How Heap used Appcues to launch our Appcues integration
“Hockey stick growth” is a term product managers use to describe a specific growth pattern: one that stays linear until a sudden spike sends its trajectory upwards. When product managers release a new product or feature, they usually shoot for hockey stick growth in user adoption — linear growth in adoption in beta, and then a sudden spike in adoption rates after general availability.
One way your product team can help drive hockey stick growth for a product release is by defining — and then fine-tuning — what a good launch process looks like. In January, the product and product marketing teams at Heap got together to do just that. Then, we folded the changes we made to our process into launching our recent integration with Appcues.
Small changes, big impact
When our two teams met to talk through our general launch process, we had a clear agenda: we wanted to understand what had been working, what hadn’t been working as well as we wanted, and what adjustments we could make to improve our release process and boost adoption rates.
Looking at recent launches, we realized one thing: making a few small changes often had an astounding impact on our feature releases. On our most recent launch, we expanded our in-app communications and refined parts of our in-app messaging — and we saw feature usage skyrocket in the week after launch. And adoption continued to grow, doubling during launch week and more than tripling since.
A few weeks later, we put the process to the test again when we launched our integration with Appcues. This launch was even more exciting because we were not only launching a new integration — we were able to use its new feature in our launch process. We believe that product teams can learn so much from dogfooding their own product, so we couldn’t wait to be our own end users.
Using Appcues to launch Appcues
Before we dig into the specifics of our launch plan, let’s take a look at Heap, Appcues, and our integration. Heap is a product analytics platform that enables product teams to extract customer insights, and Appcues is an experience platform that product teams use to create in-product experiences, such as product walk-throughs, surveys, and announcements.
Our new integration allows teams to personalize in-product experiences in Appcues based on behavioral data and segmentation taken from Heap. As it turns out, the integration is not only useful for customers — it was a game-changer for our launch process!
Here’s how: before launch, we decided to use our Appcues integration to personalize our launch messaging to customers. (Yes, we were using Appcues to personalize our messaging about launching Appcues.) During launch prep, we identified three segments of customers and defined launch goals for each:
For users who participated in this feature’s beta program, we wanted to offer a feature walk-through to refresh their memory of the integration and encourage ongoing use.
For users who use Appcues but haven’t set up the integration, we wanted to guide them through the setup process before we walked them through the feature.
For users who don’t use Appcues, we want to share why Appcues is a great addition to their toolkit.
In segmenting our audience this way, we also defined what a successful launch looks like: addressing the distinct needs of these three groups, creating personalized in-app experiences for each, and then targeting them accordingly. Audiences were defined in Heap as Heap segments and then synced to Appcues to target the announcements.
Personalized messaging for target audiences
We took a different personalization and targeting approach for each user group. Let’s take a look at what we did for each segment below.
Segment 1: Users who participated in this feature’s beta program
Personalization: The announcement copy we used recognized that the user has enabled the integration and offered a feature tour. It was extremely easy for us to do this in Appcues, and to use the Appcues builder to create a personalized experience that looks native to our app, without having to rely on engineering or design.
Targeting. Constructing the segment was easy with Heap. We defined a Heap event that represents using the beta feature and created a segment of “users who have done the event at any time.” The segment was synced to Appcues using the new integration.
Segment 2: Users who use Appcues but haven’t set up the integration
Personalization: The announcement copy we used recognizes that the user hasn’t set up the integration and offers to take them through setup.
Targeting. We received a mutual customer list from our partners at Appcues and then removed users who participated in the beta (i.e. users in segment 1 above).
Segment 3: Users who don’t use Appcues
Personalization. The announcement we used here is mostly informational and links to our launch blog post to share common use cases and benefits of the integration.
Targeting. This experience was shown to everyone except users in the other two segments.
Results and next steps
Our initial results have been pretty positive for all three segments, and we were able to hit our launch adoption goal for this quarter!
The biggest surprise was the engagement level we saw from users who don’t use Appcues (segment 3). We’ve received more interest from these customers than anticipated, which gives us an opportunity to share more information about the power of personalized in-app engagement.
The integration allowed us to leverage a user’s past behaviors to target the user in a more personalized way. More specifically, we used it to create three different segments of users (beta testers, customers who have Appcues, and customers who don’t) and build out differentiated announcement experiences based on which segment the user is in when he or she visits our product. Before this integration, we used generic announcements to launch new features, which often resulted in lower feature engagement.
Of course, launching a feature is just the start. The most exciting part is learning from the launch and evolving our product strategy accordingly. For me, this means moving on to analyzing usage data, getting qualitative feedback, and developing an after-action report.