February 21, 2007 Webinar
On the Money...Savvy Database Marketing Strategies

Question and Answer Session Recap

Do you recommend that an RFM analysis be done on school buildings/districts or on the individual customer names? And can RFM be done on two levels with a parent/child-type relationship?

Romans: If you have never done an RFM analysis before, I would encourage starting at the institution level. RFM at the name level is more difficult because of turnover. While names are very important, purchases are controlled by school budgets. Once your product is being used in a school, loyalty to a vendor really resides at the institution level. RFM is a pretty straightforward analysis. But you can weight it--put more toward frequency and recency rather than monetary and not have a straight line and use other attributes of the district. I would recommend sticking with a pure RFM analysis and let the scoring lie where they lie.

What if you're just starting out and you don't have much of a customer base yet, what should you do?

Stimolo: For starters, you should be aware of the fact that you're just starting out. It has to do with your planning, budgeting, and if you're involved with raising funds. You can tell your investors or your husband that this is going to be as bad as it is going to be, because the customer file as it starts to grow is going to improve the return on your promotion investment over time. Another consideration for you, especially if you're in a business where you don't have unique products…let's say you're a distributor selling the same products that many other companies sell. You're trying to get started; you need to have an edge, so you look at discount or special service consideration or perhaps some combination of the two. Try to give your prospects a reason to buy from you rather than from your competitors.

How do you measure the value of mailings to customers who may be ordering your product through a wholesaler? This participant's experience is mailing a catalog to them is still valuable even though they don't buy directly from them.

When you're selling to a wholesaler, how can you tell if it is a current customer or a prospect?

Stimolo: I don't think you can address it. You basically look at what your overall sales are doing, and if your sales are going up and they are commensurate with the investments you're making in your promotion, you declare a success.

Winter: You look at that distributor as the customer, and I think you can support them by providing some tracking and some incentive. And you have to realize that some distributors will not share that customer list. But I also think this market is predictable enough than when you're selling direct and selling through channel; some patterns you see emerging from your own customer base are somewhat applicable to the customers who will be buying through channel.

How do those mailing intervals translate into e-mail intervals?

Romans: When you talk about e-mail, it is in support of acquisition program or in support of direct mail. If it is in support of direct mail, you want your e-mail to get there just prior to or just as the direct mail is arriving, especially within the two-week time frame. With e-mail you want to do it as frequently as you can afford to do it but still showing results. There is no magic number as to how close the interval is. You want to watch your open rates and click-to rates very carefully.

You mentioned that it is important to test to be sure using profile information for prospecting yields expected results. How can this be done when it is so difficult to track in the education industry? And what are the best practices for tracking mail response?

Winter: That's where e-mail becomes very, very efficient. It is a great way to begin to test, and you can do it in smaller quantities.

Romans: One way that I know how customers have done is to take a look at their customers they have acquired in the past school year and look at their characteristics. Do they follow the same pattern as their mailing strategy, or do they fall outside of their mailing strategy?

Stimolo: There are three different techniques that I'm used to using for testing. You want to test the efficiency of a particular catalog…let's say I mail in January and I mail in August, but I want to see how March does. I put a prefix or suffix to the item code--that prefix or suffix will be invisible to my accounting inventory system. It will only show up in my marketing information system. I'll get close to 98% trackability back to that item code. If I want to test an offer or a creative approach, I'll isolate the geographical area--the number of states that are equal to the sum of the test quantity and the control versus test--and then I'll split it by odd and even ZIP Code. When the program is over, I'll have my data processing people go into the sales data for those odd and even ZIP Codes and compare the sales…and that will give me virtually 100% trackability. The most difficult thing to test is list segments. The only way to test them is to use source codes. The best of us are able to collect approximately 20% of the source code data. The trick is to test in a large enough quantity so the 20% return is statistically significant and to make sure that both the president of the company and the manager of the order entry area are on board as to how important it is to collect that data.

Can you give different ways to track school catalog mailings that are not source coded?

