Product design: when private beta isn’t the next step
We know that not all projects move from alpha to private beta. But there’s an expectation that it’s the next logical step… then to public beta, then live.
So, when we realised our project wouldn’t make it out of alpha, it felt a bit sad. Almost like something had gone wrong.
We won’t be the only team who find themselves in this situation and we wanted to share what we have learned and more importantly, how it’s actually been a positive thing.
Our alpha focused on ‘mandatory reconsiderations’. This is where a citizen is unhappy with a benefit decision and asks the Department for Work and Pensions (DWP) to look at it again. The citizen explains why they think the decision is wrong and can give evidence to support their case.
Our discovery work pointed towards this being a key area for us to improve for several reasons:
It felt like we were operating in an area that had so many opportunities for change and improvement, it was hard to know where to start.
Our discovery work showed that Personal Independence Payment (PIP) receives the highest volume of mandatory reconsiderations, so this is where we decided to focus our efforts. We set out by mapping our user journeys, starting with the PIP application, and ending with the appeals process.
Our service was in the middle. At first that felt ok as there were clearly defined ‘start’ and ‘end’ processes to an application, mandatory reconsideration and then an appeal. But what emerged very quickly was that citizens didn’t see it chunked up that way. They saw it all as their ‘PIP application’. They didn’t want to start a new process. Our user research was telling us we needed to build the mandatory reconsiderations process onto the existing service.
We also learned more about all the work going on to improve the PIP application process. Their goal was to ensure that fewer applications get to the mandatory reconsiderations stage in the first place.
The team who were developing the online PIP service were already building in excellent feedback loops to improve the service, learn from mistakes and make it easier for citizens to give DWP the right information and evidence at the start.
This work was well underway and would affect anything we knew we’d need to build to help citizens. It started to feel like we didn’t have a ‘new’ service to work on that would fix lots of problems. It felt more like we were trying to squeeze into a space that was rapidly changing size and shape. We felt that digitalising the mandatory reconsiderations service could be more of a hindrance than a help at this point.
The second Government Digital Service (GDS) service standard was screaming out to us. We couldn’t solve a whole problem for users alone. It wasn’t going to be possible to look at this ‘mandatory reconsiderations chunk’ in isolation even though it seemed logical to start with. These were problems that we were unable to solve anytime soon, and certainly not during a private beta phase.
The time wasn’t right to progress our alpha. But that’s not to say it never will be.
We learned so much about the needs of our users in terms of communications, updates, expectations of the service and much more. We packaged up all our work – user needs, thoughts, designs and plans for private beta. This means we can share our insights with teams when they start to look at this stage of their user journeys. Or we can come back and pick up this work again when the time is right.
We have also been able to recommend changes to the notifications people receive after raising a PIP mandatory reconsideration. Our work showed that these could be timelier and reassuring. A trial has now started to test these notifications and their impact.
Our alpha work highlighted that our agents are using systems that aren’t fit for purpose when dealing with mandatory reconsiderations and appeals. For appeals especially, we saw some key areas where we can improve the user experience for our agents.
We have already started to look at this and identify where we can add value. We know that anything we can do to improve processes for agents will ultimately help citizens too.
Sometimes as you go into an alpha, you can learn that your service is much, much larger than you first thought. The need to solve the whole problem for the user can show how much you need to navigate across boundaries both within and outside your organisation. This doesn’t mean it’s impossible – we know we should do the hard work to make it simpler for the user. But equally, the timing also needs to be right, or you risk developing the wrong thing. We all strive to do what’s best for our users, and sometimes, that is to wait until the full lay of the land can be seen.
We could have pushed on and built something independently that could have helped people in the short term. But longer term there’s a real possibility that we’d have created a beast that caused more problems for users and was unsustainable due to so many moving parts and dependencies.
We have learned, very quickly, that the route from alpha isn’t always private beta. That’s ok. Sometimes it’s just not the right thing to do at that time.
Originally posted here
