Old World Computing Supports Predictive Analytics Solution for DonorVoice and Amnesty International Belgium – Flanders

“On a scale from 1 to 10, I would give OWC a 9 for the result and the cooperation. So that would be an ‘Excellent’.”

 Ilja De Coster, – Senior Fundraising & Data Strategist for DonorVoice and Amnesty International Belgium – Flanders

 → Click here for a PDF version of the interview.

Ilja De Coster, –Senior Fundraising & Data Strategist for

Amnesty International Belgium – Flanders,

photo: private

Dr. Maximilian Fischer of OWC,

photo:  © Anke Sundermeier, ankesundermeier.de



Amnesty International Belgium – Flanders utilizes fully automated telemarketing campaigns to increase the amount of donations, employing heuristics and machine learning techniques. Old World Computing provided inspiration, best practices and laid the groundwork for a predictive analytics solution Amnesty continued to build on. François Baumgartner, Head of Communications and Marketing at Old World Computing, talked to Ilja De Coster, Senior Fundraising & Data Strategist, and Dr. Maximilian Fischer, data scientist at Old World Computing UG (OWC), about the newest successful project.


Mr. De Coster, what does DonorVoice represent?

Ilja De Coster: DonorVoice is a consulting agency focused on donor retention. We are active worldwide and use data analytics and insights from behavioral science to support charity organizations in understanding and retaining their donors: Donors are among the most important stakeholders of such organizations as the humanitarian projects and human rights activities can only be funded and successfully continued with the help of their membership and donations. The donors’ experiences, expectations, and wishes are therefore in the center of our attention. All this has to be taken into account in fundraising: we do not only collect data, but learn from it and actively implement our knowledge gained to support charity organizations.

In which countries is DonorVoice active and what exactly does your consultancy service entail?

Ilja De Coster: DonorVoice is active in the US, Canada, Great Britain, and selected countries of the European Union, primarily in consultancy and research projects. We are further involved in projects focusing on data-oriented process optimization. Our focus is donor retention and to that end, we use a data platform, through the use of which many data processing, analytics and donor relationship building tasks can be automated. This reduces operational stress and costs.

Accordingly, how would you describe DonorVoice and their distinguishing feature using only one sentence?

Ilja De Coster: In a nutshell, we do the research and then implement the insights into day-to-day operations: DonorVoice has unique research-based expertise in the field of donor relationship management and successfully employs this expert knowledge with regard to donors and in the spirit of charity organizations.

Dr. Fischer, please describe the project order for DonorVoice and Amnesty International.

Maximilian Fischer, Data Scientist at OWC: The project was called Campaign Automation and Optimization using Predictive Analytics, which really puts the objective in a nutshell. We helped to develop new and more efficient processes for Amnesty and their telemarketing campaigns and implemented these in cooperation with DonorVoice. Moreover, and perhaps even more importantly, we were able to pass on best practices and expert knowledge that do not only apply to the project at hand, but to many tasks at DonorVoice, thus improving overall work efficiency and serving as an inspiration for many projects to come.

What did the process look like before?

Maximilian Fischer: A telemarketing campaign works as follows: people already donating to Amnesty are phoned again and are asked whether they were willing to increase the amount of their monthly donations. Previously, the people who received calls had not been specifically targeted. The only restriction had been the exclusion of certain groups, for example recently acquired donors or those who had already been called in the past. Additionally, there had not been a solution for the automatic integration of the call results into the IT systems at Amnesty.

What is new?

Ilja De Coster: We collect and document the call results in order to know exactly who is more likely to be willing to gradually increase their donated amount. The call results are then transferred to the respective IT systems. After this, a weekly, automatically generated list of people who will be called in the upcoming week is constructed. This is an ongoing learning process and we are still optimizing our results, benefiting from the foundation Old World Computing has laid out for us.

Which methods are used for this and what does the result look like?

Maximilian Fischer: Machine learning techniques, i.e. predictive analytics, now support the mentioned heuristics in the selection of who to call. The call results are used to identify persons very likely to increase the amount of their donations. Thus, the selection can take place in a significantly more targeted fashion.


“RapidMiner is a great tool, but Old World Computing made it even greater”

Ilja De Coster, –Senior Fundraising & Data Strategist


What are the new advantages?

Ilja De Coster: The work of Old World Computing is really excellent, we can work even more effectively with the new tools by OWC; furthermore, the fundraising procedures are now faster and better. But what is even more important to me personally is how OWC was able to help us with general RapidMiner matters, imparting knowledge we can use on a daily basis. This also led to the development of some new functionalities for the RapidMiner platform, now available to everybody as part of the Jackhammer and Advanced Reporting extensions. They even wrote the Web Automation Extension for our project, which still saves us many working hours every week. For example, with the groundwork laid out by them, the procedure for the entire telemarkting done by Amnesty, and the procedures on the DonorVoice Retention platform, could be greatly improved. RapidMiner is a great tool, but Old World Computing made it even greater.

Mr. De Coster, are you as well as Amnesty International satisfied with the results?

Ilja De Coster: The feedback from Amnesty International was great. OWC has excellent knowledge in the field of big data and machine learning. What’s more, the cooperation with the team of OWC is always most pleasant and highly professional. The first benefit was increased efficiency and productivity in many areas of our work. We are still profiting from the project in that we have adopted best practices and better procedures. Even more, telemarketing with the model for better targeting shows very promising results.

Could you rate your evaluation on a scale from 1 to 10?

Ilja De Coster: I would give OWC a 9 for the cooperation and the result. So that would be an ‘Excellent’. I’m only not giving a ten to leave room for improvement.

Would you recommend OWC?

Ilja De Coster: Yes, I will most definitely do so. We will also be happy to turn to the Bochum start-up company again in the future, if we feel the project and the resulting project order and requirements apply to OWC.

Dr. Fischer, would you recommend DonorVoice?

Maximilian Fischer: Absolutely. Cooperation with DonorVoice and Amnesty International was exceptional and will be happy to receive more project requests.

Thank you both for the conversation.