Lessons learned running an HTML email marketing campaign

Posted in: Archive

Last week we did some stuff and it was fun and we learned a lot of things. Our Content Manager, Richard, took the lead as we embarked on a week-long odyssey to retool and refocus on digital marketing. A great chance for the content team to blow off some cobwebs, and for the development team to get an introduction to the "dark art of marketing" (their words).

As well as providing some amazing presentations and great discussion points on digital marketing during the week, Rich also acted as a client with the requirement for some digital marketing in order to put the team through its' paces.

In a nutshell we had a predetermined (small) budget with which to promote the (fictional) launch of an awesome new way for students to communicate with us. It was to tie in with the University's Open Day on June 21st.

First thing to do was to decide how best to spend the budget - what would give us the biggest bang for our buck?

Although it's been around forever, and despite there being much more shiny attractive things to use as the basis for your digital marketing campaign, still nothing has the same impact as an HTML email in terms of views and conversions. That's not to say it's a magic bullet, there is still a heckalot to consider when planning the content and the message, never mind the nuts and bolts of building something that's based on the kind of HTML that would embarrass you at parties.

One good thing our limited existing stats showed though was we no longer needed to worry about Lotus Notes, the bane of many an email marketing campaign in my past. In fact, figures for recent undergraduate and alumni campaigns showed a very different trend than expected (Outlook still does come out top, but I was surprised to see the traction that the iPhone had as a device for checking email).

Top clients for Alumni campaign:

  1. Outlook 28.7%
  2. iPhone 22.3%
  3. Apple Mail 10.0%
  4. iPad 9.8%
  5. Hotmail 7.6%

Top clients for  Undergraduate campaign:

  1. Outlook 33.1%
  2. iPhone 20.8%
  3. Hotmail 18.1%
  4. Chrome 8.1%
  5. Internet Explorer 7.1%

Another thing that has changed is the amount of support you have for creating your HTML email, whereas before I've had to manually create the code and then test send it to half a dozen different mail clients - now there are numerous (freemium) online campaign handlers that come with preset templates and customisable examples. Of course, the code is still something best not looked at too closely, but they do a lot of the heavy lifting for you.

We identified three candidates to use for our campaign; CampaignMonitor, DotMailer and MailChimp. Right now I have to give a big shout out to CampaignMonitor's resource pages, which I plundered extensively. They're really worth reading, I'm told they put a lot of work into them 🙂

In the end, we went with MailChimp as the client (that's Rich remember) expressed a strong (possibly monkey-influenced) preference for it, though other departments within the University tend to use DotMailer for their email promotions (it may not be the sexiest of the 3, but it does have great analytics).

Now that we had our medium, and to a certain extent our message worked out - we really needed to get down to planning the nitty gritty of the campaign, identifying and slicing up the audience, delivery dates, campaign length, and what and how to track all the information.

Enter the Planning Board

We agreed the audience would be people in the UK who were in their last 2 years of non-HE education. From this pool we divided them further into those just starting to think about coming to uni, and those who had already decided to continue in education and were looking at their options. We then further divided this second category to shake out those who had already contacted our uni. This last group was deemed the most likely to be interested in the product we were offering, as they had already engaged with us at one level. This meant we had 3 different groups to target with our campaign:

  • Considering the future
  • Decided on university
  • Considering our university

We actually later identified a fourth group - those that had a confirmed place at our university and we created a separate campaign for them, but they do not feature on the Board of Plan.

You'll also notice Facebook integration and a whole offline campaign - we really went for it! - but I'm only going to cover the email side of things in this post, as it was the one where we learnt the most. Also, calling it Project Gandalf was Kelvin's idea.

We arrived at our targets for open rates and clickthroughs based on the national averages for our sector, plus a bit of good faith that the more engaged audiences would be more likely to act on the email. We had very little real data of our own to back this up, so it was a little bit finger in the air for a lot of the percentages we settled on. The obvious benefit of doing a more large scale campaign like this is of course the rich veins of information we would be able to mine afterwards, allowing us to improve our campaigns in the future. We've never really looked at a potential audience in such depth before, we even considered breaking each group down further using gender and location information to really tailor the message.

The other main thing we learnt from this was about delivery timescales - we again had no on-hand information so needed to research when the best time to start a campaign would be. We had the launch date, and we figured the emails needed to go out long enough before that date to allow people to act, but not so long as to be something filed for later, and then forgotten. Fingers in the air, 2 weeks felt about right. We looked at this date and found that it coincided with the summer half term dates for schools, making it ideal. In terms of narrowing it down even further, we did a lot of searching online and found a great guide from the MailChimp guys showing their data on this. So Thursday at 3pm was a no-brainer.

It may turn out that our audience is more likely to act on emails received at a completely different time, but without any data to go on we felt it best to run with MailChimp's.

daysofweekgraph

What we learned from doing all of this is really how easy it is to set up and run an HTML email campaign once you've fully understood the audience and your message. The planning does take a long time and there are a lot of variables to keep track of at once, but the rewards far outweigh the costs.

Unless you blow your budget on branded nylon fold-up frisbees, which we almost did (they were super-great though).

Posted in: Archive