Bob Kemper

Email Subject Lines: Do symbols hurt email marketing response?

February 26th, 2010

Editor’s Note: The MarketingExperiments community is an interactive group with a great deal of questions and answers between marketers and their peers as well as with the MarketingExperiments staff. Occasionally we publish these interactions on the blog when we think there is a particularly good question that our readers can benefit from…

QUESTION:

I recently watched The Five Best Ways to Optimise Email Response seminar by Dr Flint McGlaughlin. I found it extremely enlightening and it provided a lot of food for thought. However, I have a quick question with regards to slide no. 22.

I appreciate your time and I’m sure you receive plenty of mailings of this nature; therefore I will get straight to the point.

In this slide, the recommendation is to change the subject line of the mailing from “Thank You For Making Us Your Florist Of Choice” to “15% Off – Our Way Of Saying Thank You!”

I understand why the wording would be changed to make it more endearing to the receiver but I wondered if the symbols added would increase the risk of the mailing being filtered and more inclined to be highlighted as spam – therefore reducing the success of the mailing. 2964298027_a32d8f75bc

In my experience I steer clear of any symbols in the subject line when sending large mail shots, especially %, ! and £. Am I being too cautious?

Kind regards,

Chris, BA(hons) Business & Marketing
Marketing
London

ANSWER:

Hi, Chris. Thanks for your question.

If I might broaden the question slightly to interpret its essence as a transferrable principle, could I restate it as…

How much validity is there to the conventional wisdom that, in the Subject Line of an offer email message, numbers, certain symbols (especially £/€/$, %, and !) and “SPAM words” such as “Free” and “discount” will cause a dramatic reduction in deliverability, and consequently effectiveness?

… if so, then it’s surely an important one.

In the case of the particular company and study referred to on Slide 22 – that was precisely one of the questions we set out to answer.

What you couldn’t see in the context of Dr. McGlaughlin’s presentation at the MarketingSherpa Email Summit in Miami is that this particular two-treatment comparative vignette was just a tiny part of a much larger and broader study. We intended to test the specific, widely accepted presumption you mentioned.

We were also exploring a host of other best practices to see how valid they remained through the evolution of regulations as well technical filter changes by email service providers (ESPs) since the time they were first introduced and anecdotally adopted (around 2003-2005).

This was important because we know from our foundational Offer/Response-Optimization principles of “clarity trumps persuasion” and “specificity converts,” that the clearer and more specific subject line – i.e., the one with the “15% Off…” copy – should convert better.

What we found was that there was, in fact, a small but significant difference in deliverability – interestingly, it was more pronounced among the smaller ESPs. In addition, as we had predicted based on the “eme” heuristic, the Open Rate actually declined (…by more than 25%).

In the end, though, the central research question was “Which email subject line will result in the greatest projected net revenue?” As revealed in Dr. McGlaughlin’s presentation, despite the slight dip in Delivery Rate, and the (what would otherwise have been alarming) drop in Open Rate, the Click-through Rate (CTR) to the landing page was 60.3% higher.

What he may not have mentioned is that, in direct answer to the research question, the Treatment subject line yielded a 56% increase in projected net revenue vs. the Control.

So, while it appears there is still at least some validity to the commonly held belief that special characters in the email Subject Line reduces deliverability, our research (this experiment plus two others conducted with different products and industries) suggests that when they serve to do so, these negative factors are dwarfed by the power of clarity.

I hope that’s helpful, Chris.

All the best,

Bob Kemper
Director of Sciences
MECLABS Group, LLC

Dr. McGlaughlin will next be teaching and speaking about email marketing at MarketingSherpa Email Marketing Germany 2010 in Munich on March 8th and 9th.

Dr. McGlaughlin’s four-hour workshop and keynote presentation will cover email capture rate and quality, open rates, conversion, and building customer trust and loyalty with email. He will also be conducting live optimization of audience submissions – a lively and always-popular segment.

Share and Enjoy:
  • LinkedIn
  • StumbleUpon
  • Facebook
  • del.icio.us
  • Digg

Categories: Email Marketing, Research Topics Tags: , , , , ,



  1. Susan Gibson
    February 26th, 2010 at 14:42 | #1

    This was a good discussion and very helpful. Testing always helps determine the best subject line. In my former company, we went completely away from using such words such as “Free” due to SPAM filters. But I am not sure we tested the CTR or delivery rate.

  2. February 27th, 2010 at 13:43 | #2

    Another great experiment.

