You often hear the advertising wisdom that it takes seven impressions to produce a buying decision. You also may have read in our Membership Marketing Benchmarking Report the power of continuing contact with renewing and lapsed members.
But how do these concepts apply to the frequency of contact when looking at membership recruitment or acquisition? Is there a correlation between the number of new member solicitations and the likelihood of a prospective member joining?
We asked this question recently and set up a test to help find the answer. Here is what we did.
With one membership marketing client that uses fairly static prospecting lists that we have compiled, we went back in our data to identify the number of times we had sent a mailing to the prospective member over the past three years. We then tagged prospective members by the number of contacts.
Then in a subsequent campaign, we defined our lists by the number of mailings that the potential member had previously received and tracked back our responses to these “lists”. Here is how responses broke out.
Clearly our analysis highlights that for this group there is an erosion of response from people who have been mailed multiple times. So with this organization, to maximize net revenue and work with tighter budgets, we dropped the prospects who had been contacted more than seven times from some upcoming campaigns. By eliminating these prospects, we saw response rates jump up. However, after resting these records for several campaigns, we will re-introduce them in future promotions.
In the coming months, I will be doing more analysis with mailers who have a longer history and higher frequency of contacts.
But how do these concepts apply to the frequency of contact when looking at membership recruitment or acquisition? Is there a correlation between the number of new member solicitations and the likelihood of a prospective member joining?
We asked this question recently and set up a test to help find the answer. Here is what we did.
With one membership marketing client that uses fairly static prospecting lists that we have compiled, we went back in our data to identify the number of times we had sent a mailing to the prospective member over the past three years. We then tagged prospective members by the number of contacts.
Then in a subsequent campaign, we defined our lists by the number of mailings that the potential member had previously received and tracked back our responses to these “lists”. Here is how responses broke out.
Clearly our analysis highlights that for this group there is an erosion of response from people who have been mailed multiple times. So with this organization, to maximize net revenue and work with tighter budgets, we dropped the prospects who had been contacted more than seven times from some upcoming campaigns. By eliminating these prospects, we saw response rates jump up. However, after resting these records for several campaigns, we will re-introduce them in future promotions.
In the coming months, I will be doing more analysis with mailers who have a longer history and higher frequency of contacts.
Posts
4 comments:
Thanks, Tony, that's valuable information and good food for thought.
~Tina Horn
American Association of Naturopathic Physicians
Tina -- Thanks for the feedback. Interestingly, in real life when we dropped the bottom group from our last mailing we did see a nice upward bump in response rates. The key decision here is whether you want to maximize members volume (the number of new members) or net revenue. Good marketing is a balancing act between the two.
Tony
This test seems a bit flawed, though. The respondents (those most inclined to join) come out of the first few mailings. If the list is, in fact, fairly static - then you are mailing less-and-less qualified prospects as the # of mailings goes on. Try some "chronic non-responder" analysis to boost response in the later contacts.
Thanks for the feedback. What we have done with this client in order to maximize revenue is to take out prospects who had numberous opportunities to join and give them a rest. The test showed us that they were less likely to respond.
What we are doing is not looking at the list as one entity, but as a diverse group that has some segments that respond better than others. Previously, we treated everyone the same. Tony
Post a Comment