Stimolo: The most common practice is to prefix or suffix the item codes with a catalog code. For example, if the sku number is 1234 and you want to track the performance of a January catalog as opposed to a March catalog, prefix the January sku to J1234 and the March sku to M1234.

Winter: Add a tracking code (you can track by list) to the mailing address. MDR and other list providers can include this in the data they send you, so you can track by list. You can also create unique landing pages and send prospects and customers to separate online landing pages, tracking offers and lists that way as well.

Is there research to show that marketing campaigns that include offers (discounts, free materials, etc.) show greater response rates than those that do not?

Winter: Yes, in fact, the Direct Marketing Association (DMA) offers several research studies that show the efficacy of specific offers. You'll also find data from the American Marketing Association.

Stimolo: Tons of it. What's unique about school marketing is the timing, how you test and track results and the mailing lists. Everything else--offer, copy, design, formats, etc.--is directly transferable from consumer and B2B marketing. The Direct Marketing Association has a library of books about different offers and merchandising techniques that have been successful, and you should avail yourself of that information. Remember, not every offer will be successful. One offer will be more successful than any other. It is your job as a marketer to discover it.

With e-mail, how do you use the correct verbage to not be snagged as spam?

Winter: Avoid words like FREE, SAVE, HURRY, SALE, etc., in your subject lines.

Ziemnicki (MDR's E-Marketing Product Manager): Generally speaking, try to create an e-mail message that communicates your value proposition and your relevance to the reader, and don’t get hung up on spam filter flags. Spam filters generally take into account a large number of attributes of the message, including the words used, the percentage of the message that is image based vs. HTML, reputation of the sender, etc. Use common sense when creating your messages. Clearly, overuse of things like FREE and CA$H in your messages is probably going to raise your score, but used sparingly, they don’t add significantly to your overall deliverability on average. A good technique to employ is the use of a spam filter testing application (such as the one that MDR offers; contact us for details), which can give you a spam filter score of your creative before you send it. While this will give you only one opinion and not guarantee deliverability, it can be used as a guideline as to the likelihood that it will encounter issues.

Any advice if your goal is simply to develop awareness (by mailing out a free product) of the organization and cause as opposed to selling a product? We are nonprofit, mailing to 5-12 grade schools.

Winter: Be sure to ask educators to "share" your materials and messages with colleagues, invite them to visit your Web site to learn more about you, include testimonials from other educators who are currently using your program successfully. It is also important to provide information about your organization, so educators can "trust" your credentials, agenda, etc.

Romans: The concept remains the same, but the approach needs to be different. Even though it may be a free product, some institutions and personnel will be more responsive than others. The key to measuring awareness is to create some interaction between you and the end user. Perhaps create a feedback device on the product and ask for name and address information. Educators like to give feedback. The feedback response would be one measurement, and a profile of the feedback will help you target the best audience to send the product. Another idea is to use e-mail as the method for feedback. This way you can use an option like "send a friend" and encourage them to tell their colleagues in the district about your free product.

E-mail marketing hasn't been successful for us because many schools have built-in firewalls which do not permit e-mails to be received from the outside. Has this been a problem for anyone else on this webinar?

Winter: While school districts' firewalls have become more sophisticated and stringent, proper wording of subject lines will help you improve your deliverability rates. You might also want to consider smaller "pulsed" mailings that don't overwhelm a district's e-mail server.

Ziemnicki (MDR's E-Marketing Product Manager): Generally speaking, just like B2C and B2B, there are some (small percentage) of deliverability challenges facing educational e-mail marketers. The good news is that they do not present a significant deliverability issue and are very far behind those of the B2B space. If you institute best practice techniques when building your own customer based e-mail list--such as white listing instructions, user-defined preferences, etc.--you should have strong deliverability. Additionally, when prospecting, it is a good practice to your campaign results and reply-to messages (the feedback loop) as you can identify schools or individuals that might need follow-up.

Do those same buying cycles apply to the college market too?

