Boomtrain University
BME Guide

Cohort Monitoring in Boomtrain Marketing Engine

BME’s cohort monitoring dashboard gives marketers a great tool to manage the health of their email marketing programs in a proactive way.

Screen Shot 2016-04-21 at 4.07.49 PM

This tool lets you see trends in your subscriber base, which can give incredibly valuable insight into how to optimize your email programs. Every marketer has unique challenges to deal with, but today we will break those down into 4 categories:

Sign-up Rate

How many subscribers have you added in the last 30 days? How about the past 90? It’s important to monitor your add-rate in context. Break this down over several key time frames, by newsletter, and by sign-up source to give yourself actionable insights that can be used to drive your rates even higher.

New Subscriber Retention

Simply implementing a welcome series isn’t enough to ensure that your new subscribers convert to engaged subscribers at the highest possible rates – the onboarding process should be monitored and optimized. Monitor the performance of your onboarding process by looking at the users who signed up last week/month/quarter/year and were engaged with you this month. If new subscriber attrition is a problem area for you, split this out by newsletter or signup source to see if you can get some hints at the root of the issue.

Also, if you’re using double opt-in to ensure your list quality, make sure you keep an eye on the conversion rates in that process.

Engagement Rates

For your subscribers who make it through onboarding, we can use more traditional engagement metrics to monitor health. Think in terms of monthly active openers/clickers/viewers for out of the box metrics. If you have other high-value end goals in mind, you should track those as well – monthly active purchasers/social sharers/commenters/etc.

Deadwood Rates

This is pretty much the opposite of engagement – this is disengagement – and because your list is always growing, it should be measured separately from engagement. For monitoring the population of users that has become disengaged, think in terms of subscribers who haven’t opened, clicked, or viewed a page in the past 30/90/180 days. Again, if your real objective is some other high-value action, you should monitor the cohort of users that hasn’t made a purchase or shared an article in the same time frames.

A pro-tip on monitoring disengaged users: if you are running re-engagement campaigns (for disengaged readers or shopping cart abandons, for example), your monitoring time frames should take that into account. For example, if you trigger a re-engagement email after a user hasn’t been to your site for 30 days, then you should set your initial monitoring time frame to 35 days to give those users time to re-engage and fall out of your deadwood metrics.


How to set up cohort monitoring:

In order to monitor a cohort, you first must create a dynamic segment with the definition of your cohort. Here is an example of the filter definition for a “Monthly Active Openers” segment:

Screen Shot 2016-04-21 at 4.18.11 PM


Screen Shot 2016-04-21 at 4.18.24 PM

Ensure that you save this segment and click the “keep this segment updated” checkbox:

Screen Shot 2016-04-21 at 4.18.55 PM

Now that your segment is saved, navigate back to the “Dashboard” tab and click the “+” icon to add a new widget in the “Users at a Glance” tool.

Screen Shot 2016-04-21 at 4.06.28 PM

Select your segment and submit, and a new widget will be created

Screen Shot 2016-04-21 at 4.06.51 PM

Repeat this process for all of your desired cohorts. Here are a few example filter “recipes” for creating segments that you may want to monitor:

Total double opt-in verifications last month:

Screen Shot 2016-04-21 at 4.28.40 PM

Signed up last month, opened(or clicked, viewed, purchased) this month (or quarter, half, year):

Screen Shot 2016-04-21 at 4.32.21 PM


Screen Shot 2016-04-21 at 4.32.40 PM

Existing users that signed up for a new newsletter:

Screen Shot 2016-04-21 at 4.32.53 PM

Note that “last_startedmembership” uses a special user property type that can be used to create more sophisticated cohorts.  This one is set up with this configuration:

Screen Shot 2016-04-21 at 4.07.23 PM

Share This