Alexa, what’s my Forecast?

Arnulf Hsu

By

SalesDirector.ai is an AI based Forecasting Platform.

Here we try to explain how we derive your Worst Case, Most Likely and Best Case Scenarios.

SalesDirector.ai builds highly advanced Dynamic AI Models based on your historical data and applies them to your existing Deals. Read more in depth here.

Each Model contains Sales Stages, duration and win rates based on a whole bunch of criteria (activity, personas engaged, sales cycle attributes such as MEDDIC or other fields, etc…).

First, we forecast by a Period, meaning you set your forecast for this Month, Quarter or Year.

Next, we predict your Worst Case, Most Likely and Best Case using the following scenarios:

Best Case:
Takes all your existing Opportunities expected to close within the Period and runs them as fast as possible (20th percentile) and then applies the most statistically significant Win Rate against those deals.

Most Likely:
Takes all your existing Opportunities expected to close within the Period and runs them using your mean and then applies the most statistically significant Win Rate against those deals.

Worst Case:
Takes all your existing Opportunities expected to close within the Period and runs them as slow as we have seen (80th percentile) and then applies the most statistically significant Win Rate against those deals.

Take this example below:

If we focus on the Most Likely scenario, we are only forecasting $175 total, since for Deal C we have never ever seen a deal of this type/size/rep, etc. close within 30 days, and since the deal started in March 1, the likelihood of closing this by March 31st is extremely low so we exclude it. The other 2 deals are within what we have seen before, and as such we apply our win rate probabilities and then come out with a risk adjusted number.

While we realize deals usually have a binary outcome (closed won or lost), using probabilities against the deal amounts is the best way to get to the Sum number.

Once we have the “CRM” adjusted number, we then look at the deals and evaluate where the AI thinks the deal stage is. The system will upgrade or downgrade the deals accordingly and then apply similar modeling per above. The system will then come up with a “System Recommendation” for your Forecast.

Really what you are aiming for are numbers where there isn’t a lot of variance between what “CRM” says and what the AI says. If there is a large variance usually there is risk in that area of the business.

Hopefully this helps clarify a bit around how the Forecasts come to be.

Ready to see your AI forecast? Book a demo today. 

Get the latest content delivered to your inbox.