Richard on his last day at the Email Summit:
I’m just really beginning to notice how crowded it is here at the MarketingSherpa Email Summit here in Miami. There are hundreds and hundreds of people. I can only assume that the restaurant / bar hosting tonight’s after-party must be pretty huge. I hope they have a dance floor large enough to contain both me and all these people together. A normal dance floor can barely contain me alone.
What these sessions (the first hosted by Liz Bullock of Dell and the other by Tara Zanecki of Workshare, Inc.) conveyed above all else is the massive amount of variables you can tweak when testing. Note that if you haven’t done any testing in the past, the first thing you’ll want to learn how to do is use our list segmentation tool to do A/B split testing. (Basically it's testing two things and cutting your list in two equal parts to do so.) We’ve done a webinar on list building and segmentation - hosted by myself and Erin Jacobs, our Director of Marketing - and we also have a short tutorial showing how to put the tool to use. I’d recommend checking both of them out.
To quickly sum up what both hosts discussed:
Liz has very little control over the pages on the Dell website (there are apparently more than 4 million pages on the Dell site, which is beyond ridiculous), and a request to have something changed can take months. The only thing she can directly change is the landing page she uses for deals offered to current and potential Dell Small Business customers. As a result she wants to make sure her landing pages are optimized to a point that they perform better than any other page on the Dell site.
She did a variety of tests which involved splitting her lists between current customers and subscribers who had not yet become customers. She discovered that the Dell Deals page, which lists several computer deals Dell has going at an given time, did better than the landing page among non-customers and that the "Configurator" page, which allows site visitors to configure the best system for them, did better among current customers.
To improve things she removed clutter on the landing page and reduced the number of options visitors were given to navigate away from the page. She also adjusted the layout of the page and added a quick link to schedule a call with a Dell Sales Rep to answer questions. The page then did better than both the configurator and Dell Deals pages (this improvement in performance took a lot of changes and testing). She continues to make changes and test those changes to try and perform even better.
Tara faced an issue in which a long lived version of Workshare’s software was being phased out and replaced by a new piece of software they’d created. She needed to convey this information to as many customers as possible (meaning those customers needed to open and read the email they received) to make sure they understood why it was being phased out and how easy it would be to make the change to the new software.
She tested two versions of the email: one that was fairly short (a bit less than a page), somewhat generically salesy, and to the point, and another that was personalized with an actual employee’s name, was written in a more personal style as if it was coming from that employee, and was quite long (three pages) in order to explain what was happening.
Though general best practices would call for a shorter email to perform better, the longer, heavily personalized email performed significantly better. Though this performance was almost certainly helped by the personalization, it still shows that a longer email can perform better with the right audience if it’s providing them the value and information they’re looking for.
That does it for day two. Now I’m off to the networking party! I’m planning to walk, through Miami’s downtown?
Comments