Must-Have KPIs to Track to Ensure Forecast Accuracy.
Sure our forecast measures our sales performance, but how do we measure our forecast performance?
Without the right metrics and procedures in place, we have no insight into how well we actually forecast. Moreso, when it fails, we don't know how to make it better.
So, in order to make your most accurate sales forecasts ever, make sure you have these metrics in place...
- Annual forecast vs. quarterly results
- Forecast by region and/or product group
- Pipeline coverage
- Forecast category and movements
- Opportunity age
- Risk score
We gathered a panel of some of the best revenue and sales operations leaders, consisting of:
- Stephen Daniels, Head of Revenue Operations, Branch
- Victoria Moss, VP, Revenue Operations, Greenhouse
- Anu Krishnakumar, VP, Global Sales Operations, Smartbear
The group discussed how exactly they measure the accuracy of their sales forecasts. Here’s what they had to say. You can also watch the full recording at any time, here.
1. Annual forecast vs. quarterly results
Victoria Moss, VP, Revenue Operations at Greenhouse recommends that at the end of each quarter you re-examine your annual forecast through the lens of the quarterly results.
Check and see if your yearly goals are still attainable, or if you will surpass them. Regardless of the outcome, this will either prevent an end-of-the-year scramble to make goal, or give support, onboarding, and/or success teams a chance to prepare for additional users.
2. Forecast by region and/or product group
For greater forecast accuracy, it is important to adjust your forecast KPIs to fit the needs of your situation. Two examples are within geographic regions or for different size companies.
Stephen Daniels, Head of Revenue Operations at Branch breaks his forecasts up by NA, EMEA, and APAC. Each of those regions has different working cultures and therefore deals have different expectations. Applying the same sales cycle to all of them will only result in a less accurate forecast. To find the benchmark, he recommends applying a historical win rate.
Tori echoes the sentiment, recommending a best practice of breaking up your forecast by the company size. Such as the following:
- Small business - 30-day sales cycle. Forecast on deal volume and trends, calculated by opportunities x win rate x average account value.
- Mid-Market - Deal-by-deal forecast with a “gut-check” based on quota capacity.
- Enterprise - Deal-by-deal forecasting with weekly MEDDPICC-driven inspections by both directors and managers.
3. Pipeline coverage
"Pipeline coverage is table stakes,” says Anu Krishnakumar, VP, Global Sales Operations at Smartbear. Every company should be looking to have between three to 5 times their forecast covered in their pipeline. Stephen adds that he calculates pipeline for different geographic regions based on the historical amount needed to meet expectations.
4. Forecast category and movement
Forecast categories: included deals, excluded deals, and why opportunities fall within them are very important. If sales reps and managers are doing a proper job of forecasting, you should be able to gather good information about what will close, and what’s at risk.
Tori likes to look at what was called “likely” on day one of month one of each quarter, compared to day one of month two. She looks for deals that have been downgraded, and which ones have been added, and why.
5. Opportunity age
Also consider the age of an opportunity in these forecasts. If one starts to draw out for longer than the average sales cycle, risk on it should be gradually increased. As extended deals are statistically less likely to close as time goes on.
6. Risk score
Stephen also loves to use BoostUp’s AI risk score to judge the accuracy of his forecast. The AI monitors deal activity, including buyer and seller engagement through all sales channels to determine if there is meaningful communication occurring and if the deal is being progressed. If the forecast includes a lot of high-risk deals, it’s time to take action.
To watch the full recording, click here.