Hitmetrix - User behavior analytics & recording

How valuable is the open rate for e-mail?

Loren McDonald
VP of industry relations at Silverpop
More than 24 years of experience in market­ing, consulting and strategic planning

The open rate is a tired, flawed e-mail marketing metric that should be mothballed. It retains some minor value for diagnostics and benchmarking, but it is largely unreliable and inconsistent. Most importantly, it gives marketers little meaningful data to act on. With so many consumers using preview panes, blocking images and reading e-mails on mobile devices, the concept of an “open” no longer reflects how they interact with the e-mail. This renders the tracking image as useful as a buggy whip is in making a car go faster.

Just one example: Ibrowse an Amazon e-mail on my BlackBerry (text version, no images rendered, no open counted) during the day. Later, Iread it again in my Gmail account on my home PC (images blocked, no open recorded). I“opened” the e-mail twice, yet it will be reported back as unopened and unread.

Beyond its unreliability, the open rate simply provides little value in making decisions anymore. Smart marketers are focused on strategic and output aspects of their e-mail programs: How many conversions did a program drive? Revenue per e-mail? Cost sav­ings from the e-mail channel? No mar­keter will get a raise in 2008 because of their e-mail open rates. Marketers are more accountable now for their budgets and actions. Do they want to base perfor­mance reports and marketing decisions on such a flawed and inconsistent metric? Not the ones Iknow.

Evolve the “open” metric into a new one — “images rendered,” for example — which better reflects what a track­ing image actually measures. Also, all marketers need to focus on more reli­able e-mail measures that reflect their employers’ business goals.

Morgan Stewart
Director of research and strategy at ExactTarget
Former executive at Draftfcb and Targetbase

While its significance may have waned as a measure of overall campaign success, open rate still offers insight into key aspects of program performance.

Campaign optimization begins with subject lines — encouraging recipients to engage with your e-mail and setting expectations for the enclosed content. Open rate is the primary metric for evaluating subject line tests.

Test groups for subject line evalu­ations should be selected through random sampling so each e-mail record has the same probability of being selected. This creates equality between samples and allows conclusions to be drawn from test results. The number of people within each test group viewing your message with images blocked will be the same — within a margin of error that testing for statistical significance will keep in check. As such, under re­ported open rates affect the test groups equally and test results remain valid — a statistically significant group, based on the comparison of open rates, can be declared the winner with confidence.

With image blocking becoming wide­spread, e-mail opens are consistently under reported. However, the vast ma­jority of click activity is still preceded by a tracked open. Looking at Exact­Target’s data for the past two weeks, 88% of e-mail clicks followed a tracked open — the images were displayed and the open action tracked 88% of the time before a link was clicked. While not ideal, the use of opens and open rates as tools for optimizing e-mail performance are unaffected by under-reporting — if accounted for.

From subject line testing to open-to-click ratios (which help assess quality), open rates provide valuable insights unique to the e-mail channel.


McDonald makes a sturdy case for the shortcomings of the open rate, however one can’t ignore Stewart’s contention that the metric, whileimperfect, does serve a purpose in a marketer’s tool kit. Direct marketers can rarely rely on just one metric when judging any campaign; it’sbetter to use the insight of a full set of resources, and consider each metric’s strengths and weaknesses when applying it to future plans.

Related Posts