Single directory: initial testing with practitioners

In our first design history on creating a single directory, we explained why we merged the ‘Find’ and ‘Connect’ directories to create a single directory that simplifies access to services.

We are now beginning to validate that thinking through usability testing. This design history is about the first 2 rounds of testing we conducted in November and December 2024. This research focuses on practitioners who work with families, the next phase will include parents and carers.

The problem: ensuring value and usability

Our challenge is to design a single directory that balances usability and value for both practitioners and public users. If practitioners can’t consistently find services, they may revert to existing tools like personal lists or their colleagues's knowledge.

This has been captured in previous insights about user needs, such as:

“As a practitioner I need to know what help is available locally for parents and children in different circumstances, identify what’s relevant for them and know how to request support”

To understand how we might address this need, we needed to gather more insight and validate any assumptions around their use of the single directory. We wanted to explore:

• existing ways of working
• how practitioners search for services
• what information they expected to see
• the factors that influence decision-making when selecting a service
• any barriers or frustrations they faced when using the directory

Our research approach

We conducted the first rounds of testing with practitioners because are a key focus for the team. We need to work closely with local authorities (LAs) to ensure the directory meets their needs.

We started testing with the existing ‘Connect’ prototype as it already had many features we expected to be needed for the single directory minimum viable product.

We then:

1. Made some changes to the filters to make them compliant with the Open Referral data standard we are using.
2. Populated the prototype with additional dummy data to support our testing scenarios.

Who we did research with

We conducted research with 10 practitioners in total, made up of:

• one voluntary, community and faith sector (VCFS) practitioner
• one family hubs co-ordinator
• one family hubs business support worker
• 7 early help and intervention practitioners

Participants were from the Midlands, Greater London, Yorkshire and the South of England.

How we tested it

We used the same methodology across both rounds of testing to ensure consistency. This approach allowed us to compare results and identify patterns in user behaviour. The process for both rounds included:

1. A discussion about the practitioners’ ways of working and how they currently search for services.
2. Initial impressions, where we shared a prototype version of the search results page and noted participants’ reaction and anything they found confusing or unexpected.
3. Real-world use case, where practitioners were asked to search for a service based on a real example from their professional experience. This provided insight into their natural workflows.
4. Scenario-based testing, where we provided structured scenarios, such as helping a parent looking to stop smoking or finding support for a child with autism. These scenarios highlighted how practitioners navigated the directory in varied contexts.

Before each round we set some key research objectives. Then following the round we analysed the insight as a team to produce some key insights and recommendations.

Round one: objectives and insights

For the first round our key objectives were to observe:

• the practitioners’ journey for finding a relevant service to identify any pain points and blockers 

• how search results are narrowed down to gain a better understanding of needs
• the potential impact of some of the restrictions that our data standard places on the filters and fields we can use, such as if a service is online or in person

Key insights gathered included:

• participants were generally positive about the experience of the single directory
• participants were positive about being able to filter their results, however some usability issues were flagged
• whether a service was provided by a LA was not a key consideration
• the key consideration when choosing a service was whether it could provide the right support for a family
• that not having information like whether the service was delivered in-person was not off-putting, if there were contact details so they could follow up

Iterations made to the prototype following round one

Following a team analysis of the feedback we developed some ‘How might we’ statements, including how might we:

• make it easier for users to filter results?
• improve the service details provided?
• make it easier for users to find more service details?

Based on these and other observations we made some updates to our prototype ahead of round 2.

Category filters changed to accordion style

We changed the filter navigation for categories to use an accordion pattern instead. The aim was to reduce the need for excessive scrolling and to improve navigation. We would then test this in the second round by observing if users interacted more effectively with the category filters and how clear the headings for each section were.

Additional changes to other filters

We made changes to 3 of the other filters based on observations from the first round. For:

• ‘Days service is available’ we added prompt text ‘Select specific days, if applicable’ as we had observed users unnecessarily selecting all days
• ‘Cost’ we changed this to have just one option ‘Only show free services’ as we hadn’t observed any users specifically wanting paid for services
• ‘Age’ we changed the ranges based on feedback about the groupings being too large for younger children, and more closely aligned them with key stage ages

Changes to the search results page

We made some small changes to the way search results were displayed. These were:

• adding the name of the user’s LA to provide clarity on the area they were searching within
• adding the distance from the postcode next to the service name to make it clearer to users how the results are ordered
• increasing the size of service titles as we had observed some users struggling to identify these were clickable links

Adding the organisation name

The name of the organisation providing the service was added to the search results and service details pages.

