Direct Mail - Where the Future Lies
Panelist Perry D. Drake
Questions fielded by Perry are below along with his answers.
Question #1 - A big concern is postal hikes that are looming on the horizon. How are your clients addressing this issue?
Answer #1 - Good question. Actually, there are two things I am noticing. First, I am seeing more direct marketers trying to better integrate email marketing with traditional direct mail marketing. And secondly, I am noticing more direct marketers trying to embrace modeling and segmentation techniques both to their house and outside list campaigns to gain further efficiencies.
I have one client, a publisher of a single title, that is in the process of building a new marketing database. We are incorporating email addresses as well as certain other pieces of data from the company website. Doing so, should allow them to more efficiently communicate with their customers via email regarding such things as renewal reminders. And hopefully in doing so, they will see a bump in response.
In fact, based on research conducted by ARM Research, this will most likely be the case. ARM found that when one combines e-mail campaigns with traditional methods, response rates are positively impacted. In fact, two-thirds of the direct marketers employing both marketing methods said they saw a 5 to 10 percent increase in response rates. Some even stated to have seen as much as a 15 percent or more increase in response. This is very encouraging evidence to support the integration of traditional marketing and e-mail marketing.
Question #2 - What is your take on email vs direct mail?
Answer #2 - They both have their advantages and disadvantages. The advantage of email is that you can hit your target quickly and inexpensively. But we must be careful not in inundate our customers with too many messages as their relevance lessons. As Dawn said earlier, we as marketers, are all to one degree or another guilty of training our customers to ignore our email messages.
As I see it, email should not be considered a replacement but rather a supplement to our direct mail marketing messages. It is not an either/or situation. The two should work together. Again I cite the research previously mentioned by ARM Research stating that when one integrates messages we see positive results in response rates.
Also one needs to keep in mind the results of another recent survey done by NFO Research. According to their research, 32 percent of e-mail users stated they change their e-mail address yearly. An amazing number! The top three reasons cited in the research were: a change is ISP, a new job, or a desperate attempt to reduce spam.
Regarding direct mail, the primary advantage is that we can more effectively target our customers given the wealth of information available on customers or prospects. For email lists we are not there yet. However, there are some companies like Markets On Demand that are building fairly solid email compiled lists. What sets them apart from the others is they are building the file off of postal addresses. Everyone on their file has both a postal address and an email address. Unfortunately, they only have 7 million records on their database so far but they are building on that. So, we still have a long way to go.
Question #3 - What modeling options are available so that a marketer can get even more out of direct mail?
Answer #3 - From an outside list perspective, a marketer has three basic options to try and increase response above and beyond applying simple manual selects: (1) a "best customer" or "clone" model, (2) a traditional response model, or (3) a zip penetration model. Regarding "best customer" models, they are cheap and fast. You can build one today if you like and use it tomorrow to select customers from a prospect file. For a response model it takes a bit more time and requires testing first. Regarding a zip penetration model, you base the analysis solely on census level data at a zip, sub zip or census tract level. This works great for lists for which you have no other information available. You can build a zip penetration model to predict best customers or to predict response.
From a house list perspective, the marketer can build reactivation models, cross sell models and segment the customer file for more effective targeting as well.
Unfortunately, not many direct marketers employ such sophisticated techniques. Based on Direct magazines 2001 survey of subscribers, only 31% of consumer and business-to-business database marketers employ such techniques. We have a long way to go to educate database marketers. And, another problem I find is that when some direct marketers do attempt to test modeling for the first time, they often cheap out. I have seen many failed attempts at cloning, for example, that simply failed because the universe was not properly defined. A very important detail in ensuring the "best customer" model works is the selection of the best customers from the database for cloning. Make sure, when trying these quantitative techniques for the first time, you have someone looking out for your best interest, for example a third party. It may cost a bit more but it is well worth it in the end.
Question #4 (follow-up to Question #3) - What if you are trying to model your house names and you just recently made a drastic change to your promotional strategy, say from a soft offer to a hard offer. Can you still build a response model on your customer file?
Answer #4 - Yes, but with extreme caution. You must realize that when you are modeling responders to a product offering, you are not just modeling the customer's likelihood to purchase the product but also purchase the product with the offer given. Does the package offer a free premium? Do I have a 30 days to decide if I want to keep the product? If I order, will I get a discount off of future purchases? A model built to predict orderers based on a soft risk-free offer, for example, will be much different from a model built to predict orderers of the same product but with a hard offer. As such, you are advised to proceed with caution. Applying a model in such a case will be weak at best. I strongly advise against it.
Question #5 - So, thumbs up or thumbs down for direct mail?
Answer #5 - Definitely thumbs up. Again, it is not an either/or situation when it comes to direct mail and email. We should use each together in what ever manner makes the most sense. Again, I restate the NFO research that found 32% of people change their email addresses each year. That is 3 times the national average for postal address changes. For the foreseeable future, direct mail is here to stay.