This has come up in a few conversations on EmailRoundtable, and in a conversation between me and @LorenMcDonald, and I thought I’d put my thoughts here. I don’t like list rentals. But to elaborate, let’s talk about the various ways of (in)organically acquiring email addresses:
- For a fee, you use another company’s email systems to send your email. It’s on their system, but they send your content. All links in email go back to your site.
- Some companies sell their lists. So they actually hand over part of their customer base. You insert into your system and drop the email.
- Some companies share part of a newsletter with you, so you can insert a form, and acquire sign-ups.
- Some companies do back-end overlays of data models, to determine who in your company list, fits the model and is thus a good fit for some kind of segment.
- Some companies specialize in giving you extra data on your email list. So I have the email, they will tell me the email’s favorite flavor of ice cream.
I’ve avoided using the marketing terms for the above processes as that’s a completely different discussion.
Issues to think about if you consider any of these options
What’s the email’s provenance? How did the consumer give their permission? The minute you use that email, you are potentially a spammer, if you are unaware of how it was given. And check back a few generations.
Whatever route you take, will the customer understand the relationship? The email I used to opt-in to Zappos emails, and suddenly I’m getting Gap emails. Does that make sense? Don’t underestimate the consumer. They know how they interacted with your company. It doesn’t take a lot to be considered spam.
Are you giving over more value than what you’re being provided? If I give 100K emails to a datafarm, I need to understand that I’m providing them with value. They only exist by the customers they have, and the lists they get.
As Loren McDonald says very well in his post, If Someone Says Buy A List One More Time…”
After all, marketers who ask about buying lists could just be asking, “How can I build my list quickly, and where can I acquire email addresses?” Unfortunately, there is no easy way to build a good list quickly. If there were, presumably we’d all be doing it.
Here’s the truth: In the email world, you can’t buy legitimate email addresses. You know those $399 CDs with 50 million email addresses? Most of the addresses are probably harvested or gathered in some less-than-stellar manner. Many are probably either out of date, converted to “honeypots” by ISPs looking to trap some spammers, or otherwise undeliverable. The owners of those addresses certainly haven’t given you permission to email them.
There are many methods of increasing your email list, “organically,” a term I use just to say, it’s part of the normal process of business. It varies by company and organization, and it’s largely to do with getting out the word that you have interesting mailing campaigns, that you make it a priority to take email addresses at F2F events, strategic parts of your site, at the cash register, etc. Viral campaigns are great, and parternship marketing.
For clients who have explored the acquisition routes above, I have never seen one of them that has exhausted the organic methods. Lifecycle, “triggered” emails are probably the most unsung hero in acquisition channels. It enhances the relationship, it is targeted, and personalized, and 24×7. But it’s a little tricky to execute. I think marketers go the “easy” route by back-end data models, because it’s something they understand, whereas lifecycle emails are not one-hit-wonders but slow growth. Still, when you compare cost and response rates, lifecycles win every time. Web 2.0 companies understand this- their emails are short, text-only (or with maybe 1 image) and triggered according to user activity on their site. They notify you of social relationships- and they create a stickiness. Unfortunately retail and consumer goods haven’t launched onto this as much, they’re still in the image-heavy, HTML one-drop-a-week world, barely inching up from the “cart abandonment” email campaigns. They can go there, and some are trying, but it’s a hard row to hoe.
“We saw it coming because renting names from a list can be very risky for a marketer,” says Julie Katz, analyst at Forrester Research. “Those people don’t necessarily have any affinity with your brand. Also, if the names are bad, you could get caught in a spam trap and it can ruin your reputation.”
Bulk email lists: good or bad? by Mark Brownlow on Email Experience Reports
(By the way, I’m in Park City for Email Insider Summit, sponsored by the lovely folks at MailChimp. If you’re going to be there drop me a line or stop and say hey!)
Great post on Email Wars by Dylan Boyd, on How Agencies & ESPs Hurt Themselves, in which he is dismayed that more agencies consider “… ANYONE and EVERYONE with a pulse and a wallet [able to drive] email campaigns.” Also had a conversation with a colleague I’ve worked with for going on 7 years, on and off, about various places we’ve worked, and his assessment of the current situation: “Our email solution here has been so successful that management depends solely on email.” We both tutted that having such a money-making revenue channel ended up blinding them to other customer segments, methods of contact, and potential viable channels.
