Donating makes you happy.

Dietmar Hopp

The Diak

The Evangelisches Diakoniewerk Schwäbisch Hall e.V. (a church charity, nicknamed ‘the Diak’) is a charitable service organisation for elderly and sick people. With around 2,300 employees, it is the second-largest employer in Schwäbisch Hall.

It was founded in 1886 and has the following business areas:

  • Work with the elderly with a current total of 321 spaces.
  • Out-patient services with 330 ‘brothers and sisters of Hall’ in 20 welfare centres. Around 125 nursing professionals work in the Diakonie daheim nursing service, along with nearly 125 care specialists, nearly 100 household assistants in care roles and over 60 volunteers in nine care teams in the Schwäbisch Hall region.
  • Hospital: The Diakonie hospital offers centralised treatment services and is an academic teaching hospital for Heidelberg University with a 492-bed capacity. Each year, around 24,000 patients are treated as in-patients and more than 36,000 as out-patients.


In fundraising, tight budgets are used to achieve the highest possible volume of donations. The advertising and fundraising efforts represent a significant proportion of this. This gave rise to the following questions:

Where can costs be reduced without causing unwanted side effects?

How can we understand donation behaviour better?

To answer these questions, Proventa used classic data-analytics processes to save costs where possible – for example, to optimise the response rate to donation mailings.


To start with, the donor and donation data were consolidated in a central database and prepared for further analysis. In the next step, the central question of which donors react particularly positively to donation mailings was answered by using data-analytics processes:



This determined global donation profiles that could be used for the general management of mailing activities.



This determined specific donor and donation profiles that were suited to the optimisation of specific mailing activities. The knowledge won was then applied to a simulation of actual mailings so that only donors fitting the donor profiles were contacted. Through this process, for selected mailing activities, it was possible to achieve an increase of 40% in the donation surplus.



An approach based on the CRISP standard was selected that basically proceeded in phases as follows:

  • Data consolidation
  • Exploratory data analysis
  • Data preparation
  • Iterative application and benchmarking of data-analytics processes such as neural networks, cluster processes and decision-tree processes



In the different phases, various tools were used:

  •     MySQL database for central data storage
  •     MS Excel for analysis purposes
  •     Pentaho data-integration tool
  •     Various data-analytics tools (WEKA, KNIME)