change
when they would return
Several participants said they would choose "I'm not sure" when asked whether their work status or earnings would change after leave because they had not yet had the necessary conversations with employers or made final decisions.
Others wanted examples to help clarify what counts as a change in work circumstances.
One participant with dyslexia also struggled to understand whether a question referred to circumstances before leave or after leave, suggesting that time periods may need to be communicated more explicitly for some users.
These findings suggested that uncertainty was not an issue with understanding content but a part of the life stage users were in.
Parents wanted to understand future eligibility
A recurring theme from this research was that some participants were using the service to help plan their return to work.
Rather than looking for confirmation of their current entitlement, they wanted to understand what support might become available if their circumstances changed.
One participant questioned whether the service should recognise people who were about to return to work. Another wanted to understand childcare support options before applying for jobs because childcare availability would affect whether returning to work was realistic.
These findings raised important questions about whether the service should focus solely on current eligibility or also help users explore future scenarios.
The findings reinforced the need to balance accuracy and certainty
The research showed that many of the questions in the leave journey were necessary because childcare eligibility rules depend on future circumstances rather than current ones. However, it also showed that users often cannot answer those questions with complete confidence.
This left the team balancing two competing needs:
collecting enough information to provide accurate eligibility information
avoiding questions that users could not realistically answer
For our next iteration to this elements of the leave journey were simplified due to delivery constraints. We will document this in a future design history.
What we still need to learn
Several questions remain unresolved.
We still need to understand:
how much future information users can realistically provide
how well different types of leave are represented in the journey
how confident users feel answering questions about future circumstances
how users understand child-specific eligibility and application dates
how well users on leave understand their results and apply them to their circumstances
how effectively the journey works for people using shared parental leave and self-employed users with more complex circumstances
We also need to understand whether some eligibility decisions could be communicated better through results and guidance rather than through additional questions.