It wasn’t the subject line: dozens of other factors impacting your results

August 19th, 2010

Traditional direct marketing has the 40/40/20 rule.
40% of your success is dependent on the audience, 40% on the offer and 20% on everything else.
Email marketing has the 90/10 rule.
90% of your success is dependent on the subject line and 10% on everything else.
At least you’d think so, given the volumes of space dedicated to the topic of writing subject lines. Mea culpa.
When a promotion or newsletter pulls in unexpected results, our first thought is to point the finger in the direction of the subject.
After all, we sent the email to the usual list with our usual sender name and standard design template, with a typical offer or tidbit of content. It must be the subject line that did the heavy lifting or screwed up.
Or not.
Even if you exclude the offer/content and list from the equation, there are many other factors that could be skewing results one way or another. It may indeed be the subject line, but it may be something else.
The wider we cast our analytical net, the better we understand why we got the results we did and how we can exploit that knowledge to do (even) better next time around.
Here’s my list of other factors influencing results…what would you add?
Deliverability
Hear that noise? Me neither. It’s the sound of an email not landing in an inbox. A depressing non-sound.
The total response you get depends on how much email actually landed where people can see them.
Bounce rates
Anything unusual happening with bounces? A jump in soft bounces, for example, might be a sign that your mail was triggering spam filters. An example from my own reports:

That harmless soft bounce is actually mail rejected as spam.
Inbox deliverability
Low bounce rates translate to a high delivery rate in a typical campaign report. But, as this article explains, this delivery rate only tells you that email didn’t come back as undeliverable. It does not tell you if the email was deleted before delivery or redirected into a junk/spam folder.
Getting a handle on true inbox deliverability is tricky, as most ESPs or email marketing software simply can’t deliver the information.
One possibility is to segment your list by address domain and compare responses across the major domains and across emails.
If you notice that, for example, hotmail.com addresses produce an unusually low response compared to other domains and previous emails, then you likely had delivery problems there. If the response is unusually high, perhaps a delivery block was lifted.
Another method is to use an email analytics tool, like Litmus or MailboxIQ.
And another is to set up accounts at common ISPs and webmail services, add them to your list and then monitor what arrived and what didn’t.
This kind of “seed list monitoring” is also offered by a number of services, including Return Path, Pivotal Veracity, EmailReach and Delivery Watch. These companies also offer other diagnostic tools to identify the cause of any delivery issues.
Delivery delays
Delivery is also about when email arrives and not just whether email arrives. Sending delays (which some of the services mentioned above can also monitor) can throw out campaign timing and make a hot offer meaningless.
A Friday afternoon email advertising a Saturday morning in-store sale won’t perform well if the emails don’t actually reach inboxes until Saturday lunchtime.
Remedial action
Did you do anything before the send to change your deliverability? For example:
* Did you just finish a dedicated “please whitelist us” email campaign?
* Did you add whitelisting instructions to your welcome messages?
* Did you or your ESP use a different IP address to mail from? One with a better sender reputation? Or a worse one? Or a new one with no reputation?
* Did you start authenticating outgoing messages?
* Did you start certifying your email?

Copywriting
Let’s assume the general quality of the copywriting is the same from email to email. But what seemingly innocent copy elements can throw off results?
Preheader
If you’re using snippet text to promote the email’s content, like this…

…then that snippet is like a mini subject line itself. Not nearly as powerful, but still capable of drawing people deeper into that message. if you’re not using and optimizing that piece of text, perhaps it’s time to rethink the preheader.
The call-to-action
Many a good piece of copy is ruined by a weak call-to-action. Even small changes in wording and placement can affect CTR dramatically. Factors worth considering are:
* Wording and length (what words did you use?)
* Position in the email (was the CTA at the top of the email only? Bottom only? Did you move it closer to (or further away from) related copy and images?)
* Repetition (how often did you repeat the CTA for a particular landing page? Did you use a dedicated CTA only, or also link images and headlines?)
* Competition (how many different actions were competing for attention in the email? Were there too many distractions from the main CTA?)
* Design (did you change the color, size, shape or font? Did you change to an image-based CTA? Or to a text one?)
Link robustness
In an ideal world, you tested all links before sending out your email. Even so, things can still go wrong. For example:
…if you set up the campaign well in advance of the send date, one or more links may have broken in the meantime.
…the landing page link may work, but maybe the tracking link broke because of maintenance work at your ESP.
…the link “worked”, but it was the wrong link.
[I once got a surprisingly large number of clicks to an article featured briefly right down the bottom of my newsletter. Turned out the main article link at the top of the email was pointing to that minor article, rather than the intended one. Ouch.]

