Litmus Email Analytics: a review
Jump to:
We don't have a lot of ways to measure how people interact with our emails. We can look at clicks and the much-discussed open rate, but that's about it.
The premise behind the Litmus Email Analytics tool is to offer more insight into your subscribers by revealing details on where they see your messages and how much attention they give them.
That kind of information would let you build and design better and more robust emails. And help measure how successful you are in engaging your readership, which in turn provides clues on how to improve that engagement.
I took the tool for a spin with my last two newsletter issues, so let's review my experience...
The concept
The tool is very easy to use. Once you've signed up for an account, you can track analytics on between one and five different mailing lists (depending on the price plan you choose). For each mailing list and campaign, you get a small bit of code to add to the HTML of your email.
Since the only requirement is to edit the HTML in the email you send out, you can use the tool with your existing email marketing software or service.
The code contains various calls to remote images hosted by Litmus, and their technology uses that as a basis to record engagement data. As such, the tool's reach is limited to those recipients who are not blocking images. So, in effect, you're getting data on people who register an open, not on the whole list.
What do you get?
The data collected by Litmus is displayed on a real-time basis at your account page. The main features are:
1. An overview of how many of those "opens" printed out your email and how many forwarded it (those are organic forwards, not people using forward-to-a-friend links):
2. An overview of the clients used by your readers to view your email. These are also broken down by version (for desktop clients) and browser (for webmail services):
There are also summary charts for overall browser use and the background technology (rendering engine) used by desktop clients to interpret your email's code:
3. An overview of the proportion of openers who "read", "skimmed" or "glanced/deleted" the message:
These three figures are also given for each broad reading environment (desktop email client, webmail, mobile email device). For example:
...and for each of the most popular clients, webmail services or mobile devices used by your subscribers:
What can you use it for?
OK, so what does all this tell us? Just how useful is the data?
The key is not to think in absolutes. Remember, we're seeing data for those who "opened" the email, not the whole list. So there will be a natural bias toward those clients that don't block images and those recipients that took action to activate images on those clients that do block.
Instead, think in terms of rough indicative data and (importantly) data you can use to track trends.
Forwards and prints...
Standard campaign reports offer no data on those who forward manually (rather than via a FTF link) so this is completely new insight.
Forwarding is an indication of how valuable and shareable your content or offer is. It gives you a broader overview of how viral your message was, especially if you combine the data with any stats you might get from "Share-with-your-network" links.
Another new metric is the indication of print outs, which would prove valuable for those campaigns using, for example, printable coupons, maps or invitations.
Not everyone can properly record whether emailed coupons are eventually redeemed offline. And even those that can have to wait days and weeks for all the results to come in. The print count gives you a rough, but fast, idea of how many people are at least interested enough to print out the coupon for possible redemption.
Engagement...
The overall engagement data is like a more accurate (though still not perfect) version of an open rate in terms of telling you how much attention your email is really getting.
You can see immediately if you have an engagement problem, and clever use of code supplied could, for example, let you compare engagement between particular segments of your list.
The numbers can also help you understand the factors behind click patterns. Consider two emails that get the same clicks and opens. Engagement numbers help you understand, for example, that one has a good call to action, but fails to draw the reader into the email proper to see that CTA, while the other is read more thoroughly, but has a weak CTA.
In essence it gives us intelligence on a part of the conversion chain that used to be a mystery: the "read" part of the open - read - click - convert chain of action. And, of course, you can compare numbers between emails to pull out patterns on what's working (and what isn't).
Engagement numbers for specific clients also alert you to problems you might not otherwise know about. In my case, those using mobile devices would seem to give my emails more attention than those using desktop or webmail applications. It seems I am not alone.
So while mobile devices may make up a relatively small percentage of the list, they make up a greater proportion of engaged readers. The attention I give to mobile email marketing (design) issues needs to be greater than I might otherwise have planned for.
All this information is particularly important for those publishing standalone email content, where clicks are not required or a goal, and open rate is too vague a metric. Now such senders have a way to better assess the success of that standalone content in holding attention.
Display environments...
The data on clients, rendering engines etc. is also valuable, offering five insights I detail in an earlier blog post.
Since the data also includes numbers for total opens, you can cross-reference that with the equivalent figure from your ESP or software to see if there are any discrepancies. if your service is under- (or over-) reporting opens, you need to know why (a topic for another day).
Limitations and scope for improvement
As is clear from above, lots of insight pops out of the tool. But there are limitations and features I'd like to see in future versions. For example, I'd like:
- a facility to export or manipulate the data yourself. In particular I'd like to easily chart changes through time to engagement metrics so you can automatically compare campaigns
- a summary chart on the email clients report, splitting total opens between "mobile", "desktop" and "webmail". This would let you track the growth of mobile email more easily
- the option to record both unique and total opens (the latter is currently measured)
- more transparency on definitions. There is no explanation, for example, as to how "read", "skim read" and "glance/delete" are defined. If based on time thresholds, it would be interesting to know what the cutoffs are (and in future adjust them individually). An email with very little content may want to define a "read" differently than one with several pages of text in it
- a chart displaying when opens are recorded, broken down by type of client. This would let you see, for example, when most people are viewing the message and whether mobile email users exhibit different behavior in this respect
Summary
For years now, we've stumbled along with little real understanding of what happens between an "open" and a click. Despite the limitations noted above, the Litmus Email Analytics tool gives us access to extremely useful data that can help us refine our content and pinpoint design, copy and strategy issues that need more attention than others.
Highly recommended for those doing big email business. Those, like me, with small lists might want to get a few snapshots across a month, then save fees and return to the tool further down the road to pick up on new trends and changes. (There are no minimum contract lengths or sign-up fees to bother about).
Disclaimer: I have no business or financial relationship with the team behind Litmus. I did, however, get free 2-month access to their testing and analytics tools for the purpose of this review.
Need more email marketing guidance? Try the email newsletter or blog.