How to map complex customer journeys
3 ways to visualise and present your data
Mapping data on customer journeys can help organisations understand their audience better and make their journeys easier. Whether these journeys are within a website, through their decision process, or are actually physical, there are plenty of ways to represent them. When it comes to straightforward journeys, flowcharts or online tools can do the trick, for example for a simple online checkout process.
But what do we do if our customers have complex, looping and messy journeys? I came across this problem recently at Geckoboard, while trying to understand what our customers get up to on our website. It’s got a lot of pages, as well as a blog and an online help centre, so our visitors had anything but a linear journey.
And for me, it wasn’t only about extracting insights from the data, but also making it easy for non-analysts to understand and explore our customer journeys. So I experimented with some attractive and interactive ways to visualise journeys: here are a few tips if you want to try them out yourself (even without coding!).
1. Sankey diagram
Sankeys are widely used and easy to understand. The height of the bars represents the quantity, and colour can be used to distinguish between flows or themes. They’re great if we want to show journeys with a few events, or customers that all follow a specific order of steps, such as a checkout from an online marketplace.
However, they aren’t so good for journeys where we care about the full sequence of events, as we can just see the number of people who move from each particular event to another. They also might not work for journeys with loops or that have events with no set order, for example, shops visited during a trip to the mall. In such cases the diagram would get confusing.
Do it yourself
- With R: https://christophergandrud.github.io/networkD3/#sankey
- With D3: https://github.com/d3/d3-sankey
- With no coding: http://sankeymatic.com/
Why we didn’t use it
In our case, there isn’t a single ordered flow through our website. The users visit and revisit pages in lots of different orders which makes for a very messy and confusing Sankey (see below!).
2. Chord diagram
Another beautiful and interactive way to explore complex journeys are chord diagrams, which are also often used to show connections between different people or items. The size of the arcs represent how many people move from one event to another.
Chord diagrams are good for journeys with loops or with two way flows between events, such as trade flows between different countries. However, they aren’t so good for understanding the entire sequence of events, as they only display the relationship between individual events.
Do it yourself
- With R: https://github.com/mattflor/chorddiag
- With D3: https://github.com/d3/d3-chord
- With no coding: http://mkweb.bcgsc.ca/tableviewer/
Why we didn’t use it
In our case, our goal was to understand the entire journey of our users, not just how many move from one particular web page to another. Unfortunately, this detail gets lost in a chord diagram.
3. Sunburst diagram
A third pretty and interactive way to understand flows and the order of events is through a sunburst diagram. In these charts, each ring represents a different level of hierarchy, starting from the center and moving outwards. Every ring is divided in arcs of different size and colour based on the quantity and category of event.
Sunbursts are great for journeys where it’s important to show the order or hierarchy of events. They also make it easy to get a clear overview of the data, and offer the ability to dig deeper. A good use case is visualising a journey through a website. However they may not be useful if we need people to assess the exact values, as humans aren’t great at measuring arcs with their eyes.
Do it yourself
- With R: https://github.com/timelyportfolio/sunburstR
- With D3: https://bl.ocks.org/kerryrodden/7090426
- With no coding: http://app.rawgraphs.io/
Why we used it
Finally, with our sunburst diagram, we could clearly see what the major common journeys through our site were. We could also easily spot quite a few uncommon and unique journeys, revealing that lots of people are up to really different stuff on our website.
So which is the best visualisation?
Choosing the best visualisation all depends on your data and your audience. There is no “best” one! After starting with a clear idea of what you want to visualise, there are plenty of online resources to help you research your options. For example, we love the Financial Times’ Visual Vocabulary and the Data Visualisation Catalogue. And to avoid the many design pitfalls that come with data visualisation, check out our top tips on how to Play Your Charts Right.
