This year, we did something that we’ve been wanting to do for a few months. We got members from our marketing, sales, and engineering teams together and decided to completely overhaul our customer newsletter using our own Lytics customer data platform. We learned a lot during the process:
- We learned what it was like to be a Lytics customer. It felt pretty good.
- We learned what it was like to work with multiple stakeholders.
- We learned that you can start delivering awesome, personalized content in a matter of a few weeks instead of multiple months.
Before we revamped our newsletter, it looked like, well, it looked like any old newsletter. It had good content that we thought people would be or should be interested in, but which people and how interested was sort of a guess on our part. That probably describes every newsletter you receive.
But we wanted something better and we knew we had the perfect tool to do it already. We wanted to be able to send out a dynamic newsletter that could be customized based on the recipient’s interest. Specifically, we wanted to use a recommendation engine to align the content in our newsletter so that readers would be more interested in reading it. Because, let’s face it, most people aren’t waiting with bated breath to read the next newsletter that comes into their inbox.
In the interest of full disclosure, not everyone was as excited about redoing our newsletter as I was. Some members of our new, cross-functional, newsletter-overhaul team worried that the redo and adding personalized content effectively would take too long. On top of that, we had the COVID crisis driving our sense of urgency to communicate more effectively with our customers. The time to value and data issues are a pretty common concern we hear from other customers too.
You may remember Drew’s blog post about having the right data versus all the data, “Not All Data Are Created Equal.” You don’t need to have completed an entire IT data project creating a full 360-degree view of the customer to get start getting insights from Lytics. In fact, it can be done in about 7 days. It turns out that the data we collected based on which blogs our customers read, the content they viewed in the knowledge center, the articles they read on Learn.lytics, and the webpages they visited was plenty robust enough for us to kick off the new newsletter with personalized content. And when a new contact is added to our distribution list, we can still provide more accurate content recommendations than before. Lytics starts collecting behavioral data out of the gate and creates aggregate identities for users until more data is available. So, when you encounter a new reader, you can recommend content right from the start based on what the average reader is interested in.
We also had an ace in the hole: Kyle. Kyle Rothenbaum is one of our software engineers who walked us through the process of setting up our data profiles and KPIs (since you’ll want to be able to measure success early on), making sure we had the right behavioral data and linking everything to the other marketing tools we were using, like Iterable. He also advised us on KPIs. Kyle is part of an entire team of employees and partners who are available to help Lytics customers.
We mapped out our goals in phases but the process went so smoothly, we were able to launch at Phase 2, immediately. While measuring long-term results will take a little while, the short-term results have been very encouraging. Anecdotally, when comparing my newsletter content with my colleagues in marketing, there were differences in what we were offered, and as we moved out to other departments, those personalized suggestions became more apparent. Kyle and I, for example, didn’t have a single piece of newsletter content that was the same. Big surprise, right?
With Lytics, we’re able to get closer than ever before to the goal of delivering a unique newsletter to every one of our customers. Compare the next newsletter you get from Lytics with your colleagues and see what differences there are. It was a fun exercise to show the importance and differentiation of the Lytics Content Recommendation Engine.