Identifying @risk users in lifecycle marketing
CloudBees is the leader in the Continuous Integration (Jenkins) and Continuous Cloud Delivery (DEV@cloud) world. It was inevitable that we would start examining our lifecycle marketing practice through the continuous-improvement lens.
A while ago, we started grappling with the issue of understanding how developers used our Platform as a Service (PaaS) and how we could identify at-risk users. This blog describes the mechanism underway to solve this issue in an initiative called Continuous Engagement (what else :-)?).
Cloud application providers can micro-segment users based on actual usage of software. Examples of micro-segmentation are: users who deployed a database, users who deployed a database greater than 5GB in size and so on. Such segmentation is feasible, as cloud providers run applications in-house and have visibility into every nook and cranny of the system. Running software in the cloud is a tremendous advantage over the previous generation of packaged software where, even though the software could be instrumented, in-house adopters would be wary of sending any data outside the firewall.
We started by instrumenting our PaaS to emit interesting events that are logged into a third-party Software as a Service (SaaS) provider called Totango (Totango helps analyze the usage of software).
Since we are in the application lifecycle business, the events of interest are things like:
Login into our web console
Deploy an application
Jenkins job started
… you get the idea.
We can then start doing interesting things with this data.
We divided our funnel into 3 segments, each with multiple points in the adoption lifecycle. It is perhaps better to show this visually than to talk about it. The accompanied image shows a simplified view of our funnel.
The top end of our funnel are the “onboarders,” essentially users who signed up with us in the last 4 weeks and are using the platform free (CloudBees has a free tier). Moving down the funnel, we have the “established” sub-segment, free users make it here if they have a met a threshold of usage on the platform. Now this is where Totango shines as we can use the actual usage of services on the platform to move people from onboarders->established and down the established funnel. The paying list is a list of users.
In traditional inbound marketing, you would walk these people down the onboarders funnel through marketing campaigns on services like Marketo (in fact, we still have a few campaigns like those, but I suspect they will bite the dust compared to the “continuous engagement” juggernaut). However, with those models you cannot move people up and down the funnel based on actual usage. Here, you can be an established-yellow this month and back to established-green next month. The last bit is that at each of the checkpoint locations within a particular funnel (say “established-lost”), we can send them personalized emails that are based on the actual usage or have a customer success manager reach out to the users.
I will blog about specific use-cases over the next few months, as we continuously improve our continuous engagement process (for example: how do you set up weekly usage reports in a cloud-based environment) and build on this image (we are actually farther along than what I have shown here).
PS: I am giving a short use case presentation at the Totango Summit this week about this use case.
Senior Director of Product Development, CloudBees
Harpreet has 14 years of experience in the software industry. Prior to CloudBees,