The model works by scoring all individuals on a database in terms of their likelihood to respond, using hundreds of variables.
Charities can maintain their existing mailing segmentations, and select the most promising donors from within a segment instead of blanket mailing the whole list.
According to Occam, most charities mail their supporters five or six appeals each year, using a typical recency frequency value segmentation, and achieve response rates of 12%-15%.
It claims that occam.predict increases response rates and return on investment as well as reducing costs and mail wastage.
It cited the example of a campaign for one client where after the model was used the response rate climbed by 3.3 percentage points.
Scott Logie, managing director of Occam, said: "Charities can now truly engage with their supporters on an ongoing basis at a level appropriate to them individually. It will enable fundraising to be far more efficient in the future."
The model takes approximately three weeks to build and deliver, and one day is required before each appeal to run the model and score data.
If you have an opinion on this or any other issue raised on Brand Republic, join the debate in the .