Stop drinking the kool-aid! Your subject line test strategy needs a complete overhaul.
Last week I released a whitepaper on subject line length and its effect on open, click and click-to-open (CTO) rates called Subject Lines – Length Does Matter. At first glance, our analysis of over 640 email campaigns supports the widely held view that short subject lines are best for opens. But what it goes on to demonstrate is that when it comes to subject line length, more opens = less clicks!
We discovered that although subject lines with 60 characters or less make more people open the email—the traditional view—these people are less likely to then go on and click on content or offers within the message than people who open an email with a longer subject line. There is an inverse relationship between opens and both click and CTO rates.
The study goes on to show that CTO rates start to be optimized when the subject line is over 70 characters in length and continue to rise until well beyond 100 characters. This is true for both the number of characters and the word count. The more words there were in the subject line, the better the CTO rate (if anything the results are stronger for word count).
Given the strength of the whitepaper’s findings, I wanted to find out what subject line lengths marketers were using as common practice. As you might expect, we monitor a large number of UK email campaigns, from a cross-section of sectors and companies including British Airways, Hilton, Apple, Amazon and Reuters and we were astonished to find out just how many subject lines sent by these companies were short (i.e., 60 characters or less).
Out of 700 subject lines sent to our monitoring accounts in the last 90 days, the vast majority—87% of them—were under 60 characters in length. A further 7% fell into the ‘dead zone’ between 60 and 70 character where neither opens nor the CTO rate is optimized, and only 6% of the subject lines were over 70 characters long and therefore likely to optimize click and CTO rates.
Does his mean that everybody out there is only interested in opens and doesn’t care about clicks? Or perhaps that email marketers, having extensively and regularly tested longer subject lines, know for a fact that they don’t work? What’s more likely to be the case is that as an industry we’ve done such a fantastic job of drinking the kool-aid that we have stopped testing outside of the accepted norms.
Our whitepaper also found subject lines with a higher word count also optimize clicks and CTO rates. So how do the numbers break down when it came to word count? The numbers are equally amazing. Only 13% of subject lines monitored contained above 10 words—where clicks and CTO are optimized. 60% fell into the ‘dead zone’ of between 6 and 10 words, where neither clicks nor opens are optimized; and 26% of the subject lines contained fewer than 6 words, and therefore optimized open rates.
So what I have learned from this exercise is that email marketers need to completely overhaul their subject line test strategy:
● Subject line tests should be more granular—long and short just isn’t good enough. Subject lines need to be broken down into more character groupings (1-10, 11-20, ...91-100).
● Introduce word count testing. Words are a much better way of conveying meaning than characters.
● Assess the impact of the number of propositions contained in the subject line on your campaign performance.
● Finally, open rates are just a small part of the story. Your tests should assess the impact of subject lines on clicks, CTO rate and conversions, as well as sales.
My greatest fear is that the people reading the whitepaper will be looking for a simple answer such as “when it comes to email subject lines, short is best”, when in fact the central message is keep searching, keep optimizing and keep on challenging assumptions.
—Dela Quist of Alchemy Worx
.gif)
Comments (8)
Great job testing something outside the box!
One question that I have though is was the CTO rate affected only by the open rates going down? In other words, did you find that you actually receive more clicks with a longer subject line than with a shorter one?
Posted by Tim | July 10, 2008 9:01 AM
Posted on July 10, 2008 09:01
The CTO rate was affected by the click rate going up AND the open rate going down.
It appears that when it comes to subject line length Click and open rates are inversely related.
Dela Quist
Posted by Dela Quist | July 10, 2008 1:28 PM
Posted on July 10, 2008 13:28
Dela,
Would you be able to do some analysis of the click performance of these messages without considering the open rates?
For example, what if you took the total number of clicks and divided it by the total number of delivered messages for each group?
I would be very interested in seeing those results.
Posted by Tim | July 10, 2008 4:57 PM
Posted on July 10, 2008 16:57
Tim
We did analyse clicks in the white paper.
Click rate (total clicks/total delivered) is represented by the yellow line in the 2 charts. The CR trends upwards as character and word count increase.
Posted by Dela Quist | July 11, 2008 4:17 AM
Posted on July 11, 2008 04:17
Thank you for this analysis... we've experimented with sooooo many combinations and variations in our subject headers and have found simple + incentive has been our best bet as a clothing company. We are always looking for an edge to get to get past the first hurdle of opening.
Example: Summer 08 Launch! (Free T-shirt with order!)
Chris
PeakeUSA.com
The Peake Pals (http://peakeusa.ning.com)
Posted by Chris Richards | July 13, 2008 9:27 AM
Posted on July 13, 2008 09:27
Hi Chris
Glad to see you do a lot of testing. It is interesting that you say you are always looking for an edge to get to get past the first hurdle of opening.
What about the 2nd and 3rd hurdles (clicks and sales)?
Have you ever run a test where the subject line wich made the most people open delivered fewer clicks and or sales?
Posted by Dela Quist | July 15, 2008 6:36 AM
Posted on July 15, 2008 06:36
What about the word count of articles. I have always though shorter is better, but does a dead zone exists with content as well?
Posted by Elle Wheeler | October 16, 2008 12:44 PM
Posted on October 16, 2008 12:44
Great study. I had a client who insisted that we always go short on the subject line to the detriment of the message. Wish I had this white paper back then.
Posted by Scott Posner | October 16, 2008 5:52 PM
Posted on October 16, 2008 17:52