Here we will show you the method for sales forecasting using pipeline stages, and give you this online tool to see how it works.
No doubt you are here for the tool itself, so let's begin with that.
Sales Forecasting Tool
Firstly, decide what period you are forecasting over - for example this month or this quarter. Then for that period, get the historic (but recent) win rates per stage, and currently open pipeline per stage, and input below to generate a forecast.
Intuitively, the later the sales stage an opportunity is, the more likely it is be closed won. We would expect there to be a strong correlation between the sales stage and the probability of winning the deal.
Therefore - it makes sense to calculate a probability of winning, per sales stage, and use these for extra accuracy compared to simply multiplying total pipeline by your win rate.
Calculate Win Rates
Firstly, you will need to calculate your win rate from each stage of the sales funnel. Make sure to go far back enough that you have enough data, but also stay recent enough that the data reflects your current business and environment.
It is important that you use pipeline win rate ([sum of closed won] / [sum of everything that was open pipeline]) rather than regular win rate ([sum of closed won]/[sum of closed won and closed lost]). If not, you will over estimate win rate by the amount of deals that slip to the next period.
Get Pipeline Stage Breakdown
Get the amount of open pipeline, closing in the period you are reporting for, per stage. You should be able to get this from your CRM, analytics platform, or do it in a spreadsheet.
Multiply and Sum
Multiply the pipeline per stage by the win rate for that stage to get forecast per stage, and add them up.
This is a relatively simple method, and as such there are some limitations.
- It only forecasts based on open pipeline, so can only be used to forecast in the near-term
- It does not take into account other important factors such as age of an opportunity
- If your company is growing quickly, the win rates used can be inaccurate if too large a time period is used