Earlier in the year, we decided to switch email newsletter providers. Like many small businesses, we used MailChimp with a fair degree of success. But as time went on, we found several problems. Regular readers will remember my rant about GDPR compliance, but there was more.
On the one hand, while MailChimp provides a ridiculously easy way to prepare emails visually, the reality was that we knew almost nothing about deliverability and getting any real insights into open and click rates. The analysis was, how shall we say, sparse. At the same time, we needed a better understanding of what readers care about as a way of guiding what we publish and what we leave to others.
A big risk
We took a BIG chance in opting for PostUp, an enterprise-grade solution that few seem to have heard about. As I was digging into the product, it was clear from the get-go that this was going to be the enterprise equivalent of jumping from QuickBooks to (name your favorite ERP here) endeavor, with all the costs and grumbles that go with it.
One of the main reasons I went with PostUp is because it bundles ReturnPath and some Tableau functionality. For those that don't know, among other things, ReturnPath runs what are known as render tests on emails. The idea is that you can see how a particular email looks across multiple hardware/software combinations. This is vitally important because, as was clear to us during our MailChimp experience, no two hardware/browser combinations are the same. In turn, this means that even if you have what looks like a well-formatted email in the visual editing screen, the chances are it will not look the same when rendered.
The HTML nightmare
Useful? Sure. But it comes with a massive headache which has taken time for us to work through.
You may not know this, but there is very little commonality in the way hardware/browser combinations render HTML. There's a whole bunch of stuff you need to know about how browsers work and the HTML they accept. This is a PITA if you're a non-technical person because while many consider HMTL as a way of merely declaring how content is displayed, it is much more. Here's one definition:
HTML is a computer language devised to allow website creation. These websites can then be viewed by anyone else connected to the Internet. It is relatively easy to learn, with the basics being accessible to most people in one sitting; and quite powerful in what it allows you to create. It is constantly undergoing revision and evolution to meet the demands and requirements of the growing Internet audience under the direction of the » W3C, the organization charged with designing and maintaining the language.
Note the use of the word 'Language' and the reference to 'constant evolution.' Besides, it helps if you have an understanding of CSS, which can be used programmatically. Some people find CSS challenging to understand. I don't experience the same problem, although there are occasions when I go - WTAF happened there?
Once you know that, then it is not much of a leap to understand that while visual editors are great for simple layouts, they carry risks. This was what we found. That means (almost) every one of our emails needs some HTML check to ensure we don't inadvertently introduce formatting anomalies. More to the point, we have massively dialed back our use of images and other 'pretty making' devices to ensure that we keep the HTML as simple as possible.
What does this mean? We quickly discovered the effort we need to expend to achieve 98-99.5% deliverability. It is not 'point, shoot, forget' as it is with simpler systems. You have to take the time to make sure you've got your mailings in good order. PostUp provides us with the capability of resending where an email is not delivered or soft bounces. That's helpful since there can be any number of reasons behind a soft bounce.
What else can PostUp do? As you'd expect, extensive conditional sending capabilities are depending on how you want to slice and dice your email list and if you wish to address particular behaviors. This has proven useful with events coverage.
Right now, we are in the middle of Dreamforce 2019 and have a segmented list of some 1,050 people who expressed an interest in that topic. We've done the same for a couple of other vendors. Fine segmentation allows us to better 'target' interests, although that is far from the whole story. In another segment, I have a list of what I call 'fans' - people who regularly (but not always) open and/or click on something we send. Over time I will use this small group to understand better the things they care about and how those change over time. Why? Because experience and history teach me that if we address the needs of the few, they are more likely to act as force multipliers for fast followers and, eventually, the central majority.
What's more, I can see all or any of the campaigns through a single pane of analytical glass. This gives me an easy way to assess performance visually.
Why lists rot
I've also started to use PostUp segmentation analysis to cull our lists. Yes - you read that correctly. I actively dump people off lists. Why? There are several reasons. Top of mind are:
- Interests change over time
- Some move away from technology altogether
- Some got what they wanted at a point in time and are no longer interested in a topic or topics we cover
- Some are plain fed up with all the email newsletters they get (I am on about 50 lists a DAY - yikes) and send our stuff to a 'read later' (which they rarely do) folder. Suggestion - try SaneBox as an alternative to dumping material straight to a 'read later' folder. It's much more efficient and allows you to still push content that matters back to your inbox.
- Time gets too crunched to care enough about the stuff we put out
- Content consumption methods change. Jon Reed recently talked about how to get the best out of RSS, which can act as an excellent proxy for email newsletter content.
- In some organizations, IT policy actively strips out email newsletter messages as a way of keeping server-side cost operations in check. But then I've discovered that a good number of our subscribers use private Gmail addresses through which to read our content.
Some will argue that dumping people is a sign that your list is crappy, but aren't they all to some degree? Yes, I know that quantity matters, but equally, I know that quality matters much more. I also know that if you don't engage people quickly, then they don't come back after the first look-see. So it is vitally important that we not only get a new subscriber's attention but follow up to discover why they may not be opening emails to which they have subscribed. As Bill McDermott, ex-CEO SAP, now CEO ServiceNow once famously said: 'Trust is earned in drops and lost in buckets.' Also, dumping people means that my open and CTRs are industry-beating by some 40%.
Is there more we can do? Heck yes. PostUp has us signed up for audience development, and I've yet to take advantage of that, despite shelling out a four-figure per month sum for that service alone. We have a unique problem that will challenge PosUp's audience development people - we're not selling anything, so there's no apparent 'buyer journey' down which to steer readers. Neither do we have an advertising model, so that's another well-worn path we can't tread. It will be interesting to figure out what we do. I have some ideas I want to test, and I will revisit this topic later.
What about the overall experience?
I've got to say that PostUp is a beast. It's clunky when compared to modern mailing systems. The UI and UX are frighteningly old fashioned by current standards, and I can see many 'shops' eschewing it in favor of something similar. To make it work, you have to be very clear about all the things you want to do, and here, PostUp ticks a lot of boxes. And it is because of that I am learning to accept some 'why is this so shitty?' compromises.
The upside is that the insights gained from this first six months' experience are actively helping us shape what we do for the future, which, in turn, helps everyone in our small world. Now all I have to do is figure out who in the team, I can hand some of this off to without creating a support crisis. Should be a doddle (said nobody.)