Helping taxpayers when they’re due a refund

April 2019
Project length
1 year

I helped HMRC to better serve the taxpayer by designing a service to track VAT repayments online.

In a hurry?

This case study will take you around ten minutes to read. Here’s a summary if you don’t have the time.

I worked with a government digital team to help taxpayers better understand when they’d receive a VAT repayment. Users got a better experience and HMRC saved money. You can skip to the outcomes.

The problem

When a VAT repayment is delayed, it can spell disaster for business as predicting cash flow is vital. Understandably, people want to know when they’ll get paid.

In 2018, HMRC had no online service to allow people to track their repayments. 50,000 people called the VAT helpline asking “where’s my money?”. To further complicate things, if a user hadn’t supplied bank details, HMRC had to send a cheque. Combined, telephone support and cheques cost HMRC around £1M a year. The business objective, naturally, was to reduce these costs.


In the Discovery phase, we:

When a company submits a VAT return, it triggers a complex set of interactions between the user, employees and software. To understand this journey, we set about creating a service map.

This was a significant challenge; VAT is complex, involving many people and processes. There isn’t just one VAT team, so we spent a lot of our time trying to find out who could answer our questions.

For example, understanding why a repayment might be delayed was particularly tricky. The vetting process for repayments is a closely guarded secret. It’s done on a case-by-case basis, and fraud can be a problem. Naturally, HMRC didn’t want to give away anything that might make it easier to cheat the system.

Mapping the service took weeks. At our daily standups, we slowly filled in the process on our “wall of knowledge”.

Getting to know our users

Speaking with users, we found that for the majority, the repayments service was working well. But for around 15% of people who did have a problem, this often resulted in confusion and multiple calls. These were the users for whom we needed to improve the service.

Furthermore, even those having a good experience were confused about how they’d be paid. For example, users thought that setting up a Direct Debit meant that repayments would be sent back to the same account. This isn’t true, people have to provide a “repayment account”. Without one, HMRC would send a cheque.

Our user researcher, Kathy, created a set of user needs:


When expecting a VAT repayment, most people simply wait for the money to arrive in their account. It’s only when this goes wrong they start looking for help. People did this in an unexpected number of ways. In fact, uncovering the routes-in was quite the challenge. It was well worth the effort, though, as it revealed alternatives to building a digital service, namely:

  1. Building an API to allow third-party software vendors to provide tracking information in their MTD enabled software;

  2. Improve the automated phone assistant to give tracking information;

  3. Improving various other user touch-points, such as letters and online guidance;

  4. Extending the current Business Tax Account service to show repayments.

Pain points and opportunities

I translated all our findings into an “as-is” service blueprint. This allowed us to identify pain points and opportunities by comparing the blueprint against the user needs. We were then able to rapidly spot areas for improvement.


In the Alpha phase, we:

Going wide

We now had a clearer understanding of the existing process and the problems facing our users. It was time to “go wide” to create as many ideas as possible. I usually like to involve as many people in this part as possible. By engaging with a big group who have familiarity with different elements of the service, you often end up with unexpected solutions.

For this project, I ran an ideas workshop with stakeholders from HMRC and other designs from the digital centre. We focused on two, deceptively simple-looking questions:

This is always my favourite part of the design process: Working with people who wouldn’t call themselves creative and seeing them come up with ingenious ideas. Having other designers in a workshop is excellent, but involve others who actually work on the thing you’re trying to improve, and you’ve got magic!

Jon is hands-down the best UX designer I’ve worked with. His creativity, technical mindset, and ability to bring people together makes every project fun with guaranteed results. I’d jump at the chance to work with Jon again!
Louise Maloney, Content designer, HMRC

Buy-in from the business

In HMRC Digital, we work alongside other parts of the revenue—what we loving refer to as “the business”. There’s naturally a lot of red tape in government, and solutions can be narrowed early. So we needed to get buy-in before we did any serious testing with users. I put together some options in Sketch. The intention was to demonstrate a general direction to the business.

The simple tracker

In some ways, this solution would have made a lot of sense. We proposed that we allowed users to track repayments directly as part of their VAT account dashboard.

The integrated tracker was never going to fly as it meant we’d have to hand the work over to another team. When cost centres have more sway than user need all you can do is roll with the punches, but it’s a shame when business structure imposes itself on the UX.

The third-party tracker

Thanks to the Making Tax Digital programme, VAT payers were now submitting tax returns from MTD compatible software. We thought it probable that people would want to see any repayments progress in the same place, too. We thought providing an API would be an ideal addition to any service hosted on GOV.UK. The business agreed but proposed funding the API work as part of a possible second phase.

The deluxe tracker

The deluxe tracker was created to show complete information about repayments. By providing a full history of statuses with an explanation of what they meant, we could hit every one of the user needs. Thankfully, the business agreed and green-lit further work on this solution.

Moving closer to a solution

Now we had a direction agreed, we could start testing and improving the deluxe tracker.


One of the things we explored was the idea of alerts: where a user could sign-up for a notification when their repayment moved from one stage to another.

I mapped out how this could work. Based on the states we knew a repayment could move through, I knew that some users would receive some alerts in quick succession, then non at all for weeks. We decided to concentrate our effort elsewhere and revisit alerts later. This turned out to be a wise decision; during testing, we saw there wasn’t really any user need for alerts. People generally prefer to log in and check up rather than be pushed to by a message. Even those that said this might be useful admitted they’d ignore an email. Text messages were a better option, but expensive to develop for not much benefit.

When will I get my repayment?

One thing that kept coming up in testing was the question “when will I get my repayment?” We felt that without showing dates, we wouldn’t put a dent in the volume of calls to the helpline.

The problem was that we couldn’t just pull a date out of an API as no such data existed. We did, however, know when the user had submitted their return and how long HMRC allowed for repayments; at 30 days HMRC would have to pay a “repayment supplement” if the delay was their fault. So we tested a version with an approximate date. This worked great!

If repayments were delayed beyond 30 days, we added a new timeline item to the history acknowledging this and providing guidance on what the user could try next.

This also prompted us to redesign the tracker history page. Making the amount and estimate date front-and-centre worked well during testing.