Level Three Feedback: Using Feedback to Model

March 9, 2018      Kevin Schulman, Founder, DonorVoice and DVCanvass

You had an acquisition package that beat your control like the control owed it money.  Turns out that putting in a coin and a notepad and address labels and a back-end premium in that package really increased response rate!  You have a new control!

Fast forward a year.  Your donor file isn’t growing the way you want.  When you look at the reasons, you see the culprit: a poor second-year retention rate of donors from your new, premium-laden acquisition package.

That’s because the donors that your premium-infested new control was acquiring were loyal to the tchotchkes and not to you.  Once they saw they weren’t going to keep getting hits from you, they went to find a new dealer.

If only you could have determined the quality of donors at the initial point of acquisition.  This is not a challenge restricted to mail programs.  With online, we measure click-through rates, opens, and impressions often without knowing whether the right people are clicking, opening, and impressed.  The street canvasser likewise is using age as a proxy for quality, if quality is measured at all.

But wait!  You can!

By asking donors simple questions like identity, commitment, and satisfaction at the point of acquisition or as close as humanly possible to it, you can see whether the donors you are acquiring are likely to be with you through the long haul.  From that, you can change the way you are acquiring people to get quality, not (just) quantity.

We’ve had a good amount of robust discussion on The Agitator about whether transactional or user-provided information is more predictive of future giving, with both Team Transactional and Team User-Provided Information making good points.  (Team User-Provided Information had better points, but still – both had good points.)

Even the best case for Team Transactional, however, will acknowledge that it is backward-looking.  Granted, there is some predictive value in the amount someone gives as the first gift (namely, bigger gifts are better, but with a trough around the lowest amount on your default acquisition source – people who write in an amount under your lowest ask amount are likely to be frequently low-dollar givers).  But the difference is relatively small.

On the other side of the coin, feedback can be forward-looking.  If collected at the point of acquisition or immediately after, you know, for example, that the high-commitment donors have twice the lifetime value of low-commitment donors.

Armed with this information, you can sack the canvassers who are bringing you poor leads or have them retrained by the ones who are bringing you the best.  Amnesty Belgium did this (plus the type of individual feedback we talked about in the post on Level One Feedback) and cut their cancellations of monthly donors in half.

You can run display ads that are targeted to better donors and absorb the lower click-through rate, knowing you are setting yourself up for success.

And you can consign the kitchen-sink package to the recycling bin before the donors you really want to acquire do it for you.

Retention begins at the point of acquisition.  If you get quality donors to start with, your success isn’t absolutely guaranteed.  But if you put garbage in your system, it’s guaranteed you will get garbage results.

Nick

One response to “Level Three Feedback: Using Feedback to Model”

  1. Really appreciate this garbage in/garbage out analogy. It can be applied to so many types of fundraising strategies. Not just direct mail and online acquisition, but also to acquisition through events. Targeting is important. So is leading with mission and values, rather than tchotchkes and golf tournaments. Thanks for the reminder.