Combined with those two comments, I was explaining to someone how I did permission email marketing, and they were shocked that “…it costs money? Isn’t email free?” Again, the perception that good email marketing is free, easy and open to all (of all skill levels) is sadly not true. Don’t believe the hype.
Alongside the other comments, someone recently said (I wish I could remember who!) that “the low cost of email is our problem- that’s why the ROI is so poor.” Which does and doesn’t make sense to me. Yes, it means that we don’t make much money if it’s $.015 to the email address, and we’re sending out millions, but it does make money if you send out 150 transactional notifications and the click-through is 70%, and completed orders (for cart abandonment campiagns, for example) end up being product revenue X 105 orders, each day, minus a dollar in email costs. Even without any discounts or offers. So in a way, if it’s easy, you’re probably doing something wrong.
So in parting I just want to say, the perceived easiness of email is our problem. It brings in a lot of people who use the term “blast,” which alters consumer perception in regards to what is spam, and what isn’t. If grandpa is used to clicking “spam” button on emails he didn’t ask for, and clicks “spam” on ones he did, such as a gorgeous certified email from Eddie Bauer, despite their stellar reputation as a sender, customized content, user permissions, and top of the line, au courant policies and techniques, it makes it harder for all of us.
I’ll put in a screenshot so you see what I mean:
The part that alarms me: “Copy the message below into an email. Send the email to friends to invite them.”
Wow, so old-fashioned. Are they avoiding possible forward-to-a-friend (FTAF) send issues? Is this just old-school for old-school users? Or, is this just a way they can avoid the whole hassle of spawning bulk issues from their mail servers, and otherwise storing and managing email accounts, unsubscriptions, etc.? Perhaps they don’t want this to be thought of as a harvesting or acquisition ploy, and this is an indicator that the market in general is shifting away from FTAF kind of methods? Intriguingly old-fashioned.
The “invite a friend” aspect of social networks like Pownce, Plurk, and countless others has spawned a few “is this spam” fears, especially with CAN-SPAM laws getting refined. Perhaps GoodReads solves this problem by offering APIs into 4 major inboxes, and then this clunky manual version by “copying it into an email.”
At first, being an online marketer, I thought they were encoding the link to source their referrals- but then realized, no, it’s so the invitee can find the inviter on these social networks. Still, pretty clunky, and I wonder how many people are doing it this way.
More Reading: CAN-SPAM and Forward-to-A-Friend
Return Path’s “CAN-SPAM Rules Update: What Senders Need to Know” Scroll down for the peer-to-peer clause, basically you’re OK if there’s no inducement to forward.
More small type from Exact Target
Getting kind of addicted to LinkedIn’s Question & Answer area, and a couple of questions are around “why are my web2.0 emails considered spam.” I explored a little bit last post about why my email servers think Plurk invites are spam.
Invitational emails are that strange gray area of promotional and transactional- like “forward to a friend” functionality that some find spammy. Similar to “forward link” or “forward article” to a friend, many of these friends don’t appreciate their email being stored on NYTimes’ site forever, for example. Some sites explicitly say “we will not remember this email”- but otherwise it’s a little unclear, unless you read the fine print, exactly what a company will do with your friends’ emails.
That paragraph was to explain why precious Web2.0 emails are considered, at times, spammy. Another common reason: the site uses the source-friend’s email address in the “from”. This will set off problems with spam filters that look for a resolution between the From address and the real sender domain. If those don’t match up, it’s in the spam folder. So, don’t spoof the email to get automatically white-listed, it won’t work.
Use of “no-reply”- I don’t like this feature as it seems to take the position that the site can shoot out emails, but not handle any unsubscription/deletion requests, like it’s a one-directional firehose. I don’t know how SpamAssasin treats this, but I did notice the Plurk address was that way. What they should do is setup a simple unsubscription filter on the reply, like almost all GNU mailing list apps can handle, and a forward to customer service or some inbox that spits an autoresponder out “We got your email and we’re going to answer within X days,” with a list of helpful site link.
