Step 2: (part 2): Defining the types of data to drive your decisions

May 25, 2017

In part 1 of the blog I discussed the different types of data that you should consider for fundraising decision-making.  In this blog, I continue with the structuring of your data for analysis and targeting.

Data depth

Strong indicators and segments usually incorporate a certain amount of history in their criteria. While some indicators are “of the moment” (i.e. monthly vs non-monthly donor) others require a certain amount of historical depth in their calculation (Ex: to determine active donors, we need the giving history of each donor over the past 24 months to assess whether a gift was present – active – or not – inactive).

As you build your indicators, some data depth rules will be quite obvious.  However, when it comes to building behavioural segments, the data depth rules are usually in a gray zone.  How far back should you be looking to determine which behavioural group someone belongs to?  Given unlimited budgets and resources, you would, of course, want to track donor behaviour from their first gift all the way to the present moment, but that’s not always feasible or possible.  And it may water down a donor’s more recent behaviour.

For analytic purposes focused on determining ask sequences, 3-4 years of giving history is sufficient.  For behavioural patterns to emerge, however, a minimum of 6 years of giving history is preferred; 9-10 years is optimal.

Information Hierarchy

Much of this blog has focused on donor-level data used to identify and group donors together.  But, analytically speaking, most of the time we look at summarized data for decision-making.  So, though we may have donor status indicators, we would tend to ask: How many active donors do I have?  This would use the indicator as a “group by” value to count the number of donors by status.  This same indicator can be used to calculate the total revenues of each group, the average gift size and even the number of donors broken down further by province.  This would provide a type of “profile” of active donors by province, to be able to answer questions such as: What is the average gift size of active donors in Manitoba?

Information hierarchy deals with the different, nested levels of “group by’s” for which your data will be summarized.  The hierarchy is set up to fulfill the fundraising informational requirements you set out initially during Step 1 (

An example of a hierarchy could be:

  • Highest level: Donor Type (Monthly, Non-Monthly)
  • 2nd level: Status (Active, Lapsed)
  • 3rd level: Life cycle phase (New, Growth, Mature…)

Suppose, for this example, you summarized the following fields for each level of the hierarchy:

  • Total number of donors
  • Total revenues in the past 12 months
  • Total number of gifts in the past 12 months

You could now calculate, for each level of the hierarchy, new variables:

  • Average revenue per donor
  • Average gift size

These new variables could be considered to be KPIs (Key Performance Indicators) that would allow you to evaluate the performance of each of the donor groups from the highest level all the way down to the most detailed.  This type of “drilling down” is quite useful when trying to determine which of the details is highest yielding.

A great article to help inspire you to begin your KPI building is “9 Key Fundraising Performance Indicators Every Nonprofit Should Be Tracking” (Desmond, Andrew, March 10, 2015,  He states: “When it comes to determining how efficiently your nonprofit raises money, there’s nothing quite as helpful as a large pool of accurate, relevant fundraising performance indicators that can objectively show you exactly how well you’re doing.”  Couldn’t agree more.