    I would back up from experience too that use of free and symbols are not a ‘no go’ in the subject line. Just be sure you are measuring and testing using metrics as close as possible to your ultimate goal. In this example, revenue.

    Interesting to see that subject line based spam filtering ‘was more pronounced among the smaller ESPs’. This is because the smaller ISPs do not have such advanced reputation based filtering that the likes of Yahoo, Hotmail, AOL etc have. They are still using the techniques more of two years ago with content being a bigger scoring factor. As will be the case for B2B emails too.

    Note if you wish to use the Britsh pound symbol (£) in email subject lines you need to use “Quoted Printable” or “Base 64″ encoding of the subject line. Otherwise on many email clients it will not show correctly. The original email standard did not allow for non-US characters (not too surprising considering its origins go back decades).

  3. February 27th, 2010 at 20:23 | #3

    Great Blog! I think the subject line in the e-mail is critical since it determines where you will end up in a spam filter or deleted before opened. I think the more you can personalize with demographics, geography or data about the recipient the better your chances of surviving. E-mail isn’t the most effective form of prospecting but it is good for building relationships with current customers in the form of service.

    Tim Little
    Publisher, MArketingListBroker.com
    Publisher, MarketingListBroker.com

  4. March 10th, 2010 at 08:13 | #4

    This is an excellent response. In email marketing, we are constantly facing these types of questions, and it ultimately boils down to the return. We often end up recommending copy or design options that might reduce deliverability or opens some, but if you can get more conversion from the “targeted,” relevant audience with the clearer message, your return often does dwarf the reduction in some of the other metrics.

    Mitch Lapides
    President, FulcrumTech, LLC

  5. March 10th, 2010 at 08:15 | #5

    I don’t know how many different tests you might have done.

    But I’d like to point out that the test line “15% Off” has other variables that might have skewed your test. Was it the “%” that made the difference or the whole measurable discount “15% Off”.

    I’ve always told my clients that “using a measurable result” or a “measurable value” will increase the attention getting ability of an article title, or an email subject line.

    So, was your 50% increase due to the “%” or to the whole concept of a 15% discount?

    Did you test other subject lines with other symbols in other ways?

  6. March 10th, 2010 at 09:48 | #6

    My question is: Was the 56% increase in projected net revenue measured against the number of contacts sent to, or the number of emails delivered? The reason I ask this question is because logically, in my mind anyway, the number of deliveries will significantly drop due to various spam catchers put in place, whether by ISPs, email providers, or corporate I.T. departments. So if you measure against the delivered emails only, The 52% increase wouldn’t give a real indication of what special characters do to an email subject line. If measured against the delivered email, then of course the more concise subject will prompt a higher open rate which will in turn (hopefully) lead to a revenue increase.

    I would imagine the delivery rate would have been higher, resulting in an even higher revenue increase if the same thing were said without special characters. for example “15% Off – Our Way Of Saying Thank You!” could have been put as “15 Percent Off – Our Way Of Saying Thank You”. Here you’ve removed the special characters of “%” and “!”, but have said the same thing.

  7. March 12th, 2010 at 17:32 | #7

    Thanks for your comment David. I passed your ideas by our Director of Sciences, Bob Kemper, who was very involved in the initial research, and here’s what he had to say…

    That’s a great point of clarification, David, thank you. Because, as you suggested, the gross revenue from an email send is based upon the total number of actual clickthroughs to the site, the revenue projections are based upon the list size (total messages sent) rather than on messages delivered, which could vary considerably with message content, ISP mix and other factors.

    Thanks again, David, for pointing that out.

    Regarding the “%” symbol vs. the word “percent”, you may very well be right. Even so, I would want to test that presumption. Remember that for many ideal customer segments, each message receives only a minute fraction of a second’s attention in visually scanning the Inbox for topical relevance. If indeed I am one for whom a 15% discount is a compelling decision factor, the word “percent” may fail to trigger the “visual reflex” that the symbol “%” does—especially to one who is thus wired—and thereby inadvertently miss the Open. So once again, David, your hypothesis could very well be right; in any event, it is decidedly test-worthy.

  8. January 6th, 2013 at 21:03 | #8

    There’s a great list of symbols for use in email marketing subject lines on wikipedia http://en.wikipedia.org/wiki/Miscellaneous_Symbols

  1. February 26th, 2010 at 16:16 | #1
  2. March 3rd, 2010 at 09:12 | #2
  3. June 25th, 2014 at 12:46 | #3
  4. June 25th, 2014 at 13:15 | #4