Plurk especially is missing out on a great opportunity to leverage the cutesy UI and design, by sending out these barebones text-only emails that aren’t even making it through spam filters (which really is the point of text-only transactional emails, in my mind.)
I know how these emails get built- been there!- an engineer (me) somewhere rolls out the new feature, “let’s let users invite friends by email!” and nobody checks the content of the email minus a cursory copywriter edit, and out it goes into production. Decisions like sender “real names,” domains, etc. are made on the fly by someone who can send themselves a message, so it must work! No monitoring on the systems, no ability to tell if the emails are going out, or how they’re being viewed. It’s a toughie, and not unique to any single system.
Update 6/14: I finally got one of my invites and it’s a simple text-only email. It did take forever and a day to get the darned thing. Wonder if they’re having scaling issues.
Looks like Plurk transactional email invites are considered spam by many webmail systems. I wonder why.
I noticed a couple of times this morning inviting people, that there was a caution: ”
“Spam issues: If your friends can’t find your Plurk invitation, please tell them to look into their spam / trash folder”
I sent 2 to other of my own email accounts, and haven’t received them. Anyone have one that I can use as an example for deconstructing exactly why these are considered spam? (And do Plurk some free work, ha!)
This is really not saying very good stuff about Plurk- and if you want your friends to use it, are they going to be that thrilled to check in the Spam folder? Otherwise the site has been a real delight. It’s a combination of Twitter with threaded messaging & a timeline.
A quick guess: badly formed HTML, strange subject line, and because it’s a social network, probably bombards networks and they don’t have good deliverability and sender reputation. But this is all guesswork until I see one!
I wrote about Political Spamming, and how our dearly elected circumvent and generally ignore CAN-SPAM regulations, and general email etiquette. Well lately the spam in my inbox is from public relations professionals. At issue:
- no reference to how they got your email
- strange subject lines “News item for ya, Anna” (???)
- largely irrelevant content – None of my blogs even remotely get close to this kind of content: celebrity diets.
- no unsubscription process – nor indicator on how they got my email.
- no reference to who they are, where they are from or on whose behalf they act – who is this from?
And the email in question:
At the Marketing Sherpa talk last February I sat at a table with a PR person who admitted that her industry does have the most heinous email etiquette. We joked for a bit about the Wired editor Chris Anderson (author of Long Tail) published a list of all of the PR spammers — as a kind of quid-pro-quo. I like this quote from Chris, regarding what press releases he does use:
Given that each one of those side projects is more narrow and geeky than the last, it’s a rare press release indeed that is focused enough to be relevant.
That’s what we focus on in email marketing: relevancy. So yes, it’s good PR vs. bad PR, it’s knowing the audience.
Just checked and that domain- popculturepr.com- is in Chris’ list. Sigh. Well, here are to the PR companies that don’t want to be considered spammers. Tips: warm up the cold call by…
- Doing some research & show your work. Do any of my posts seem at all interested in celebrities? If so, reference them in the email so you draw the connections for me.
- Work on the Subject Line. Use more words to describe who you are from: pop culture PR, and what blog this is in reference to. I write several!
- Provide an unsubscription. This can be your company’s list, or from some parent corporation. It ensures that you’re creating an opt-in list of journalists, and you won’t spam me, and generally “I asked for it.” It will raise your reputation with the journalist and set expectation.
- More sender branding. It’s simple to the point of being… spammy. Tell me where you’re from.
I was describing to my two sisters the merits of temporary email addresses over Memorial Day weekend. It’s like using a fake middle initial in your name when you buy things, then when you get catalogs in the mail with that initial, you know who sold your address (always an interesting experiment!).
Web email now supports sub addresses (also called “plus addressing”), from back in the day with sendmail. Yahoo & Google make it possible for you to extend your email address in various ways so you can track how your email gets picked up. Next time you register for something, use “email@example.com” instead of “firstname.lastname@example.org”. If you end up on some random mailing list with that to: address, Pottery Barn has sold your list. I should have done this when I registered to vote! Yahoo has a more convoluted method, called “disposable emails,” using AddressGuard, that probably won’t be embraced by non-Geeks.
