cf6e74f923461f259f8545a45b947aa7b1c66503

5 Ways to Evaluate if your Lifecycle Model Actually Works

Blog Post created by cf6e74f923461f259f8545a45b947aa7b1c66503 on Oct 6, 2015

After building a revenue model in Marketo and letting it run for a few months, you will want to make sure it is working properly so that you can get the reporting that you want. Here are five fairly straightforward, if sometimes time-consuming, things you can do to validate your lifecycle.

  1. Create a segmentation to track which records are in which segments. Each segment should contain one revenue stage with the filter Revenue Stage. Any record that's in the Default segment is most likely misplaced (unless it belongs to an exclusion list). You can then use smart lists to further dig down into the Default segment and figure out why people aren't in the lifecycle when they should be. You can also use the segmentations to get a quick count of people in each stage and dig into any stages where you are seeing much higher or lower numbers than expected. (Yes, technically you can do all of this with a bunch of smart lists if you don't want to waste a segmentation temporarily on this, but I think the segmentation is nice because you can see everything all at once.)

  2. Check that your Lifecycle Status field matches the Revenue Stage that a lead is in. I do this with smart lists, once for each stage, to find people in a particular stage that have a different Lifecycle Status value. Any time I see a mismatch, I know there's a problem that I need to dig into.
  3. Look at a Lead Flow report in RCE. Use the Model Performance Analysis (Leads) report with Rows Stage and To Stage, Column Month, and Measure Flow. You want to be watching for any leads that seem to be moving backwards in the lifecycle instead of forward, which is a common problem with poorly built models. For example, if Known is your first stage, then you would expect to see leads moving from Known to any stage forward. If you see leads moving from MQL to Known, then your model must be adjusted.
  4. Look at the results log of each smart campaign that moves leads around in the lifecycle. Look for counts that are much higher or lower than expected (e.g. Really, we had 1000 leads go through the MQL campaign this month but we only had 50 leads actually become MQL?) Skim for things that look out of order, particularly Skipped Do Nothing messages and anything where you see a Revenue Stage Change that you do not expect. View the activity logs of some sample records to see if there’s anything unusual happening that might be causing those messages.

  5. Pick one stage and check the people in that smart campaign or stage against select criteria in a separate smart list. (You can also do this with a particular date if you prefer. For example, just randomly select September 18 and go and look at each lead that moved in or out of any stage on that date to validate. This is my version of spot-checking the lifecycle.) For example, you might look at all the MQLs to ensure that they have actually hit your score threshold or all of the people in Lost to see if any of them have open or won opportunities associated to them. You can also do the reverse - look for people with lost opportunities who are not in the Lost stage.

Outcomes