E-appending done right

With postage and paper costs on a never ending march upward, e-mail appending — in which a third party matches permission-based e-mail names to a merchant’s postal address — is becoming an increasingly enticing option for merchants who sell by catalog.

But there are downsides — pitfalls so big, in fact, that the word “pitfall” seems a little weak in describing them.

For one thing, there are a slew of shady e-mail appending operations. Horror stories — merchants paying good money for a list so filthy it is unusable, for example — are legion in online marketing circles.

Also, an e-mail append done badly can seriously damage a merchant’s brand, reputation and file. For example, if enough people hit the “report spam” button — 0.5% of recipients to a specific mailing, according to conventional wisdom — Internet service providers will begin treating all the merchant’s messages as such and the marketer will start experiencing delivery troubles.

Translation: If the people holding those newly acquired addresses call the merchant a spammer, that merchant will have trouble communicating by e-mail with its entire file.

But appending can be a great way to grow an e-mail list and spur postal customers into multichannel buying if done right. It simply takes some careful planning and a good vendor.

“Appending is absolutely not for everyone,” says Austin Bliss, chief executive of e-mail appending firm FreshAddress. “We frequently turn prospects away, but if you decide it is appropriate, then vendor selection is very important because not every vendor follows the correct process.”

Founded in 1999 as an e-mail change-of-address company, FreshAddress began offering appending services in 2001. The firm claims it has 385 million e-mail addresses of consumers who have opted to receive offers from third parties.

The first step to a successful e-mail append is to select postal names from the files that are active. This way, the effort has a greater chance of reaching customers who are likely to respond positively.

“It should be a fairly recent customer list,” says Bliss. “You don’t want to dust off customers who haven’t participated with your brand in eight years.”

According to Bliss, there are vendors who will append names from telephone books. Avoid them. “That’s one place where appending can go horrifically wrong,” he says.

Rick Buck, director of privacy and ISP relations for e-mail service provider e-Dialog, adds that what constitutes an active address worth an append attempt will vary from merchant to merchant.

“You’ve just got to put some method behind the effort that helps determine there’s some level of engagement, and that this person’s worth e-mailing,” he says.

Once the merchant has thoughtfully identified postal names that are good targets for a possible append, the next decision is which type of append to do. In an individual append, the vendor tries to identify the exact person holding the address, while with a household append, the address of anyone in the household of the postal address will do.

Not surprisingly, household appends yield higher results. Which road to take depends on the marketer’s relationship with address holders. “If your relationship is with the household — let’s say you’re a non-profit and the whole household participates — then a household append is really appropriate,” says Bliss.

But if you’re, say, Victoria’s Secret, “my guess is a household append isn’t as appropriate,” Bliss says. “You’ve got to think about who in that household is really going to buy.”

Bliss added that the vendor should also suppress against unsubscribes the merchant has already received and the Federal Communications Commission’s do-not e-mail list of wireless domains. This is a list the FCC publishes under the Can Spam Act to which marketers may not send unsolicited e-mail messages because they go to wireless devices.

(The list can be found at http://www.fcc.gov/cgb/policy/DomainNameDownload.html.)

“If they’re a Direct Marketing Association member, they may also want to suppress against the DMA’s e-mail preference service,” says Bliss. “A careful vendor is going to talk to the client about suppression.”

Asking for permission

Next comes the permission message, which is typically deployed by the vendor on the merchant’s behalf. The address holder is notified that the merchant wants to add him or her to its e-mail program.

The purpose of the permission message is twofold: To make sure the address is deliverable and to make sure the address holder wants to receive e-mail from that merchant. “Maybe they just want to get the merchant’s catalogs and they don’t want e-mail messages,” says Bliss.

“It’s also possible that the e-mail append will hit the recipient at an inappropriate address,” he adds. As a result, a good vendor will enable recipients to respond positively to the initial permission message with an address-change request, says Bliss.

One mistake marketers commonly make — or at least attempt to make until a sales rep talks them out of it — is to include an offer in the first e-mail they send to newly appended addresses. “Those are very good marketing instincts at work, but they’re not appropriate here,” Bliss says.

“An append permission message should be explicitly that,” he continues. “It needs to be a clean message simply stating: ‘Hey, we’re about to add you to our e-mail communications. Here is why we think you’ll be happy about it, but if you’re not going to be happy about it, please opt out now.’”

So what kind of match rates can marketers expect from an e-mail append? According to Bliss, this is like asking what a good response rate is.

“In April, we had match rates ranging from 3.6% to 23.5%, and that’s April,” he says. “That’s why I shudder when I hear customers shop based on promised match rates.

