When data is licenced for use, both parties are acting in good faith. But just as good fences make good neighbours, so seeding data keeps brands up to the mark and gives owners proof they are being paid what they should be, says David Reed
“There is a view that everybody knows how to do seeding and do it properly and that it is bullet-proof. That is not true.”
Imagine you are a retailer selling produce by weight. All of your customers pay you the exact price for what they buy. But one of them secretly takes an extra kilo of goods while you are not looking. Unless you catch them in the act, it is possible you might never notice or find out it is happening.
Now consider that what you are actually selling is data, provided on a single-use licence priced by volume. The payments you receive from customers match the invoices you send out, so everything seems to be working well. But what if one or two of those customers is using your data for a second time without paying you? Would you notice and be able to prove what they are doing? If not, just how much money could you be losing out on through this unauthorised activity?
Seeding is the way data owners generally keep track of what their customers are doing. In principle, it is an easy way to monitor end users and compare their activity against licence terms. In practice, seeding techniques are not being kept up-to-date by many companies and may have even been cracked by unscrupulous users. Yet a new seeding service from DQM Group offers not only a reliable approach to seeding, but also a closed-loop royalty management process as well.
Lynn Stevens, managing director at Lloyd James, notes that: “All of our clients use seeding. We offer the service based on a simple system of friends and family around the country. We add them to clients’ data using unique initials each time.”
For data owners selling files to direct mailers, adding in a small set of seeds in this way provides a straightforward tracking method. It also means that the items received by names on the list can be compared against the sample provided by the marketer ahead of the campaign. Back-checking in this way is an important part of the work done by list managers in the marketplace, even if it remains largely unremarked upon.
“It is part of our order process, not something we discuss at the start. We are trying to control what happens and control how people are using the information,” says Stevens. Both direct mail and email files can be controlled using a simple technique like this, but telemarketing is more difficult, so few management companies offer a seeding service in this channel.
Stevens argues that seeding can be used on a simple basis like this because the data industry is largely self-regulating and well-behaved. “It is amazing the number of times we have seen re-use over the years - I can count them on one hand. It is not commonplace because marketers are aware that data is seeded,” she says.
However, Stevens does sound one warning note. “If data owners were complacent, that could affect the perception among customers and we could see more misuse,” she says. Evidence that this has already happened can be found, albeit anecdotally.
You do not have to dig around in the data industry for very long before you will hear a commonly-traded story about a broker who has cracked the seeding process used by one very large data owner. Apparently, the seed set embedded into its targeting universe has not been changed for many years. Through careful matching and comparison, this broker has been able to identify all of the seeds, remove them, and is happily reselling the data on his own account. With no tracking mechanism in place, the data owner is either unaware of this misuse of intellectual property or is unable to prove the breach.
Rosemary Smith argues that it is data owners, rather than users, who have a perception that seeding is keeping everybody honest. “There is a view that everybody knows how to do seeding and do it properly and that it is bullet-proof. That is not true,” she says. “I am amazed when I talk to some data owners and quickly identify that they don’t apply seeds properly.”
Data owners bringing a file to market for the first time may be forgiven for not knowing about the seeding process. Long-established providers, especially where it is their core business, have little excuse for not seeding.
“To me it is a no-brainer,” says Smith. By introducing seeds into a file, its owner can track what has been done with that data and whether the outgoing messages conform to what was expected. “Seeding can show you if there was any switching. You may have received a sample pack that ticks all the boxes of the Code of Practice, then it gets switched to an illegal lottery. That is what used to be done,” she warns.
Although overseas lotteries are no longer the heavy direct mailers they once were, there is always the risk that a file could be sent an inappropriate message. Seeding not only tracks actual usage, it provides real evidence of any abuse.
“Years ago I had a case taken to court on the basis of seed names showing misuse. They settled out of court for a significant amount of money. When we went to get advice from counsel, those seeds were seen as absolute proof,” says Smith.
Gauging the actual level of misuse of data from outside is difficult, not least because there have been very few court cases. Where seeds exist, the abuser will nearly always settle. This absence of high-profile legal punishment could be leading to the complacency that Stevens warns against.
In other jurisdictions, it is the penalties available under Codes of Practice that help to prevent abuse. Germany’s direct marketing association applies a penalty for misuse of ten-times the cost of the original data. That is a serious sanction which brooks little argument.
In the UK, it may be new solutions that provide a more holistic approach to seeding and royalty management that bring data owners up to the mark. Peter Galdies, director of DQM Group, says that using elements of its new Secure Distribution Hub will transform the way data is tracked and controlled. “It may be that an organisation places a few its staff into a file as seeds. That shows no real understanding of the the process. For example, reconciling what is received against what was planned takes effort, especially to compare it against agreed terms,” he says.
The Hub service incorporates correct formatting of seeds as well as terms and conditions of data contracts. So if the rental is for targets in Aberdeen, seeds will not be London-based - an obvious difference that would allow the unscrupulous to remove those names. Equally, the service can identify if other aspects of the campaign are within agreed terms.
Galdies believes that abuse of data outside of licence is in the order of 10 to 15 per cent. While this may be largely accidental - such as multiple, rather than single use because of a failure to flag correctly - it still means lost revenue for the data owner. “Almost every day we have been able to prove through the seeding process that things go wrong,” he says.
Without the right management process, data owners really have no idea whether their goods are being used and paid for properly. Good seeding gives them that insight and control. As Smith says: “It’s like micro-chipping a dog. Seeding is chipping for data.”