User journey maps are created during the initial phases of projects and rely heavily on qualitative and anecdotal data. They provide information on behaviors, motivations, and feelings that occur during touchpoints while completing a task (e.g. purchasing movie tickets). While these journey maps are incredibly directional to start, they often don’t evolve over time.

What if we were able to combine journey maps with quantitative data (once the project is in production) to understand: Were our initial thoughts about the user journey correct? If not, what has changed? How do these insights influence further research and strategy?

The User Situation

Let’s walk through a hypothetical example. You and a group of 10 friends have decided to rent a lake house through Airbnb for Memorial Day Weekend. Work has been exhausting and you all need some sunshine to recharge. Your group of friends has also decided to put you in charge of the planning process, which includes (wait for it) the payment.

Just from a ballpark perspective, say you are planning to go for 3 days, and the rental is around $500 a night, that totals to about $1500 dollars.

You recognize that your organizational skills are far superior to the others and that they are going to pay you back, but that financial burden is a lot for one person to take on. (Inner dialogue: Why did I do such an amazing job planning Jeff’s dog’s first birthday party last month, now I’m going to be picked EVERY time.)

Feeling stressed? Us too. Feeling empathetic? Many of us have been in this exact situation before. The next step here is to use this information to create a user journey map. Taking what we know about the hypothetical (or not so hypothetical) situation above, it might look something like this:

Journey Flow 1 (1)

It seems understandable that you would want to receive the payments from your friends before you book the place both for financial reasons but also for accountability reasons (If Jeremy doesn’t pay me before I book, who knows when he will actually pay me back…talk about unreliable).

At this point, let’s remember what we said at the beginning about user journey maps. They are incredibly directional to start, but often don’t evolve over time. Instead of creating a stagnant journey map, let’s continue this example to explain what happens when iteration and evolution is involved.

The Airbnb Employee Situation

To continue this hypothetical example, let’s say that over at Airbnb HQ, their Insights team is picking up on this behavior. They’ve been noticing that during holiday booking seasons, 18-25 year old users take 3x the amount of time to complete the booking process and have a 50% higher abandonment rate than average*.

From a quantitative perspective, the Insights team identifies a couple additional things:

  1. This user group starts browsing through options over 2 weeks ahead of time*.
  2. The majority of listings this user group is looking at are for groups of 6+*.

From here, the Insights team turns to the Research team. The Research team looks at the situation from a more qualitative perspective: Why is this behavior happening? What else is going on?

Through user interviews and sentiment analysis, the Research team identifies sole financial burden as a large variable in influencing behavior. The pressure for one user to make the payment is a tension point that could lead to a longer booking process or even lead to abandoning the plan altogether.

How can Airbnb work to develop a solution that takes away the sole financial burden and speeds up the booking process for younger users?

*The data is made up for the purposes of this blog post.

The Solution

Towards the end of 2017 and the beginning of 2018, Airbnb launched two features to alleviate financial stress: partial payments and split payments. Partial payments allow users to pay a 50% deposit up front and the remainder of the balance at a later date. Split payments allow for booking payments to be split between up to 16 users with 72 hours for everyone to complete their portion. The reservation will be locked for the 72-hour period, but if the payment is not completed within that time, the reservation will be lost.

These new features, as you can imagine, shift the user journey. Instead of so much time spent gathering payments from friends to secure the lake house for Memorial Day Weekend, it might look like this:

Journey Flow 2 (1)

Conclusion

Creating the original user journey not only put Airbnb into a user-centered mindset, but set the stage for questions to answer over time. That, in turn, could influences the way they set up tracking for an analytics platform or questions they ask during user interviews.

Now, this is just one example of how data can influence a user journey. At WillowTree, we’ve been able to put this into practice. For Regal, we noticed a pain point in selecting seats during the ticket purchase process, especially for larger groups. To alleviate this stress on the user, our developers created a machine learning algorithm that identified the best block of seats for the user, based on how many tickets they wanted to purchase.

Not only did this feature solve a real problem, but also it helped to create positive sentiment and loyalty around the Regal brand.

The bottom line? Don’t let your user journey maps collect dust. Use data, quantitative and qualitative, to evolve them as your users evolve. One tiny insight can inform design, development, and strategic decisions to delight users and improve their experiences with your brand.