As we started to discuss where we begin with our Information Seeking Discovery phase, we soon realised we have loads of existing research and data, and there are lots of different areas that we could look at.
To bring some increased focus to what we’re doing – and to get a shared view of what a successful Discovery will look like – we decided to run an Inception day.
What’s an Inception day?
The idea was to bring together a bunch of people from across the NHS Choices programme – as well as our commissioners and other experts and interested parties from the Department of Health – to share ideas around the areas that our Information Seeking Discovery should explore.
To get us all pulling in the same direction, we ran through a number of exercises during the day – around Vision, Constraints, Areas of Focus, Assumptions and Hypotheses.
We began, though, with some background on why we’d chosen Information Seeking as our first main theme, or topic, for Discovery.
Jeanette, our Head of Customer Insight, ran through a summary of the existing research that was pulled together over the preceding months. This covered some of what we know about the people who use our site, and the people who don’t. One of the key points is that we know people don’t come to www.nhs.uk to browse – our users are very task-focused – they come with a specific thing in mind that they’re looking for, or trying to do.
This word cloud summarises the responses to a standard question in one of our recent customer satisfaction surveys – What exactly were you looking for today?
The word “Information” has been removed, as it would otherwise dwarf the other words, due to its prevalence.
It may seem like it’s stating the obvious, but all of this shows that people are coming to the site seeking information – about a wide variety of topics.
Drafting a Vision
We ran through an exercise to draft what the vision might be for the Information Seeking part of our service, with each group filling in the blanks in a series of statements.
There was a fair bit of consensus between the groups, with common themes around us providing a digital service, or experience, rather than a website; the service being trusted and providing reassurance, and not having any commercial bias.
We’ll keep reviewing this vision as our Discovery progresses – to see if we can refine it further as we learn more.
As the Information Seeking topic itself is still really broad – we need to zoom in on some specific areas. Although we’ll focus on these – the learning that we do in these areas during the Discovery will still be applicable to other aspects of the broader service – we’ll be building up a good knowledge-base.
Matt (Product Lead on Information Seeking) and Kenny (Analytics Lead) gave us a run-through of what our data tells us about the top site entry points and the top search terms – through both Google, and our onsite search. What are the top things that users look for on our site at the moment?
There’s a fair bit of correlation between the entry points and search terms, as you’d expect. After digesting this information, we then split into groups and each came up with our top five potential areas of focus. What we were aiming to do was come up with a representative shortlist – that covers a variety of different types of Information Seeking user journeys, with a variety of different contexts.
The examples below cover some common symptoms and conditions, including a childhood condition, which brings with it a whole different emotional context. We also have general health information in the form of BMI, and an emotive service-based subject in Local Social Care.
We’ll take the top five that each group came up with, and shortlist them further to arrive at some specific areas to focus on. Watch this space…
Assumptions and Hypotheses
As well as focusing on some specific topics, we also started to explore some specific assumptions, to focus our learning.
We all make assumptions, all of the time. By getting our assumptions around our service and our users up on a whiteboard, we can start to see a range of opinions emerge. What was great about this exercise was that no-one tried to disprove anyone else’s assumption – everything was accepted, with a view that the whole point of Discovery is to start testing these assumptions and validating them one way or another.
In our groups we came up with assumptions, and then plotted them on a whiteboard, based on our level of certainty about the assumption (the vertical axis) and how important it is for us to validate the assumption right now (the horizontal axis).
The majority of our assumptions currently lie in the top-right quadrant, which means we don’t really know whether they’re valid assumptions or not, and we probably need to figure that out pretty soon.
We can focus our learning efforts on these assumptions in the top-right, and through the Discovery phase we should see them gradually shift to the bottom-left, as we validate them either way.
We also started to discuss how we’d validate these assumptions by coming up with specific hypotheses that we can test. These hypotheses can be tested by different means – some can be tested via further quantitative research, while for others we might build prototypes to test with users.
This process of surfacing and prioritising assumptions is an earlier stage of the Lean UX approach that we adopted for the NHS Choices visual refresh we carried out over the previous couple of months.
Discovery is really all about learning, and the validation of our initial assumptions will form a big part of the learning we do during Discovery.
We had a good discussion around the constraints that we have on our Information Seeking Discovery. We listed different constraints, and then worked as a group to determine how tight that particular constraint is.
While time is pretty immovable, the scope of the Discovery is flexible. This is good, as it means we can narrow the scope by focusing in on specific areas to explore in a lot of detail. We put Technology Choice right at the bottom of the sliders, as we’re not even looking at technology yet in this Discovery, and so it should not constrain us at all. Policy is in the middle as, although it’s really important, we felt that if our research around what users need should challenge policy, then we should feel empowered to question the policy.
It was good to get shared agreement on these. These sliders will also be reviewed regularly, and should guide us in our decision-making as we progress through the Discovery.
Looking at the outputs from the day, it feels like we pulled together some good ideas and we now have a bit more focus around what we tackle next.
We need to shortlist the areas of focus, so we can really start digging into these. We’ll also be pulling together existing research, data and analytics to start validating some of our assumptions.
Getting such a variety of people together for the day was great, too. We hope to continue in this spirit of making Transformation as open as we can, with regular posts here, and Show & Tell sessions in our different office locations.