If temporary email addresses ever get widespread, I can see consumers using it to further identify and opt-in to communication that they want. If they’re getting a Pottery Barn email, and their To Address is “username+potterybarn” they are going to be more likely to access and view. They could inclusively filter select mailers into folders based on the “To” address.
I see on message boards that the tendency is for marketers to filter out anything with “+” in it, as most folks who are currently using this are using it as a manual spam filter, and to weed out the bad promotional lists that aren’t using opt-in processes, or loosely define opt-in. I would like to see some usage metrics on this- if more people are using the “+” sign as way to narrowly filter their email, or if it’s just a geek thing.
Disposable addresses of this form, using various separators between the base name and tag are supported by several email services, including Runbox (plus and minus), Google Mail (plus), Yahoo! Mail Plus (minus), and FastMail (plus).
Great post by Laura at Word to the Wise- wish I’d read it before I posted! Disposable or Temporary Addresses
I have two friends who do political email, so I don’t mean to slam the entire industry, I just want to point out that recently, as a SF voter with the Democratic Party, I’ve been the victim of a lot of unwarranted email.
Here is my story (to the sound effects of Law & Order).
I made the mistake of putting down my email (which is optional) last time I voted. Adding my email to my voter registration information was eye-opening, because in the following months, I got about 5 emails from Democratic Party candidates, and multiple emails per candidate. Doing some research and replying to the emails, I asked, “where did you get my email” and they replied with conflicting stories of: “Public records,” and “Don’t worry, these emails aren’t public.” I kid you not.
When I pointed the various CAN-SPAM compliant issues with the emails, one guy pointed out, “CAN-SPAM was created for commercial emails.” The implicit understanding here that he can disregard them because it doesn’t apply, reminded me of the issues lately with CAN SPAM – if you’re concerned about the new rules, you have some bigger problems with your email marketing campaigns, because they’re basic courtesy to your audience, not random government regulations. Ironic that a politician is seeking to avoid regulations…
My beef with his (and the other 4) emails; : no physical address with which to follow-up on unsubscription, no one-click unsubscription method, no “we got your email from here” mention, and a host of other general bad marketing elements. The most notably: making me even ask “where did you get my email.” See the end of this post for recommendations to political campaigners.
I’m not mentioning the candidates because I don’t want to give them press for doing something badly! If you have any examples, though, I’d love hear it.
Some research on the state of these email databases of registration info, and how it varies by state, and email collection at voter registration:
- Great (if old) article on Wired.com
One of our own email marketers tackles the same issue: ClickZ on Political Campaigns, Known Spammers & CAN-SPAM. Jeanne’s a lot nicer than me- just because we haven’t regulated politicians doesn’t mean they can email without a method of unsubscription!
- Initiative to collect political spam on Greg Dewar’s site.
So, in reading up on California voter info, here is a sign-up form from the California Voter Foundation. In trying to find how they’d use my email, I saw this somewhat confusing paragraph:
The voter registration forms ask for basic information, including your name, street address, mailing address (if different), birthdate, the county in which you reside, and your place of birth (U.S. state or foreign country). The form also asks for your driver’s license number, email address and phone number; although you are not required to provide this information, a phone number will help your county clerk contact you if there are any problems with your form. For information about choosing a political party affiliation, see the section above.
The Wired article refers to the confusing nature of the email collection- no explanation of how it’s going to be used, along with the “place of birth.” (at least not your mother’s maiden name like some states!) For personal identity issues alone, I’d be wary of entering this stuff in a databank that is sent out to candidates.
In general, why do politicians get such lax rules? Did the crafters of CAN-SPAM actually think that candidates would (or could) ethically deal with this? The examples in my inbox show something different. I would setup a best practices for political emailing as:
- say early in the paragraph, where you got the email form
- provide a direct one-click link to unsubscribe from the candidates’ email list
- provide a real address for follow-up questions
- guarantee a one week turnaround on unsubscription requests
- do not harvest more information in the unsubscription process
Nice to see Facebook get down and dirty on controlling the wild spam that goes on with their third party (and yes sometimes core) applications. Check out the FaceBook blog for the actions they’re taking, namely:
Applications must now give you advanced warning if you’ll need to invite friends to get information or access content. So you should always know ahead of time if that quiz you’re taking will require you to invite friends to see your results. If you see applications withholding content without warning, go to that application’s About page to report it.
