Make Co-Ops Work to Industry AdvantageIt would be grossly unfair to say that cooperative databases are not meeting mailers' expectations. Still, one would be hiding his head in the sand not to recognize issues - such as rising costs and dropping response rates - that affect many, if not all, of the current databases.
Co-op databases have delivered good prospect names and helped with customer reactivation, so they are useful, even crucial, to our industry. It does not mean that they are perfect. Some practices of the co-op databases may even be harmful. In other cases, they may be less than optimal.
For those less familiar with co-ops, let me briefly describe the concept: Mailers contribute their most valuable asset - their house file (in most cases it includes their transactional data) - to the cooperative. In exchange, they are entitled to withdraw prospect names from the database at favorable pricing and from a "modeled" environment that promises a lift in response.
Co-ops have grown to play a significant role in many mailers' circulation plans, often making up as much as 25 percent to 50 percent of their prospecting efforts.
There are six major cooperative databases focused on the catalog/Internet retailing industry. The Abacus Alliance is the largest, and Experian's Z-24, I-Behavior, Prefer Network and NextAction are all factors in the industry. Wiland Direct is the most recent, and it appears engineered to solve many of the recent problems that mailers have experienced with the traditional co-ops. Time will tell, but I must admit that chairman/founder Phil Wiland has a unique perspective on the business coming from the mailer side.
I think it's time that catalogers, direct retailers, list brokers, list managers and service bureaus join in an effort to strengthen the industry and make co-op databases all they can be. Together we can improve profits across the industry by adopting the right policies and practices. For too long, co-ops have sought to minimize the involvement of traditional list brokers, hoping to control the entire process to their advantage, often at the mailer's expense. I believe appropriate reform can add basis points to the bottom line of many direct merchants.
As the scope and penetration of co-ops continue to increase, it is progressively more important that direct marketers be familiar with key issues involving co-ops. Mailers must understand and influence the manner in which the databases use contributed data. This article aims to raise a number of issues (sadly, not all), briefly discuss pros and cons and stimulate industry discussion of co-op databases and how they should evolve.
Such discussion should begin with one fact: Co-ops are here to stay because they make so much fundamental sense. The databases use more information than is available on vertical lists. As a result, they identify otherwise invisible list universes that work. Thus, it isn't a question of whether we should have co-ops or how many such databases we need. The question is how co-ops should operate to the industry's advantage.
"Industry advantage" is easy to endorse in concept, but tough when one gets down to the details. For example, is it advantageous for co-ops to serve both traditional direct retailers and companies that are mainly brick-and-mortar retailers? The answer may seem to be an obvious "yes," but what if the retailer has few direct response buyers to contribute yet wants to mail in huge volume to hotline catalog buyers, including those of catalogs that sell similar merchandise? Clearly, the database is well served (it has more revenue). The retailer is well served (it gets access to great names). But are the catalogers well served?
Important questions such as this are answered frequently by the co-ops, sometimes with little disclosure to contributors. Here are several questions that need to be answered to industry advantage:
Should there be a limit on how many prospect names a mailer can use from a co-op database?
Pro: Yes, and it should be based on the number of 12-month direct response buyer names supplied. Without a limit, small catalogs can benefit unfairly by using an unlimited number of names contributed by larger companies. Furthermore, with no limit, it's possible for a very large company with a huge budget to mail large numbers of names, even if its direct retailing business is very small. No limit is unfair.
Con: No, all the names are available on some list, anyway. And the huge mailers are not really competitive with the smaller, niche catalogs.
Should co-ops rent catalog/Internet names only to other catalog/Internet retailers rather than being allowed to rent buyer names to companies that are not direct retailers?
Pro: It makes no sense for co-ops to make money renting buyer names to insurance companies, banks and others whose data are likely of little or no value to catalogers. If they do rent outside the catalog/Internet retailing community, there should be a revenue share.
Con: It doesn't hurt anything for the names to be used in this way. Insurance companies, banks and others will get the names somewhere. And if they provide any data, even if of marginal value, that's good for the retailers whose buyer files are used in this way.
Should catalog/e-tailers be precluded from participation unless they contribute their transaction dates and dollar amounts?