Timing
When we talk of timing in email, we tend to think of the best day and the best time of day to send out a message. So, yes, the time, day and date all impact results. But timing goes beyond that…
Weather
Pure360 found that responses to certain kinds of offers and content were closely correlated to the prevailing weather.
Intuitively we know that clicks on an email promoting sunscreen are likely higher when the sun is shining. But there are less obvious connections, too. For example, the study revealed that CTR on emails…
“…promoting business related events and products increased from 12% when it was raining to 27% in the sunshine”
..but that…
“Campaigns promoting restaurants are twice as effective in bad weather”
Other things going on
It’s hard to believe, but subscribers have a life beyond email. What else is going on that might distract them from checking their messages?
* Public holidays: maybe B2B emails won’t work on a holiday? Maybe consumers have more time to read? Are they outside enjoying the extra freetime, checking the latest DVDs or catching up on mail?
* Seasonal influences: a super subject line won’t help response much if you’re selling BBQ equipment in the fall (unless it’s a clearance sale?!)
* High-attention events: events make great themes to build campaigns around, but the events themselves draw attention away from email. Was everybody watching the Superbowl? The World Cup final? The Olympics opening ceremony? The Presidential inauguration?
* High-attention news: did your email go out just as the Berlin Wall was coming down…as the war in Iraq started…as the Prime Minister resigned?
Your wider email context
The focus on time and day also reflects a focus on each email as a standalone message. But each email is also one of a stream of past and future messages that may run into the hundreds or thousands.
So the position of that one email in that stream also affects response.
What kind of emails (and how many) did that list or segment get before this one? How might that impact responses to the current email? Is the content/offer fresh or a repeat of the last five messages, sent in the last 10 days?
Previous experience with your emails also drives future attention.
If the last email was high value, then recipients are more likely to give the current one more time and attention. If the last email was rubbish, they’re less likely to give the current one a fair chance.
Other marketing influences
Email does not operate in a parallel world untroubled by outside influences. Other marketing activities impact email’s success as a marketing channel.
Your other marketing
One of the big challenges in online marketing is attribution. If someone interacts with you online through different channels, which channel gets the credit when they eventually convert?
The fact that it’s not automatically the last channel of contact reflects our intrinsic understanding that every touchpoint with a potential customer has an impact on the eventual conversion.
Did you triple your PPC search engine spend the week before the email went out? What other marketing was your company doing as you sent out your emails?
Did the email send follow an intensive direct mail and TV campaign for the very product advertised in the email? If so, some of the conversion work was already done before the email arrived.
Spam
Spam levels vary day by day. If a botnet comes online or gets shutdown, your list might get a shortlived increase or decrease in the amount of spam that clutters up their email accounts.
A bad spam day might drown out your message. A good spam day will give it more prominence in the inbox.
Competitors
And what about your competitors?
Maybe they’ve done lots of advertising for the kind of product/service you promoted in the last email? Or maybe they sent an email with the same offer just before you did…or maybe they had a 30% off everything sale and yours was only 20% off.
Your list isn’t unique to you: if they signed up for your SEO newsletter, maybe they signed up for someone else’s, too.
Competitor influence is particularly likely during seasonal hotspots (like pre-Christmas or the start of a new semester) when everyone is pushing their 20% off seasonal sale.
List changes: same list, but…
You may be sending to the same list or list segment, but a list is an organic beast. It grows, it shrinks, it changes.
New subscribers
Did a major address acquisition effort boost the proportion of new subscribers on your list? On the whole, fresh sign-ups tend to respond more than long-timers.
Where did those new subscribers come from?
Did you just dump thousands of sweepstake participants into your list and see them drag down responses?
Or did you add those few hundred tradeshow visitors who want your newsletter, doubling your little B2B distribution list overnight and lifting responses temporarily?
Did you change the copy on your sign-up forms to better explain the kind of emails you send? New subscribers are now better matched to your content and offers: response rises.
Old subscribers
Did you launch a subscriber preference center and highlight it in the last email? Maybe your segment got a few fresh faces as a result.
Did you just finish a major subscriber reactivation campaign, with the new actives now helping lift your results?
Did you undergo a major list hygiene project, removing all those bounces and eliminating lots of unresponsive addresses? Total responses won’t change, but response rates will rise: when you take out the deadwood, total responses are divided by fewer addresses to calculate percentages, so those percentages go up.
Email software changes
Things may not have changed at all at your end, but what about at the subscriber’s end?
Migration to alternative viewing devices, software or webmail services doesn’t happen overnight, but software and service upgrades do. These can change the way your emails are displayed or processed by a significant chunk of your list.
On January 12th this year, for example, Google began switching Gmail users over to a more secure https interface setting. The result was that emails with images not hosted on a secure server triggered security warnings when users tried to open them at Gmail. I discovered a resultant dip in opens and clicks on my own list.
Did any of the big webmail services upgrade their interface, features or settings prior to your last email?
Reverse engineer
Of course, once you become aware of the various factors that might impact your results, you can plan for them…and exploit them.
If TV campaigns boost email results, then time the latter to follow the former. If a national election keeps people off email, don’t send email during a national election. If the weather plays a role, put out your umbrella promotion just as torrential rain hits the country…forewarned is forearmed.
So…any more additions for the list?