We received feedback that whether a service is provided by a LA or VCFS organisation had little impact on whether practitioner would choose them. However, the name of the service provider may provide information that would make it easier for them to recognise and or trust services.

Adding a search bar

As a team we would like to explore how search may be used within our service.  We believed that a search function will allow users to search for keywords first, before filtering any results. 

This is the first time we’ve explored search so we wanted to observe if and how users would use this feature.

Round 2: objectives and insights

For the second round, we kept the same objectives from the first round. However, we also wanted to test:

• the impact of the iterations following round one
• how practitioners respond when contact details were missing from some service details, a common issue with service data held by LAs

Key insights included:

• positive feedback on overall purpose and experience of using the directory
• several practitioners noted they would still prefer their existing directories or resources, especially if they tried and failed to find relevant support through the single directory
• practitioners preferred support services they were already familiar with
• missing contact details was a key frustration, without these most practitioners would not consider a service
• as in round 1, the distinction between whether a LA or other organisation provides a service was not a consideration
• filtering continued to be the preferred way of refining results, with search used only when practitioners had specific keywords in mind

Feedback on iterations following round one

We also gathered and analysed feedback on the changes we had made to the prototype.

Category filters changed to accordion style

The accordion-style filter made identifying filters more targeted, and users scrolled less. We’ll continue testing this with parents and carers to validate its effectiveness.

Additional changes to other filters

As a result of the changes to the other filters we observed the following for:

• ‘Days service is available’ users didn’t express confusion, but we couldn’t confirm if the guidance meaningfully influenced their behaviour
• ‘Cost’ no issues were observed, suggesting this change aligns with user expectations and behaviours
• ‘Age’ users had no issues with the updated age ranges, but the overall impact remained unclear

Changes to the search results page

Insights around the changes to the result page included where we had :

• added local authority area practitioners understood that searches related to specific areas
• added the distance from postcode some confusion about how services were ordered persisted, so further adjustments may be needed
• increased size of service titles this helped some users, but those with lower digital literacy still struggled to identify clickable elements

Adding the organisation name

The organisation name didn’t significantly impact how practitioners interacted with services. We’ll need further testing to assess its usefulness, especially for public users.

Adding a search bar

Search wasn’t used as a filter replacement. While some users considered searching, most preferred filtering. We’ll potentially explore this further in future rounds.

Summary of findings after practitioner

Our research aimed to understand more about whether the single directory is valuable and usable for practitioners.

Key insights gathered across both rounds included:

• practitioners valued the concept but needed consistent, relevant service information to consider adopting it as go to resource
• filtering was preferred for narrowing results, but some categories didn’t match how practitioners naturally think about services
• missing contact details were a major barrier, reducing trust and preventing practitioners from signposting families
• practitioners prioritised the service’s ability to meet family needs, rather than the type of organisation that provides it
• some questions were raised around whether the way that users go from entering a postcode to seeing results matches user expectations

What we still need to validate

While we gathered valuable insights from these 2 rounds of research, several areas still require further validation.

Minimum data requirements

We need to determine what constitutes the minimum viable data for services to ensure users can make confident decisions. This includes clarifying whether contact details should be mandatory and understanding what other information is critical. This poses a large challenge, as lot of data held by LAs may not currently fulfil these requirements.

Understanding the impact of technical constraints

The Open Referral data standard that we are using influences what data we can include and how it’s displayed. Some practitioners expressed frustration with missing information like whether a service was online or in-person, which Open Referral currently limits. We need to explore how to balance these technical constraints with ensuring the directory remains usable and valuable.

Public user needs

To create a directory that serves both user groups, we need to expand research to include parents and carers. This will help us understand their access needs, search behaviours, and what information they require to find and trust services.

Next steps: what we’re working on in January and February 2025

In the next phase, we will:

• expand research to include public users (parents and carers) to understand their needs, behaviours, and how they search for services
• work towards defining minimum viable data requirements for services and how we can ensure this
• further explore technical feasibility to understand the impact of any restrictions in using Open Referral and how we might mitigate these
• conduct competitor analysis to understand how well existing directories meet user needs
• conduct research and testing around possible names for the single directory