Evaluating email marketing data

A quick look at email address data

Email marketing is just like direct mail, but online. The same principles apply when creating the content and you have the same amount of time (eff all!) to catch recipients’ attention before they press delete.

One of the biggest challenges, as always with direct response work, is to identify your target market and figure out how to reach them cost effectively.

I get a good few emailed offers for cialis, super cialis, viagra, penis extension stuff and replica watches every day. So – personally as well as from a marketing professional’s perspective – I’d like to slap the folk who blast this rubbish out, giving email marketing a bad name. Effective targeting’s the only way to go… or is it?

Cheap e-lists abound. I can buy a million email addresses for $99 to hit willy nilly with my book offer. Wow! But the data  will be completely random and I’d bet my last quid that it it’s neither opted in nor verified.

Another choice: I can buy a minimum of 10,000 verified current email addresses from a list of people who’ve ‘expressed an interest in’ or ‘clicked through’ similar campaigns’ for £180 per 1000. Or I can buy as little as 1000 current email addresses from a named opted-in list for just £95.

Let’s crunch some simple numbers

  • I can buy a good quality targeted list of 1000 email addresses and test the concept for £95. To cover the cost of the data I’d need to sell 39 items: that’s a response rate of 3.9%
  • I can buy a minimum of 10,000 supposedly decent quality records for £1800. To cover the cost I’d need to make 750 sales, a hefty 7.5% response
  • I can buy a whopping 1 million absolutely untargeted, unverified records for about fifty quid. To cover the cost of the data I’d need to sell 21 items. That’s a response rate of… tadaaaaaaa… 0.002%!

Do I do the decent thing?

I could test 1000 records from the best-targeted list for £95 with a 50/50 risk of not making the money back on sales. In this case, the total available records in the list doesn’t give me much room for roll-out so even a great response isn’t going to make me much more than a few quid. And a test of 1000 records isn’t going to give me statistically relevant
results so I’d learn nothing.

I could test 10,000 records from a slightly poorer quality list at a cost £1800 but that’s a crazy amount of money to spend on a test, especially with a break even response rate of 7.5%. Far too high.

Or I could join the spammers, irritating a million people for just £50 in the fairly sure hope of getting the 0.002% response I need to break even. Bugger me if this option isn’t the best as far as the numbers go. A list of a million gives rock solid statistically relevant results. And what’s more there are millions more of them where they came from, at a
stupidly low cost.

Talk about a moral dilemma.