I think the key challenge these highly communicative, transactional systems have, is how to leverage the interconnectedness without exhausting their base. Opening up the community to third party applications, free to develop and install, is great! But with little control to protect the consumer’s right of contact, a free-for-all happens that is just ugly. One of my friends calls Twitter “death by croutons”- and I think that’s the death nell of any Web 2.0 service, meaningless inundation of messages.
First, let the user set their preferences, and like they say on the blog, let the user be warned before diving into an application if it’s going to require sharing.
Is FaceBook changing the definition of spam? I get invited to install this application from a friend, I block it, as I’m tired of getting these, then FaceBook pops up a message: “we will flag this application as spam.” But is an application, by definition, spamming based on another user’s invitation? Just like “forward to a friend,” essentially, which some folks do define as spammy. I was going to write about NetFlix’s application, which has me in a daily reminder spam loop on a quiz I’d created- but they managed to fix it while I was reporting the error.
The solution, to me, is to setup conservative contact preferences and allow the user to expand them if they want. An open “share all” preference should not be the default, or required.
To my knowledge- and I don’t know spammers individually, I’ve just pieced this together from the defensive manoevers- spammers spam because there are people who respond to spam, hard as that is to believe. And they manage to keep their costs almost non-existent, illegally. Stealing server accounts, bandwidth, you name it. They also don’t really hold any data- it’s all done on the fly. So corporations will store interactions with customers, whereas spammers randomly generate usernames/passwords/email addresses and just throw it out there and see what sticks. Also, they have a trenchcoat industry. There is no desire to maintain a nice reputation, brand identity, or customer lifecycle, as good corporations do. Also, it’s like a pyramid scheme- as long as they convince someone else that eventually someone will buy, they sell the “starter kit.” See link for some folks who are taking it seriously: FTC Launches Crackdown on Deceptive Junk E-mail
Corporate email marketers, especially the ones from very large companies, are usually very conservative in their email drops. Because it costs money to do it well. To craft a good email message, to extend the lifecycle of the customer, to collect feedback data, and to re-integrate it into another engaging communication, it’s a lengthy and consuming process, that is great when done well, but can be easily destroyed by the wrong email. So in these email marketing groups you will actually see a large base of customers who want communication, but aren’t getting it because they haven’t demonstrated anything recently to show they are active. That’s right- corporations aren’t even emailing all the people they could.
It essentially comes down to the fact that what spammers are usually selling is embarrassing “trenchcoat economy” things- they don’t need to maintain a relationship with the customer after the initial purchase. Most of them actively don’t want to know you- as they just stole your PayPal password, for example. I worked briefly for a company that had a client that was an adult toy company. My excuse was that they didn’t disclose, ahem, all of their clients in the interview (lesson learned!). Doing QA for a them was a joke. There is no quality control, no costs really. You can bring down the server for a day and nobody will complain. So spammers have such bare bone costs because they certainly don’t care about maintaining and creating the same customer relationship as legitimate companies. Here today, gone tomorrow.
There is a tiny bit of guilt that lies with the consumer. This would not go on if folks didn’t buy from them. No one is altruistically sending out millions of emails on Viagra. They’re doing it to make money- so there are people out there buying Viagra. (I’m going to get tons of hits for saying Viagra now, what twice now) At one client, I did an analysis on the response to an email that was a very bland, text-only policy change message that had one tiny call to action after a lot of legalese. The percentage of people who purchased off of that email message, that were opted-out – “don’t send me email”- was 60% of the customers. FYI, the reason we even emailed opted out customers is that kind this kind of service message does not require opt-in permission, as stated in CANSPAM.
This is probably why I’m a fan of AOL’s certification for email- the more expensive it becomes for bulk emailers, the better the distinction between spammers and legitimate companies.
Background checks on new users and customers
Since spammers are frequently kicked off the network, they are constantly trying to create new accounts. Many spammers are able to make even a few hours profitable for them and can cause many days of damage to reputation of the services they abused. As a result, many ISPs and web -email providers use CAPTCHAs on new accounts and try to verify the credit cards are not stolen before accepting new customers, check the Spamhaus Project