How do I tell if Dataro's predictions worked?

Great question! Let us introduce you to what we call the 'Dataro Curve'.

Remember, Dataro is not predicting a particular financial outcome. Our models are instead predicting how likely a donor is to take a particular action relative to all of your other donors.

Dataro uses a process called 'calibration' that allows us to convert our predictive scores into rough probabilities. These are only approximations, but are very useful for interpreting results. So a  Dataro Score of 0.9 should equate to roughly a 90% chance of giving,  score of 0.1 would be a 10% chance of giving, and so on. Of course there is a margin of error (these are only predictions based on your data, after all!), but the trend should be that higher scores correlate with higher response rates.

That's where the 'Dataro Curve' comes in. If Dataro's predictions have worked, the results analysis should look something like this:

If we look at the same campaign based on Dataro Ranks, the results will look different. But again, donors with higher ranks should have a higher response rate than donors with lower ranks.

This will also often translate into net revenue projections. While we're not predicting revenue specifically, higher response rates tend to correlate with more money raised. 

So that's how you tell if the predictions have worked!