Ava Choices user feedback
and site learnings to date
(Pilots 1 & 2)

Work completed at DPH Group

My role: Junior UX Designer, working with the Head of UX until they left in May.
Skills used: User interviews, Experience map, Usability testing, 5-second test,
Surveys, Web recording analysis
Duration: April 2020 - August 2021

Overview

The first pilot of Ava launched in September 2020 as the initial part of a phased release plan. Before launch, I worked with the Head of UX to create initial designs and prototypes for testing. We often worked on our own sections independently and then would discuss as a team. We ran usability testing sessions over Zoom to refine the designs ahead of build. After the product was live, Crazy Egg was installed to record users on the site, which also provided us with heat maps and click through information. Throughout the winter and spring, new features were regularly designed and added as per business requirements. Despite positive feedback during the usability testing sessions, Ava Choices was not converting effectively enough.

In May 2021, there was a change in ownership within the company and the Head of UX also departed. I was then tasked with undertaking a rebrand and some key changes to the consumer registration form and requests section with the aim of increasing conversions at both sign-up and payment stages. This was live within 5 weeks and is referred to as pilot 2. The marketing budget for pilot 2 was significantly lower and less users found the site during this time.

My analysis of Google Analytics and Crazy Egg found a reduction from 38% to 20% in drop offs during the Requests section following the process change. However, the conversion rates did not improve. It was then decided that a review of the business model was required to showcase the services upfront and give better consumer appeal, read about Pilot 3 here.

Users and audience

Ava Choices has two user types:

Consumers - 55+ year olds in the UK who require care at home services or are looking for a care home for themselves or a loved one. They may qualify for funding from the NHS or their local council.

Providers - Care professionals who create profiles on Ava Choices to connect with potential customers in their local area.

Scope and constraints

  • The Ava Choices platform was built by an internal team of developers. A project manager was in charge of their workflow and what amends could be implemented in each release. Due to business requirements for additional features to added for a potential client, issues discovered in the system often take a long time to get into build.

  • All testers were found through social media and word of mouth as there was no budget available to incentivise users to join our testing sessions or to use a service to recruit testers more closely matched to our target audience.

  • The marketing budget for pilot 2 was significantly lower and less users found the site during this time.

InVision screen examples

Process

Pilot 1

When I joined the DPH Group in April 2020, the first iteration of Ava Choices was already underway. For the following year I worked with the Head of UX to design, prototype, test, refine and brief to build each stage of the site. We worked independently on projects and then would present them back to discuss and refine the designs ahead of testing.

We used consumer surveys, journey flow plans and wireframes to inform our initial designs, alongside interviews with care providers who would deliver the services purchased through the platform. Designs were prototyped using InVision and their effectiveness assessed through usability testing sessions. The sessions had to be run over Zoom due to the pandemic. We would take it in turns either running the session or being the passive observer and note-taker.

Any points of confusion or frustration were of particular interest and any suggestions given always noted. These would inform the majority of the amends made to the designs, ready for the next round of testing. Once designs were developed and passed the testing phase, we would brief them into the internal development team and test the final product.

Image of Pilot 1 homepage with click throughs mapped by Crazy Egg

The first phase went live in September, collated feedback from 57 usability testing sessions fond a generally positive impression of the brand, site information and process with users saying:

“It’s pretty intuitive, it’s got everything it needs without being cluttered.”

“It’s a different experience going through looking for care this way, but I like that there’s lots of information to keep you updated and that it was nice to look at and simple to use.”

During the following months, Crazy Egg was installed to allow us to observe recordings of users on the live site, as well as giving us access to heatmaps and click through data. This aided our understanding of how organic users were progressing through the site and allowed us to identify their points of confusion and frustration. We then further developed these areas to improve the user experience and briefed them back into development.

Pilot 2

In May 2021, the Head of UX left the company and I was the sole designer remaining on the project. New management had also come on board around this time and a rebrand was requested to give Ava a quality “John Lewis” feel.

The second pilot launched just 5 weeks later, with new information pages written and a simplified consumer sign-up form - amended to lead directly into the Find Care wizard instead of the dashboard - was created with the aim of increasing conversions. A limitation for this iteration was a significantly lower marketing budget to drive traffic to the site.

I collated the results from our Google Analytics and Crazy Egg recordings compare the effectiveness of the two site iterations, which showed only small differences in site performance. I also ran a workshop with the customer services team in order to take down all of the issues that users had contacted them about, which we then used to generate some initial ideas about how to improve pain points within the system. This was all noted into an Experience Map.

Outcomes and lessons

Key findings from my analysis:

  • Users are engaging with the site but are not converting, this suggests that they are not finding what they are looking for, which we can assume to be the product/service we offer aka Provider profiles and prices.

  • Change to Find Care Wizard has lead to a reduction from 38% to 20% in drop offs.

  • Between submitting a Care Request and creating a booking is causing a significant drop-off in potential sales. Calls to the Ava Assistants are often about how to arrange the assessments, so this process must be improved in the next iteration.

  • The amount of Providers visiting the site has reduced substantially. They will need re-engagement activity to bring back to the site for the next iteration to be successful in giving Consumers quotes.

In future I strongly believe that having our product offering (provider list) before the sign-up process will have the biggest impact on conversions. The current order was a business model requirement, which does not appear to appeal to consumers. I also feel that the site needs SEO optimisation to increase the CTR from Google search, I have passed this feedback along to the management team for them to assess the budget.

To increase the conversion rate of the site in its current format, there are several key points found in the report and experience map that would be my top priority. Namely the process for arranging the care assessment has been identified as a journey stopping point that needs urgent attention. Secondly, system and email alerts need to be reviewed and written by a copywriter to ensure that the consumer is informed at every step of their journey and encouraged to re-engage and progress through to booking stage.

If I were to start the project again from scratch, I would focus on creating an effective core system before expanding the service offering. The addition of features to aid one client, who had under 10 users trialling the system, significantly reduced the availability of the dev resource and prevented several key improvements from being implemented for the benefit of all consumers.

Related projects