Making our wayfinding app less accurate to improve the user experience

Back in the summer of 2015, our Shaun in the City app reached the top of the UK iOS and Android app store. Shaun in the City was an art trail in London and Bristol, UK, featuring 120 sculptures of Shaun the Sheep painted by various different artists. Getting to the app to #1 was very cool for a couple of reasons:

  • It meant we were raising lots of money for charity (all proceeds went direct to the Grand Appeal charity)
  • We got to number freaking 1!

Rewind a few months and we were in the final stages of testing the companion app for the trail. Our idea was for a combination guide, map and collecting game, where you could cross off all the sculptures you found and get achievements for doing certain things (get 5 sculptures in one day, collect all sculptures in one day, walk backwards up to a sculpture – that kind of thing).

We were pretty happy with our internal testing of the design and mechanics, and recruited eight teams of parents and kids to put it through its paces in the real world.

Here you can see three iterations of the app, two before proper user testing, and one after.


Holy testing batman!

Notice anything different? There was one minor change to our app which made a MASSIVE difference to the enjoyment people had while using it, and that was reducing the accuracy of a nearest sculpture pointer, to vastly improve the user experience.

What is the point?

sitc_poitnerOn our first iteration of the nearest sculpture pointer, we went with a classic arrow to show the way.

We didn’t put a huge amount of thought into it first time round: surely people would want to know exactly where to go to the next sculpture? There was nothing in our user testing script about checking it specifically.

It quickly became apparent, though, that all was not well with our original thoughts about the pointer. For a start, because GPS is generally a bit flaky and influenced by want kind of device you are on,  how much water you are near, how much metal you are near, your current GPS signal and many different factors, people standing next to each other were being told to go in different directions. Occasionally the pointer would wig out and just spin around.

Also, because we weren’t doing much smoothing on the readings from the GPS, the arrow would be constantly changing direction slightly. This became the main focus of the people using the app – they were using it like a compass, and constantly paying attention to the little wibbly arrow on their device. We wanted people looking up at the world around them and exploring new areas of London and Bristol while enjoying SITC, and also knew that if they had their screens on constantly it was going to burn through device batteries very quickly.

If you look at Google Maps, the might of their algorithms don’t try and tell you EXACTLY where you are pointing, and I would imagine the team at Google working on Maps was slightly larger than the three man team at Aardman trying to tackle the same issue.

This was echoed in every test we did – clearly we had a problem.


So,  the team: designer Gav Strange, producer George Rowe, and developer Chris Underwood sat down to discuss how we could tackle this issue. We still really wanted people to be able to use the app to point to either the nearest sculpture, or the next one on the specific trail they were on.

We realised that it was us trying to be too accurate that was the issue. Accuracy was causing us three main problems:

  • The implementation of GPS across the vast iOS and Android device ecosystems was never going to be comparable, would be influenced by factors in the real world, and was never going to be perfect
  • Because a pointing arrow necessarily claims to be super accurate  as it points to one single spot, we were essentially promising the exact route to the next sculpture
  • People kept checking it to make sure they were on course, and they relied more on this than the map (using the Nokia Maps API, which was the clearest of all the options available to us in Unity) also bundled in the app

We decided we needed to make our navigation system more fuzzy – something that gave the user enough of a direction to give them a clue towards the next sculpture, to set them off in the right direction, but not enough so they would constantly check and follow it. We wanted them, once the app gave them the general idea and the amount of distance, to hunt for the sculpture themselves and explore the world around them.

sitc_pointer_2After some furious sketching and discussion (now sadly lost in the annals of time), we settled on the radar esque ping which you can see on the right here.

The ping animated towards the direction of the next sculpture, but importantly it was only broken into eight general directions. This meant that the direction wasn’t constantly updating, so people weren’t constantly staring at the radar to guide their way.

Also, it wasn’t specific enough to be relied upon to guide the exact way, so people would need to use other means (maps, word of mouth, their own eyes) to find the way.

This felt like a much safer and more fun experience, and in our next round of testing we didn’t experience the same problems as with the first iteration.

After a while, and very importantly for Aardman, we realised it gave us a great pun for its name (which sadly doesn’t feature in the app anywhere)…the RAM RADAR.

We also added some real Shaun the Sheep baas that increased in frequency as you got nearer the target, for that real Ram Radar feel 🙂

Fuzzy logic

We certainly didn’t expect that we would need to make the app less accurate when we first started off building it, but that proved to be the best thing for the Shaun in the City app. We hope it helped the many people who enjoyed SITC to find their way to more sculptures, without getting frustrated.

In the end, the app had…

  • 80,000 downloads across iOS and Android
  • Reached #1 UK iOS and Android app store
  • Featured in Google’s 50 Best Apps of the year
  • Total sculptures ticked off in the app: 1,583,821
  • Total trails (structured walks around a subset of sculptures) completed: 35,674
  • Number of people who ticked off all 120 sculptures in the app: 3,417

George Rowe is Senior Digital Producer in the interactive department of Aardman Animations

Add clips to the playlist by clicking the stars