Increasing the impact of email for KLM

Impressive improvements in open rate and click through rate by combining hard core data science with qualitative research.

KLM sends millions of emails per week to its Dutch customers. These large numbers provide a huge amount of data, so this was the perfect opportunity to discover how to make these emails even more relevant and impactful.

There are several factors we can influence in email: frequency, subject line, content. In this pilot we wanted to find out what the influence of timing is. What changes when we personalize the day and time of sending the email? By combining data science with qualitative research we achieved some surprising results!

Approach

There’s plenty of data available; KLM knows who opens an email at what time and clicks to proceed to the website (open- and click through rate/CTR) so we started to collect and connect these data sources and customer data. This connection provided a rich data set with lots of insights in the attributes and behavior of individual customers. Yet this data is hardly used by companies to personalize timing, usually an optimal timing is chosen for a group as a whole.

We established 16 time frames throughout the week and predicted individually which time frame would result in the highest open rates and click through rates. We interviewed real KLM customers as extra input for the model and this way we could connect personas to the data-driven clusters.

Result

In a campaign we sent 450,000 emails at the standard moments as a reference. The other 450,000 emails were sent in one of the 16 predicted time frames. And the results were striking!  The open rate of the time-personalized mail was 2.5% higher and the click through rate 6%. In the second pilot we even reached an 8% increase in click through rate.

Bernard Geersing, Director Direct Sales & Digital Marketing at KLM Benelux sees lots of potential for the future. “Just think about the opportunities and results when we’re scaling up and personalize more elements like content. This way we keep creating more relevance for our customers by offering them the right message at the right time. Plenty of reason to keep testing and experimenting!”