From the moment Peter Wylie and I started talking about the topic of Do Not Solicit codes in university databases, the subject of his guest post from last week, I knew it would be a hot topic. Restrictions on solicitation and contact are coming under increased scrutiny — or should be. For example, annual Giving programs are becoming more aware of growing attrition in their prospect pools, and are looking to re-evaluate restrictions put in place years ago, sometimes for inadequate reasons. Peter Wylie’s paper is important because it shows that you can identify a segment of DNS-coded people for whom you can attempt to re-initiate contact, with reasonable assurance of success and minimal backlash.
Contact restrictions can be a sensitive (maybe even a legal) subject, so proceed with caution, certainly. But I think we should look at Do Not Solicit codes in the same way we think of every other type of leakage and attrition in our prospect pools: lost alumni, disconnected phone numbers, and so on. We need to 1) follow up to potentially rectify the situation, and 2) prevent wholesale DNS coding from happening in the first place.
I suspect it is common for alumni to request not to be solicited (perhaps in a fit of pique at being called at dinnertime) and then completely forget that they have done so. Or they will ask not to be contacted in any way, and then later complain they aren’t getting their alumni magazine or invitations to events. Sound familiar?
It’s my opinion that all restriction/exclusion codes should be reviewed periodically — at least the really old ones should be. Ideally you would do this review in consultation with someone who has some institutional memory. What were the codes intended to do? Are they still relevant? Were they supposed to be permanent? Who created them?
The other half of the review consists of studying the individuals who have these codes. If you do predictive modeling, presumably you don’t exclude these individuals from your models — they should get scores just like anyone else. Pay particular attention to the upper decile or two of these folks. As Peter Wylie suggests in his paper, are there any here who really should be approached again?
I’d be willing to bet that some of these alumni are not un-engaged at all. Have they attended events? Do they have a son or daughter attending? Are there other points of contact, such as a survey response or making a submission to the alumni magazine, that indicate a restriction might be reversible?
What you do next depends on your particular circumstances, but some advice I have from others suggests that the first touch should not be an ask. Try an event invitation or some other communication from the alumni office. Study the results.
The other side of the coin is prevention.There are two angles to explore in order to stop the bleeding. First, get control over your coding going forward. And second, don’t alienate alumni by your contact practices.
If your database allows it, start tracking the SOURCE of DNS and other exclusion codes in the database. A key aspect of coding the source is differentiating between alumni-requested exclusions (“I don’t want any more phone calls”) and exclusions assigned by the institution. Clearly we need to honour the wishes of alumni, but the validity of our own determinations should be questioned regularly. Was that angrily hung-up phone five years ago still a good enough reason to avoid calling? Were a lot of codes being created across swaths of alumni at the request of development officers? (And how many of those “assigned” alumni have actually been contacted or visited?)
In short, make sure your codes are specific, explicit, sourced, and certain. If they’re solid, honour them to the hilt. If not, consider them time-limited and subject to review.
The other angle, “don’t alienate alumni,” is obviously a murkier issue. But speaking as a predictive modeler, I suggest that we should ask the way donors prefer to be asked. I’ve observed that quite a few people who hang up on student callers (and then risk ending up as DNS) are actually good by-mail donors. They just haven’t communicated their preference to alma mater. So if you do predictive modeling, build one model trained on by-mail giving, and one model on by-phone giving. I do this, and I’ve observed a fair bit of spread in the predictive scores (i.e., many alumni score much higher in one model than the other).