Michael Rogers

Predicting Donor Behavior

Yes, donor behavior can be predicted.

Most fundraisers are keenly aware of the challenges associated with acquiring new donors, retaining existing donors and increasing donor value. To meet these challenges, fundraisers are embracing campaign optimization and lifestage management. 

Campaign Optimization

For most organizations, the majority of their public contributions are garnered via direct response campaigns. These campaigns generally employ a selection methodology; whereby donors are included or excluded based upon the donor’s recency of their last contribution, the amount of their largest contribution and the frequency in which they contribute – also known as RFM. The RFM selection methodology has many inherent benefits. Namely, it is an easy method to categorize donors and track their performance. Also, RFM allows a campaign manager to easily include and exclude groupings of donors based upon their RFM behavior. This method works stellar for segments that are solidly profitable or unprofitable.  But, for segments that are marginally profitable or unprofitable, RFM alone isn’t so stellar. More times than not, unprofitable donors will be included or profitable donors will be excluded from these segments.

So, how should a strategic fundraiser improve the performance on marginally profitable segments, or be so bullish as to find ‘pockets’ of profitable donors in marginally unprofitable segments?  Enter stage right, campaign optimization. Optimizing campaigns can be done several different ways. Let’s focus on RFM+A – ‘A’ttributes.  Applying single-variant analysis to any RFM segment can provide a degree of discrimination that can easily detect donor groupings that are profitable. 

For example, let’s assume we have an RFM segment grouped with 10,000 donors who have given a gift in the past six months, with a largest single contribution between $25 - $49.99 and an annual frequency of one gift. Further assume that this RFM segment has a 3-month ROI of .91. In other words, for each $1.00 expended on this segment, the return is only $.91. Now let’s suppose that we parse this segment of donors into age bands such as 40 – 49, 50 – 59, 60 – 69, etc., and find that individuals who are 60 years of age or older are profitable and those under 60 years of age are unprofitable. 

Age is just one attribute that can be used, but other attributes such as household income, wealth, lifestyles, philanthropic giving to other non-profits, giving to other channels (e.g. online) can be used in this RFMA method. By default, clients using Convio’s campaign management services automatically gain access to these demographic, philanthropic and multi-channel variables. This provides a means to more effectively cultivate valuable relationships with constituents.

Lifestage Management

Fundraisers have a unique opportunity to proactively identify donor lifestage events.  A lifestage event can be defined as a point in time when a donor increases or decreases their loyalty to an organization.  For example, the period of time following a donor’s first-time gift is a lifestage.  For most organizations, 60% or more of their first-time donors will never give again. Conversely, a portion of these first-time donors will become monthly sustainers.  The trick is to identify which pathway a specific donor will follow.  Identifying other lifestages such as active donors who are at risk of defection; middle donors who are likely to increase share of wallet; lapsed donors who are most likely to re-engage; or even identifying event participants who are likely to become financial contributors can significantly improve fundraising performance. 

Being able to identify when a donor enters a lifestage, and predict the likely behavior that a donor will exhibit during the lifestage is critical for successful lifestage management. Additionally, having an effective strategy to influence the donor’s behavior within the lifestage is equally important. Convio’s Strategic Services utilizes data modeling techniques to effectively predict the likely pathway a donor will pursue within a lifestage. In the above example of the first-time donors, a model can determine which new donors have the greatest propensity to become sustainers, and determine which of these donors will never give again.

Powerfully, this allows timely and relevant communication strategies to be implemented against both sets of donors; therefore, increasing the retention and lifetime value of those donors likely to become sustainers, and decreasing the amount of budget expended on those donors who are never likely to give again.

One last important point – managing these lifestages can be a daunting tasks and one that should be managed with Constituent Relationship Management (CRM) technology. Manually managing a donor’s movement in and out of a given lifestage is a recipe for failure unless your database contains less than a few hundred donors. To do this correctly, one should consider Convio’s Luminate CRM™ a technology built on the Force.com platform which natively automates the identification, communication and management of donors from one lifestage to another. This takes the focus off of the logistics and lets the fundraiser focus on the strategy of retaining and improving the value of each donor relationship.

Postage Savings Through Analytics

Are you mailing to all of them?

Several of our clients have asked for my input on a couple topics in the past week: the first being the postage increase that takes effect on April 17th (first class and standard mail presort will see an average rate increase of 1.8 percent) and the second being the recent blog post by Roger Craver questioning whether the top 83 direct response fundraisers should be fired. My response to both topics is the same… it is time for nonprofits to mail smarter.

As Roger Craver points out, new donor conversion rates continue to drop, while the USPS continues to increase postage rates. This is a double-edged sword for donor acquisition. So, how can we use analytics to reduce postage costs in acquisition without harming results?

The first step is to evaluate the projected long-term value (i.e. 24-month) of the newly acquired donors by list source. (Note: long-term value should consider subsequent communication costs after the donor is acquired.) Nonprofits do not consider the complete picture when they make decisions on list selection solely by measuring initial response rate, cost per donor acquired and average gift. The more important question, "Are we acquiring the right donors, the donors who have the greatest potential to become multi-year donors?"

You may have used the same acquisition lists for several years. If so, look at your acquisition lists from 2008/2009 then calculate your average18- or 24-month long-term donor value by list. (Be sure to include the costs of subsequent communication.) Are you breaking even or making money on the majority of lists, but not others? You should strongly consider eliminating the lists where you are not receiving positive net income within 24 months.

After considering long-term value and narrowing the field to worthy acquisition lists, the next step is to use additional data analytics to eliminate unresponsive donors.  By using a statistical approach, we can eliminate "bottom-feeding" prospects by creating a cloning model to identify the profile(s) of valuable donors. This additional level of suppression saves money by removing unproductive prospects.

We use a statistical approach based on each client’s unique database; rather than a more generalized cooperative model. For one nonprofit, a specific profile may be unproductive, but a similar profile for another nonprofit may perform very well. For example, Leisure Buffs and Aging Upscales may be more heavily weighted for one nonprofit; whereas, Gen X Singles and Jumbo Families are better performers for another. For this reason, Convio through our recent expanded team and offerings, creates models uniquely for each client. 

The modeled approach allows nonprofits to focus on engaging the best possible prospects while lowering the cost to acquire a donor. Case in point, Convio created an acquisition model that suppressed 28 pecent of the intended acquisition mail universe, thus decreasing mail costs by the same margin. The nonprofit client saw a 4 percent increase in average gift from new donors (due to the fact that we were suppressing lower value donors). Additionally, the modeled test group demonstrated a 30 percent increase in net revenue over the control group. Now, that’s smart mailing.