Ultimately your vendor should be focusing on quality, not quantity, Bliss notes. “You want as many addresses as you can get, but you want as many as you can get within the constraints of them being good addresses and your customers.”

Appending costs can range from 15 cents to 50 cents per match, says Bliss. “I’ve heard of prices as low as 2 cents a match,” he adds. “But one way or another, the vendor is going to have to make money, and [in the case of 2 cents a name] it’s not going to be on quality. Pricing is a pretty good indicator of quality.”

Once the vendor has made what the merchant hopes will be quality matches, it’s time for the vendor to hand the addresses over to the marketer. What happens next is crucial.

“The vendor really has an obligation to hold your hand when they return the file,” says Bliss. “First, it’s imperative to get all the unsubscribes back, and I’ve seen people ignore this. These are people who have clicked on a link and said ‘I love your stores. I don’t want your e-mails.’ ”

Bliss also recommends eyeballing the addresses to give them what he calls “the sniff test.” He says that simply by looking at a segment of the file, a garbage list can often be spotted.

“Marketers will often say: ‘I can’t open a file that big.’ ” he says. “You can always go to your tech person and ask for an Excel spreadsheet with the first 50 rows. I have seen the results of bad appends. If you see it, you’ll know right away.”

He says bad appends will contain e-mail addresses that are clearly not those of customers — John Smith’s postal address results in [email protected] or [email protected], for example.

It’s also important not to simply add newly appended e-mail addresses into the marketer’s ongoing mail stream. For one thing, it’s difficult to tell sloppy appenders from clean ones, says e-Dialog’s Buck

And even the clean ones — by the nature of their business — often mail from IP addresses that have spammy reputations with the ISPs and don’t have the best delivery rates in the world. As a result, many of the people on the file the vendor gives back to the merchant may not have seen the initial permission message.

“They may not have opted out because they never saw the message, not because they didn’t want to opt out,” says Buck. “One of the first things you should do is send a welcome message to these people.”

Buck also warns that an appended file will probably contain spam traps, or deliverable but dead e-mail addresses ISPs use to help identify spammers. It is impossible to tell if an address is a trap.

As a result, a confirmed-opt-in process, in which the marketer sends a series of permission e-mails to the file telling recipients that if they do nothing, they’ll be removed, is a foolproof way to guard against spam traps.

One thing a marketer must think about when appending is why people would want to hear from the merchant via e-mail, says Bliss. “Then you’ve got to explain that to them,” he adds. “You’ve got to send a dedicated message that tells them why they’re hearing from you and why they should be happy about it.”

Buck advises putting a master plan together: “Why do you want to talk to these people? What are you going to say to them? Relevance is always important with e-mail, but relevance in the first engagement with an appended address is highly important. You have one chance to make a first impression. This is it.”

As for when not to append, Bliss says it is not a way to start a new e-mail marketing program. “Don’t do an append unless you already know how to make money with e-mail,” he says.

Adds Buck: “If you do it right, appending can be a great way to grow your database. If you don’t do it right, it can really cause you some pain.”


The Museum of Fine Arts, Boston, four years ago determined that its catalog unit — which had been mailing 9 million books a year — wasn’t profitable enough. So in 2003 it killed the catalog and its online store.

The museum reopened its Website 18 months later. By then, however, its e-mail list had gone stale. So it began to use its retail operations and Website to collect e-mail addresses to build the file back up.

“But you only get so many that way,” says Ellen Bragalone, director of retail operations. The museum still had a postal list of 435,000 prior customers, however. So in 2007, it decided to try a household e-mail append.

Why a household append? “A lot of our members are household memberships,” says Bragalone. Plus the museum’s main demographic is women in their late 50s, who tend to be long-time residents of the same home.

“A lot of our customers are kind of settled. They’re not college students who are changing apartments, so it’s easier to find them,” says Bragalone.

“Plus, we have a loyal customer base. I still to this day get handwritten letters asking why we’ve taken people off our catalog list.”

The append resulted in 55,000 names, according to Bragalone. “This was very good for us because our e-mail list had dwindled down to about 60,000 or so. So this really gave us a boost.”

By comparison, during the six months the museum did the append, it got about 1,200 names through the shops and about 8,500 through the site. Four months later, the museum appended its remaining 380,000 postal addresses and got back 30,000 names.

“In four months, we got back 85,000 names,” says Bragalone. She says the purchasing records from the old catalog no longer exist, so she can’t benchmark these names against catalog buying behavior.

“What I can tell you is we’re seeing people who haven’t purchased from us in four and five years coming back,” she says.