Email marketing tools that allow you to (almost literally) travel in time

I've discussed so far some of the most elementary topics in email marketing. In my next few posts, I want to introduce you to some of the most advanced functionalities in the business: robots and time-travel. Algorithms and bots will probably require more than one installment, so I will address those subsequently. First, let's hop in that DeLorean and break the space-time continuum, Doc!

Time travel in the past

There are several value-add technologies you can lump on top of your ESP or CRM. The fine folks at Movable Ink came up with a few great features that blew my mind when they pitched them to me.

For example, their technology allows you to send a few different pieces of content, just like a regular email tool, except Movable Ink allows you to test all the different things you want to test at the same time and it doesn't require you to monitor the results of your A/B tests. The concept is that you only press "send" once in your entire mailing process at what would be your initial A/B test stage. The technology begins sending your mass mailing distributing all of your variations evenly, and it eventually declares one of them the winner according to a threshold you can select and adjust. For example, you can tell the software to declare a particular piece of content the winner once it gets two percent more opens than the second best-performing content. The technology then automatically sends all of the remaining emails using only the winning content. No need for a "send to remainder" step.

So far, what I've described is hardly science-fiction. Things get really funky, though, when Movable Ink goes back into the inboxes of those addresses it sent a "losing" creative or bit of copy or CTA and changes it back to the winner of the A/B test (so long as the recipient hasn't yet opened the mail.) Retroactively changing something you already sent to someone (with the same timestamp!) is tantamount to time travel to me.

(One word of caution: always be careful about declaring anything the winner of a particular A/B test too soon. Email metrics often evolve chaotically over time relative to one another, and your ostensible victor one hour after the launch can end up being the worst performing content by far. It happened to me many times.)

On top of being really neat, the above-described method limits the loss of business derived from sending thousands of users a bit of content that performed more poorly than the one chosen for the send to remainder, as would be the case every time any one of us runs an A/B test before sending to the full segment. For example, at Change.org I would make sure I sent about 4 subject lines to samples of at least 8,000 users for each line. This method allowed me to be sure that my six-figure remainders would receive a subject line which had passed scientifically robust tests, and whose results were very predictable. However, at the end of the day, that still meant that for my average campaign, 24,000 people received a subject line which performed less well. I left hundreds of signatures on the table, every time, by definition. This is A/B testing's Achille's Heel.

Movable Ink can also deep-link some background elements in their clients' emails, such as outdoor scenery, and makes them dynamically react to data like recipients' local weather conditions. To make up an example, you could open an email from Jaguar and if it's raining outside you might see a picture of a Jaguar navigating in tough weather conditions and a message next to it praising the directional stability and security of the car. If you were to open it again later and it's suddenly nice outside, you'll be shown a picture of the same Jaguar drifting on an open road somewhere in the country with a message highlighting its comfort and class. How cool is that? The power of deep-liking objects, in the right hands, can be wielded in incredibly effective ways.

Time travel in the future

Alchemy Worx has developed a massive database compiling the thousands and thousands of A/B tests (and sends to remainder) sent by Alchemy Worx' corporate clients over the years.

From this, they built a tool called Touchstone, which allows you to predict what your key metrics are going to be for a given send before you actually send out the mailing. The concept here is that if you are trying to sell something with an email campaign, the odds are good that someone else has also tried the same thing before you at some point. How many different ways do you think there are to say "2 for 1", "short-term sale", "weekend extravaganza" and the like?

So load up all your different subject lines in the tool and Touchstone will predict what your open rate and CTR will be. They're so confident in their tool that they offer you a free trial. If Movable Ink allows you amongst other things to limit the losses you take from your recipients getting the wrong subject line, Touchstone is designed to cut those losses entirely. Knowing ahead of time what your mail metrics are going to be sounds a lot like time travel to me!

My next topic will be machine-learning algorithms, what they can do for you and what they can't (yet) do.