Tips for Brokers

With times as tough as they are it is imperative that mailers are armed with the information and resources required for investing wisely in direct mail efforts.

As a publishing broker in the direct marketing industry, it is important for me to help my clients maximize their efforts while spending the least amount of money possible.

One way I can do this is to know what works for them and stick with that formula.

In our industry, we can measure success by response to an offer. However, even though we have numbers to work with, we need to know what to do with them and how to analyze them.

Numbers play a key role in direct mail. When 3-million pieces are mailed we can determine what percentage of recipients respond and the amount of revenue that is generated. We can see which lists have worked and what types of lists have worked. There are many ways we can utilize this data to maximize our efforts in the mail.

As a broker, clients often have us review results from prior mailings on both outside prospecting lists and house files. We analyze the data by putting it into simplistic categories and utilize that information to mail to a more targeted audience. Depending on the mailer and what they feel comfortable sharing, data can be broken down in a number of meaningful ways.

One thing we look to do first and foremost is code by type (the source, buyers, compiled, subscribers, members and so on) and category (men’s magazines, fashion/apparel, food/wine and so on).

These classifications are not generic, but rather are determined for each individual mailer depending on their markets. For example, a children’s mailer may use only lists where children are present in the household and, therefore, we need to get more specific with the categories. Was the list geared towards children’s apparel, children’s toys, children’s books or maybe children’s magazines?

However, if we are dealing with a magazine that targets a general audience, we may want to keep our types and categories broad, i.e. men’s magazines, technology, news/current events.

The key to an effective and meaningful analysis is statistically valid data. The more data, the more valid it is. If you create a classification that has only 5,000 names mailed, you should expect data that is statistically invalid. While I’m not a statistician, common sense tells me that your margin of error increases as the amount of data decreases. The ideal situation would provide you with a fair amount of data in each class, 100,000 plus names mailed per class and several different lists within that category.

A classification that contains just one list may skew your results, and you may either stop mailing to that category if that list performed poorly or mail more. If you happen to have a category that does not have statistically valid data, you should speak to this. A possibility would be to recommend testing further into the category to create a larger pool and reanalyze once you have more data. If this is not done, you may inappropriately recommend staying away from the category if it has not performed well or mailing deeper if it shows that the category has performed well.

Statistically valid data can also skew with the age of the data. Although a particular category worked well in 1995, you may or may not want to mail to it more aggressively in 2002. The best thing to do is to weight the data according to its’ age—newer data receives a higher weight and therefore represents what is happening now.

Along with the amount and age of data, knowing the type of data you are dealing with is important. Compiling data into spreadsheets is the easiest part of the analysis, but what makes it valuable and relevant is the human aspect – the thought process. For example, even if the table you are looking at shows that compiled lists perform less than average they may still work well for the mailer. Compiled lists have a relatively low CPM, and therefore, may be doing okay on a P/L level. The inverse can be said of files in the technology category—they traditionally have higher CPMs so response can be strong, but it is a poor performer on a P/L level.

Beyond types and categories, I like to hypothesize what may be driving response. Things like average age, average income, gender and average unit of sale have been known to drive response and profitability based on certain types of offers. Using information provided by list managers and other sources, each list in the client’s history can be segmented out in any one of these classifications to determine if the hypothesis is true.

I have created these analyses for several mailers that have age-sensitive offers. I will take the data they provide and break it down by age, both selected and inferred (this is information provided by list manager). For example, you can compare what happens when age was not selected but the inferred age is 35-45 (list A) and when this age group is selected (list B). If you find that there is no difference between the two and list “B” has an average age of 35-45, you may want to recommend testing list “B” without the age. You may find that you are not losing any response. If this is the case, you have created a larger universe and probably have contributed towards a positive P/L by eliminating the age select fee. Again, keep in mind that the data must be statistically valid to compare the two.

Offers sometimes change, and this could drastically affect the response. If at all possible, and if there is enough data, it may be feasible to have sub-categorizations of each offer. Using the same types and categories already determined you could break this data out by package mailed to see if some categories work better for different packages. You may find that you will maximize your efforts by mailing lists falling into certain categories for specific packages.

An interesting practice is comparing what the advertising department says about the magazines and what types of lists perform well in direct mail. For example, let’s say the media kit shows that 25% of your readers are male but you target females in direct mail campaigns. After looking at a category analysis based on gender, one can determine if gender influences response in direct mail. What performs well in direct mail may vary from your ad sales reported reader demographics. For example, you may find that demographics of your overall readership consist mostly of single men, but the ones that respond through direct mail are married men.

What effects do outside influences have on your direct mail campaign? If you’re dealing with a sports offer, does the anticipation of baseball season in February correlate with response? Does a family-oriented magazine have greater response when parents are expecting a baby in the next month? Does a technology-based magazine see a spike in response during one of the major computer conferences?

Using outside statistics can be a valuable tool when trying to figure out what is causing a trend in direct mail response. For example, mailer “A” is a parenting magazine geared towards new parents. By using the data on the U.S. Census, we can determine that most births occur during July, August and September. Because mailer “A” is targeting this market of new parents, is there a lift in response during this time?

After several consecutive prosperous years, our nation has now completed a year of uncertainty and recession. It has become vital that we maximize our opportunities in all aspects of business. As a broker, we can help our client’s bottom line by finding ways for them to mail smarter, keeping list costs low and revenues high. One way to do this is by analyzing data that is provided to us by the mailer. By exploiting positive trends and meandering away from the bad, we are ultimately helping our clients achieve this goal.

Jeremy Johnson is Vice President/List Brokerage at Specialists Marketing Services, Inc.’s Bedford, New Hampshire office. Jeremy started his career at Millard Group in 1997 as a summer intern and joined the company full time in 1999. By August of 2002 he was recognized by the Direct Marketing Association List and Database Council as recipient of the prestigious 2002 “Ron Davis Young List Professional of the Year Award” for his dedication and commitment to the direct marketing industry.

Jeremy recently joined Specialists Marketing Services and plays a key role in his clients’ ability to grow their direct mail efforts successfully. He works in publishing, business-to-business and fundraising, with an emphasis on circulation planning, select refinement, model optimization, and response analysis.

Jeremy has been published in various print and online versions of magazines such as: Direct Magazine, DM News, and Target Marketing. He has been a guest speaker for the DMA Young Professionals Series, DMA List Vision, Magazine Publishers of America and his alma mater, Lyndon State College, where he holds a degree in Business Administration.

Comments are closed.