Winter: The higher ed market definitely operates with its own set of buying cycles, directly related to quarter, semester, and trimester institutional course schedules. Of course, technology solutions, supplies and equipment, and other product/service categories are also reviewed and purchased according to university schedules and their "type," such as institutions with student residences, commuter schools, etc.

Is there research to show that the type of mailing piece you send can affect the response rate? For example, mailing a simple cover letter and a flyer in an envelope versus mailing an attractive self-mailer in color. Or is it really all about the message?

Romans: The only way to really get that answer for sure is to test. The best promotion format will depend on the product price point and how difficult it is to convey the value prop of the product(s). In addition to response, it is important to measure revenue. Everything else being equal, letter mailings tend to have a higher initial response but a much shorter life span than brochures or catalogs. It also depends on the time of year and the audience. I have effectively used inexpensive #10 letter campaigns to customers during times of the year when prospecting is ineffective. This approach is low risk, and the break-even point can be easily reached since customers are more responsive.

Stimolo: Message is important but so are the physical characteristics of the mailing. For example, a book sampling is a very expensive promotion format, yet many publishers find it is so effective that on a per order basis it performs better than less expensive formats. On the other hand, self-mailers are often the most inexpensive promotion formats, yet many companies cannot get a sufficient return from them to make them work. See answer to Q2 as well.

Winter: There is research, again, from the Direct Marketing Association available. John Caples
and Bob Stone (two direct mail thought leaders) also have some interesting data about mail packages and formats. The message, the list, the offer, the timing, and many other factors all impact the response rates you generate.

Can you explain more about qualifying schools based on demographic selections?

Winter: When you profile your current customers, you'll discover that many of them will share similar characteristics in terms of school or district enrollments, per-pupil spending, college-bound percentages, and a host of other factors. You can use these "demographics" to create a prospecting profile, so your lists include educators in schools and districts that match the demographics of your current buyers.

Stimolo: Not all schools are equal as potential buyers. The schools most likely to buy are the ones that have the most money. These include schools in affluent communities and private schools. Demographic selections such as Student Need Indicator, Household Income, and Poverty Level are key indicators of school affluence.

Would you say that a personalized direct mail piece or e-mail (personal name and title as opposed to just a title) is more effective than a non-personalized mailing (e.g., addressing on an envelope and salutation in a cover letter)?

Stimolo: Generally speaking, yes. However, the nature of the name selected has a bearing on the performance you can expect. For example, if we divide the names into three groups--new
names, direct mail buyers, and everyone else--the personalized mailing will perform better for the new names and the direct mail buyers. For everyone else, you may be better off with a title.

Winter: Yes, a personalized piece is generally more effective, because it has a better chance of actually reaching the mailbox and hands of an administrator, teacher, or librarian.

What would you consider a good percentage for school library retention?

Romans: I am not aware of any published average for selling to schools or school libraries. Typically about a third of customers will buy once and not again. If you have not measured retention previously, you should go back a couple of years to establish a baseline. Retention will also vary based on the type of product. School library budgets tend to be more stable and well funded so retention could be a key metric.

Winter: If you're talking about repeat buying (vs. renewal of a magazine or other subscription), 75%-80% retention would be excellent. Typically for subscription products, libraries tend to renew at a very high rate, generally 80% or more for many products.

Stimolo: One that is financially acceptable according to your business model. There are no magic numbers. The term "retention" suggests a renewal rate. A common "rule of thumb" number for a good renewal rate is 70%. However, some companies can be successful with renewal rates as low as 30% and others require 90% to make their business model work.

If by "retention" you mean response rate, there simply is no rule of thumb. Response rates are actually quite meaningless. Response rates are useful when you have a number of mailing cells all at the same promotion cost and delivering the same revenue per order. Then you can use response rates to distinguish the performance of one mailing cell over another. What is important is what the cost of the order is relative to the revenue that it delivers. So a number like cost per order, which is the number of orders divided into the total promotion cost, is a more meaningful measure of success.

Can I get the contact information for speakers so that I can speak directly to them?

Here's the contact information for the speakers: