Find more posts tagged with
Comments
Sort by:
1 - 5 of
51
One of my modules in Data-Driven Decision Making for Customer Success course covers forecasting (also available self-paced on Udemy). Regression analysis combined with Expected Value works well for short-term forecasting, and ETS (a standard feature in Excel) works well for long-term forecasting. Both use mathematical models, and measuring and improving forecasting accuracy is an ongoing process.
A less reliable, but popular manual approach is to forecast using your customer health dashboard. These ratings tend to be highly subjective and the math used to calculate them are often suspect, but assuming your dashboard has reasonable predictive accuracy, it can also be used. You can check that by using the Customer Health Validation Analysis Spreadsheet located in the Library>Customer Journey>Engagement folder on this website. You would assign probabilities to color codes and then apply Expected Value to your renewal ACVs, much like CRMs do for sales forecasts.
Happy to jump on a call and talk through these approaches if anyone is interested: Ed@se-partners.com.
A less reliable, but popular manual approach is to forecast using your customer health dashboard. These ratings tend to be highly subjective and the math used to calculate them are often suspect, but assuming your dashboard has reasonable predictive accuracy, it can also be used. You can check that by using the Customer Health Validation Analysis Spreadsheet located in the Library>Customer Journey>Engagement folder on this website. You would assign probabilities to color codes and then apply Expected Value to your renewal ACVs, much like CRMs do for sales forecasts.
Happy to jump on a call and talk through these approaches if anyone is interested: Ed@se-partners.com.
Would love to learn, definitely something I need to improve/get better at as well!
Jordan Silverman
jordan.silverman@usestarfish.com
(914) 844-5775
https://www.linkedin.com/in/jordansilverman/
jordan.silverman@usestarfish.com
(914) 844-5775
https://www.linkedin.com/in/jordansilverman/
To start, I'm in total agreement with Matt , the direction this goes entirely depends on what you want/need to forecast.
As an example, we have a shared account ownership model, wherever there is a CSM there is an AE counterpart. Because of this, upsell/cross-sell/expansion is managed by the AE, so our forecasting in CS is focused solely on the risk of contraction or churn on the existing ARR for all accounts.
Here's what it looks like for each:
- Each CSM reviews their accounts and updates the following for accounts renewing within the next 210 days.
- Best Case: This will never exceed the current ARR (even if there is an expansion on the table).
- Commit: The midpoint between Best Case and Worst Case.
- Worst Case: What's the worst case, how much contraction (usually based on active usage) is possible or might we see the account churn entirely.
- 210 days out from the renewal date we expect the spread between the Best and Worst to be pretty large.
- As we get into the last ~30 days leading up to the renewal we anticipate the spread between Best and Worst to narrow, if not completely flatten, as the team gets more certain on the amount to renew.
Why do we do it this way? It helps get the team focused on the accounts that need extra love and determine where we want to bring in executive sponsorship with plenty of runway. We approach this really conservatively and forecast absolute worst-case numbers early, this helps gather resources and prepare our revenue team mentally for potentially painful quarters (we started this process in March of 2020...).Anyways, this is just one way to look at and approach it, hopefully, this helps get the gears turning.