Leveraging Machine Learning for Churn / Upsell Opportunities

  • 2
  • Question
  • Updated 2 months ago
  • In Progress
Hi all, I watched a very interesting webinar that was created two years ago, the gist of which was that if you had historical data for a large enough group of churned clients, you could pass that information to Gainsight and they would leverage machine learning to build an accurate health model that predicts churn (or potentially upsell opportunities if you could pass that data). Just wondering has any reaped the benefits of this? Really exciting capability...
Photo of Faelan Herriott

Faelan Herriott

  • 452 Points 250 badge 2x thumb

Posted 3 months ago

  • 2
Photo of Oscar Nembaware

Oscar Nembaware

  • 76 Points 75 badge 2x thumb
Photo of Johnny Wong

Johnny Wong

  • 134 Points 100 badge 2x thumb
We worked with Gainsight's data science team to build out four machine learning models to build a predictive scoring model at the logo level for accounts based on age and segment. This has helped us move beyond a consensus / gut feel based approach. We weight our health scores 100% based on the predictive score, while also producing sub-scores that help unpack the overall score and allow our CSMs to hone in on where to take action. It's worth noting that the data science workstream is not built directly in the out-of-the-box Gainsight product. We worked with the team that spoke in the webinar to build our "older accounts" model, and have worked with the current data science team to build out our three "younger accounts" models. Hope that helps, let me know if you have any questions.