Kew Gardens

Improving the user experience for What’s On and Festivals covering Kew Gardens and Wakehurst.

Product Design

Sector
Research, B2B

Service
Product Design

Team

Product Design & Research Lead: Myself

Product Manager: 1

Content Manager: 1 

External Development Agency

Target audience
2.45 million annual visitors

The opportunity

To create a new page template for festivals and make improvements to the What’s On page to enhance the user experience. Solutions must be WCAG 2.0 AA-AAA compliant and documented within the Kew Design System.

What’s on desk research

Gathering data on how users interact with the what’s on pages for Kew and Wakehurst through Google analytics, click maps and heatmaps.

  1. Device type (Google analytics):
    Mobile is the prominent device type for all What's On pages, with 58-77% of usage. Desktop usage sits between 18-36% Tablet usage is minimal. Therefore, all our designs are mobile first.

  2. Accessibility filters (Google analytics and usability testing):
    Accessibility filter rarely used on What’s On filter, this seemed unusual. Through usability testing, when the user clicks / taps accessibility, 0 results always appear. Further insights needed.

  3. Carousel component (HotJar screen recording and click map analysis):
    Customers were getting stuck in carousels, how can this adapted to improve search

  4. Kew & Wakehurst filter use (Click & heatmaps):
    3.4% Avg. ‘Search event by type’ filter has low engagement Kew & Wakehurst, mobile & desktop
    <1% Avg. ‘Visit Wild at Wakehurst’ module has low engagement Kew, mobile & desktop
    60% Avg. Users scrolling to view all content on What’s On Page Wakehurst, mobile & desktop
    35% Avg. Users scrolling to view all content on What’s On Page Kew, mobile & desktop

Google analytics filter interaction

Date and event type

Google Analytic findings also show that date filters are used significantly more. The 12 event type filters only account for 9% of all filter engagement. (within the 9% include, family friendly, exhibitions, price included with ticket)

Event type

There is disproportionate usage of the different event type filters, with family friendly and exhibitions accounting for 36% of engagement.

User survey

Based on desk research, I ran a user survey to with a focus on motivation, planning period, search method, priorities, find-ability and quality of information. The survey produced the following insights:

Planning period

The majority of users are planning their visit in advance. This weekend with the least count. This user data informed the 'when' search chips.

Priorities

Only 7.8% of respondents said they have specific access needs but 60% scored accessibility as very important to somewhat important. We should prioritise date for events listing filters and cards. A ‘family friendly’ filter would support visitors either with children. A price filter and pricing information needs to be prominent. Covid and accessibility information should be prominent for those who want it but not distract for those who don’t.

Solution

I restructured the category tags based on user and business requirements. Once the iteration was complete I tested the wireframes with the research user group and found a few takeaways:

  1. Usability: The filter chip UI filter component had more engagement than a search bar. This was a great find before going into hifi as search is a high cost feature.

  2. Customer need: Having filter categories that were relevant to the user more visible on mobile, test example ‘family friendly’ got positive feedback.
    “The page made it clear where I could find activities which were family friendly & this to me, meant with my kids. Also the events are likely categorized into these subsections so I won’t have to spend loads of time looking for specific things to do which are family friendly”

From first click testing, I shared the results with the development team to kick off collaboration and feasibility of the change to the design. Key areas of feedback from development:

  • Adding date pickers into the UI filter chips added a lot of complexity for both development and the content team. Managing the calendar to ensure the date picker doesn’t go too far into the future to avoid the 0 results experience. 

  • Balancing this feedback with user need, content volume and development build, we collaboratively removed the date picker. This approach meant the user can still select research approved date periods via chips.

Learnings

Accessibility: We removed the ‘accessible’ filter within the search due to 0 results in data tagging. However, within survey feedback, customers added a wide range of access needs, validating the need to improve search by access.

Global search: Within the survey, an equal amount of customers would go to search or what’s on to find listings, due to search improvements being out of scope, this is an area for iteration to improve the user experience.

Design system growth

Explore more work