How do you measure the effectiveness, or success, of an e-mail marketing campaign?
Though much has been written about click-throughs, HTML readership, and such, ultimately the metric that really counts for you is the profitability of your campaign. In other words, how much revenue was generated, versus the costs you incurred?
Clearly this return on investment (ROI) is the ideal measure, but in practice might not always be possible to determine. So let's look at the range of possible measurement criteria, and try to establish the usefulness of each one, as well as the results you should expect from a successful campaign.
The Basics At the simplest level, we can track delivery and subscriber information.
Delivery information tells us how many messages we attempted to deliver, how many were undeliverable (they "bounced"), and therefore how many were successfully delivered to the intended recipients.
Typically, if you are mailing to a list for the first time, you should expect a large proportion (perhaps as high as one-third) of the messages to bounce. This happens because your list was compiled some weeks or even months before the mailing date, so a number of those e-mail addresses are no longer valid: a user has changed his or her Internet service provider (ISP), or moved to a new job with a new e-mail address.
Subsequent mailings to that list should, however, maintain a consistently low bounce rate -- typically less than 5%. One point to note is that you should use an e-mail marketing system that automatically handles all such bounces. Otherwise, your inbox will soon fill up with hundreds, if not thousands, of bounced messages, out-of-office notifications, and "unsubscribes" (users requesting to be taken off your list).
Your system should also update each user account to indicate current status so that you no longer attempt deliveries to users who have unsubscribed or whose messages have bounced.
Subscription information tracks how many active subscribers you have, and how many have unsubscribed. This information can tell you a couple of valuable things.
First, a high unsubscribe (or opt-out) rate indicates that your customers do not find the content of your e-mail messages very compelling or useful. Second, the rate of growth of your subscriber list indicates how successful your acquisition efforts are, and can indicate high levels of pass-along subscribers (when an existing subscriber refers a new one to the program).
Ideally, you want to know where these new subscribers came from: Did they respond to a particular promotion? Are they pass-along subscribers? Or, did they come from an external e-mail list you tested?
For an e-mail program that is performing well, you should expect to see consistently less than a 5% unsubscribe rate for your overall list. Your list should also be growing, with a number of new subscribers that far outweighs your unsubscribe rate.
Today's more sophisticated e-mail marketing tools will allow you to compare delivery and subscription information across different customer segments so that you can, for example, see whether particular customer groups have higher unsubscribe or bounce rates.
But Are They Reading It? The next level of measurement involves message readership. I find it very useful to send HTML as well as plain text e-mail messages. Apart from the benefits to the recipient of an attractively formatted, customized message, HTML also offers the marketer some significant measurement benefits.
One of these is the ability to track how many of the recipients of your message opened it as an HTML message. This gives a broad indication of the number of "impressions" or "readers" generated by the HTML version.
Bear in mind, however, that it is not a totally accurate measure of readership because some users might choose to turn off graphics in their HTML e-mail browser; others might start opening the message, but delete it before they have actually read it. This can be explained by the way the process works.
Each time a recipient opens an HTML e-mail message, his or her browser will start to download any graphics within that message; the graphics are not actually sent as part of the message itself (this also helps to keep file size down).
To count impressions, the e-mail marketing system simply tracks one particular graphic within that message (usually the one that's nearest to the start of the HTML file) and counts the number of times it has been downloaded. Clearly this does not work for plain text messages, which do not download any Web-based graphics as part of the process.
Typically, you can expect to see about a 30% to 50% readership (of total recipients) in HTML format. You will also have a proportion who viewed the message in plain text format, but readership cannot be measured for these.