Saturday, February 9, 2013

e-mail marketing and the "reader engagement statistic"

originally posted on the posterous site. january 17, 2011 at 04:23:00pm

this is my first professional post, if you’re a casual reader – skip this one, you’ll be bored out of your mind.

as most of my readers may know, my chosen profession is in e-marketing and email marketing. this industry of course continues to evolve and grow – and because it can be such an instant gratification form of marketing, it is here to stay.

email is a daily part of almost everyone’s life (i say almost because i got my mother an ipad and showed her how to use e-mail and yet, i’ll send an email and get a phone call in return…moving on). pingdom reports that there were 107 trillion emails sent last year, and pingdom assumes that 89% of those were spam. meaning that 262 billion spam messages were sent by, well spammers, world wide1. i am of the mindset that 200 billion of those spam messages ended up in my personal email box, i mean seriously what sites have i visited that would put me on a list telling me that the hip replacement surgery i recently had could have included a faulty medical device that has had a recall issued? (for the record, i’m 36 and as far as i know, i have never had hip replacement surgery). to put this in to some sort of perspective, in 2009 usps processed 177 billion pieces of mail2.

that being said, and with those statistics being predicted, the last quarter of 2010 found “user engagement” being tossed around as the latest buzzword and the latest fire that e-marketing professionals needed to put out. needless to say, all of this email puts a huge drain on the resources of the internet – the servers at the isp level, the end users’ computer and inbox; all of these things are affected by all of this email.

in response to all this spam, our friends over at google in late 2010 introduced the “priority inbox”. this setting looks at your gmail box, and determines things like: who you email the most, what messages you open the most, what keywords spark your interest, your reply habits and if you classify the message as important and unimportant. eventually, the gmail box becomes so good at this that “like messages” are presented first to the reader3. great! you are helping me as a person to better organize my box, awesome. but, as an e-marketing professional, you’ve made my work that much harder.

so how do i know that my lists are actually working? how can i help google (and the other isp's that will follow with similar options) to make sure that my email is not classified as spam or pushed to the bottom of the box and never seen? sure, i’m getting into the email box (my monitering software/consultant tells me i am), but i need open rates to determine if the members of my list are actually reading it, and to deliver the promised impressions to my advertisers.

with the company i work for, we are lucky. we get into email boxes, and have been pretty successful getting in and staying in those in-boxes and most of the time we keep a low complaint rate. but we are still concerned with user engagement and always thinking of ways to measure it.

the first way we look at engagement is since our newsletters include dart/dfp /doubleclick by google served ads, and our emails tend to be long; think we have content way below the fold; and by using the reports from dart/dfp /doubleclick by google we are able to look at the number of impressions delievered per ad (i.e. is the reader scrolling below the fold) and since they are arranged vertically down the “page” we can extrapolate where the biggest number of readers clicked off the email and didn’t come back.

the second one which requires less work/calculations is to use not one but two open pixels in your email. open pixels, i have found can be interesting “creatures”. some systems will record the pixel fire regardless of where it appears in the code, in the header, in the body…just about anywhere. our system isn’t like that – our system is a homebrew system, so while it has a number of quirks -- we also have the ability to make those quirks work for us at times. our system will only record the pixel when it is “scrolled over” i.e. the open pixel at the top will fire because it is “seen” in the preview-pane or when the email was actually opened, the bottom pixel won't fire until it "appears" in the preview/reading pane, meaning the reader actually scrolled down the page. for some of our email templates, since we knew this, we place open pixes at both the top and bottom of the template, which results in two open numbers. the first being the top (which is what we use to report impressions) and the second the number of readers who scrolled to the bottom -- or far enough for the pixel to fire. doing some math, we can determine the delta, and therefore know the percentage of readers who didn’t scroll to the bottom, or the percentage of people who aren’t fully engaged with your email.

sure this isn’t scientific, but it does help you to understand more about reader engagement and what needs to be done to get readers to look at the full email or at the least get overall engagement up to a point that google, hotmail or aol say is acceptable. i am of the mindset that email marketing is more art than it is science, and that there will always been a “few ounces of assumption” and at least one missing data point because you simply can’t obtain it (especially with a homebrew system like ours).

we have found, prior to updating our templates, we had about a 30% drop off on those emails (we have one email that is all headlines at the top, and the story and headline is way below the fold that enjoys an 85% drop off rate, it is the next template to be redone) and now with new templates and better delivery of content – using dedicated sections – we are seeing about a 16% drop off rate. which was hugely surprising for us during the testing phase because we expected it to stay constant because we were increasing the number of stories/content by probably double what was being sent out before (i.e. more exits). the other surprise, is that traffic back to our website increased by 30%, increasing clicks, time spent on the site, impressions and ad revenue.

to recap, the world sends a lot of email per year (most of it spam) and to cope with the increased volumes major isp’s are coming up with processes and products that make the inbox owners' life “easier”, but the role of the e-marketer harder, but you can get the information you need to get better engagement, cleaner lists and higher in-box numbers. we’re marketers, we went into this field because we are creative by nature (analytical by necessity) and you just need to put that creativity towards problem-solving sometimes. just put your mind too it…

i’d love to hear your comments and suggestions on reader engagement and email marketing, feel free to comment below.

1 Internet 2010 Numbers.
2 USPS Postal Facts.
3 Gmail Priority Inbox Beta.


Post a Comment