Optimising towards your Customer Lifetime Value
By Mira Javora
“The State Machine is a profound change in how we measure and look at each traveller.”
When it comes to measuring growth, the old-fashioned metrics of unique sessions or the size of your email lists can no longer cut it. Users interact with your site and your online presence in multiple ways, in various locations, using increasingly more devices. Building a single, unified picture of user-interaction and relationship with your brand is crucial. At Skyscanner, that job belongs to the system we call the State Machine.
Travellers’ relationship with Skyscanner — their state
Every traveller uses our product differently. They may be high frequency travellers who know our product inside out, they may be explorers looking for inspiration or they may have just visited the site for the first time.
At Skyscanner we analyse how travellers are interacting with us across all the platforms we provide. Their activity combined with time-based rules defines their relationship to Skyscanner — their user state.
As the user interacts with the product, they may move from one state to another. Similarly, if they are inactive for a period of time, they may decay to less active states. The diagram above describes a high level overview of a traveller’s interaction with Skyscanner. We define more granular states to measure specific key points in a traveller life-cycle. For example: a searching user can be searching for something specific or just exploring options, depending on their intent.
Like many other internet economy companies, we form an interaction timeline for each user; a record of the complete interaction history between a traveller and Skyscanner.
Defining value of each state
Some user interactions generate revenue and others have associated costs. Comparing these allows us to look at the cost we spend to acquire and retain them over time. It allows us to calculate the customer lifetime value (CLV) — the lifetime revenue minus the lifetime cost.
Furthermore, we can apply predictive analytics of behaviour of each group and based on the data we know and predict the CLV of each user over the next year.
UMV -> AAR-> CLV
Knowing the future CLV of each user changes the game completely. It allows you to optimise your acquisition, activation and retention efforts onto moving people to the most valuable state while at the same time giving you the ability to justify spend and predict revenue. Historically we have moved away from UMVs towards measuring Acquisition, Activation and Retention (AAR). The State Machine allows us to move beyond AAR and optimise each campaign by CLV.
Using the data to drive personalisation and product
This unique data-set is also used to deliver personalisation and drive product features. The user activity timelines allow us to change messaging based users’ past behaviour in order to improve user experience. We can also create audiences from user segments in various states to build lookalike audiences or provide improved targeting and personalisation to our email and push notification communications. Finally, the data can drive on-site features and products (recent searches and deals) and could ultimately be surfaced to be fully controlled by each traveller.
The State Machine is a profound change in how we measure and look at each traveller. Through the unique data-set we can deliver truly personalised experience and deliver more value to our travellers.
We want to share more
For more tips, Growth Hacks and job roles from across our global offices, sign up for Skyscanner Growth hacks right to your inbox!
About the author
Hi, my name is Mira and I am a Technical Manager at Skyscanner, working in the Growth Tribe. I’m a squad lead on the State Machine and I was heavily involved in building out user services at Skyscanner such as authentication, price alerts, recent searches and automated email and push. I’d encourage you to take a look at the current Skyscanner Product, Engineering and Marketing roles we have in our Growth Tribes across our global offices.