Pro: Giving participants access to recency and dollar amount info when they don't contribute such data themselves simply isn't fair. We need a level playing field.
Con: If they think having just names and addresses will enrich the database, it's fine for such contributors to have full access to the data provided by others. Knowing that someone once purchased is enough information.
Should promotion history information be contributed to the co-ops?
Pro: We need the databases to bring contact frequency strategy down to the individual level so that best customers and prospects can be promoted frequently while marginal prospects and customers are promoted less often. Done right, this can make marginal names profitable and put money on the bottom line. Relative to prospect promotion history, permission from list owners is obviously required. The industry should cooperate to facilitate such permission.
Con: So many lists are on the market that the names will get mailed anyway. It isn't worth the time and effort to keep track of promo history. Also, obtaining list owner permission is too difficult.
Should co-ops grant a net name arrangement?
Pro: The industry needs competition among numerous co-ops. Competition fosters better value and technological innovation. But it also means that duplication is inevitable, so a fair net name arrangement is fully appropriate.
Con: The price is low to begin with. Besides, you can use balance models to reduce duplication by having your merge/purge net file sent to your favorite co-op.
Should a model score from a co-op database be available as an enhancement to other lists, including non-merchandise lists, for purposes such as outside list optimization?
Pro: It's fully appropriate for a co-op to ship names with a model score (no participant data) to a merge/purge company for inclusion in the merge/purge process to enhance/optimize other lists. If this can help marginal vertical lists perform well, it makes sense. But data beyond model score should not be used to enhance other lists. Furthermore, such usage should be restricted to participants, not available to just anyone.
Con: List enhancement is not an appropriate use of co-op database information, even if it is just a model score.
Should co-ops cooperate actively with a participant's list brokers and list managers?
Pro: Optimizing circulation strategy is the main objective. The marketing staff of the direct retailers, list brokers, list managers and co-ops should work as a team to deliver optimum results for the direct marketer. To have them at odds with each other is illogical and counterproductive.
Con: The co-ops don't need input from list brokers or list managers. Brokers and managers should stick to vertical lists.
Should databases disclose specific information about what factors are important in a model?
Pro: The black box mentality should end. The DMers own the data. No information specific to a participant should be disclosed, but categorical and general factors should be included in each model report. Participants and their list advisers should be able to glean useful marketing intelligence from every model built.
Con: Their techniques and findings are proprietary, and they should not be expected to disclose them.
Should participants contribute SKU-level data to co-ops?
Pro: This is a valuable way to find segments of names that will work that can't be found any other way. Provision of such information should not be required. Participants who don't provide it should not have access to product level data supplied by others.
Con: If a co-op contains hundreds of lists, SKU-level information will add little or no predictive value.
Should participants contribute their data, and co-ops track data, in a manner that optimizes understanding of marketing and order channels?
Pro: It's crucial that legacy systems be abandoned in favor of new systems built to shed maximum light on our multichannel marketing environment. Co-op participants need to revise their systems for providing data in a manner that segregates orders by marketing channel (what produced the order) and order channel (how the customer ordered). Co-ops need to provide low-cost match-back services to help marketers make better segment decisions. They also need to optimize their modeling techniques to address multichannel issues.
Con: A good customer is a good customer, no matter whether they order online, by phone, by mail or buy at a store.
If you answered "pro" to most of these questions, then you likely agree that the co-operative database world needs reform because some companies are operating closer to "con." It's probably impossible to be certain how the profits and growth of a catalog/Internet retailer are affected by differing approaches by co-ops, but the differences unquestionably are significant.
Some of my clients get good results from several co-ops. Others have experienced falloff in results from some of the cooperatives. Personally, I think the newer entries to the field may have better technology and operating practices. At the least, the new players recognize the need to bring greater value to the cooperative proposition. A good-for-the-industry approach is what we need from all the databases.
Here's my bottom line: Join every co-op with policies that you think operate to the industry's advantage. Give each database you join several mail cycles to achieve good performance. If a database does not produce good results after adequate testing, withdraw. If your results are weak, or bouncing up and down, with a co-op that you have used for a long time, get out and try the new ones.