Every time I’m asked to deliver a presentation on “best practices in email,” it makes me sigh. The reason is simple: Most of the best practices out there are fairly generic and low-impact. They include tactics such as:
Create a preference center.
Allow people to opt down (they may not want all your email).
Use cascading style sheets (CSS) to create responsive emails.
Employ more triggers.
Create a welcome stream.
Have a double opt-in.
They’re all good practices, but I’m not sure they are best practices. All of the above are fairly generic and speak to check-boxes in some theoretical or codified maturity model for email practices.
But I have yet to meet an email marketer who was measured on a maturity model or their adherence to best practices. Much of what I have learned can be distilled into a few tenets:
Math is your friend.
Go beyond the email channel for segmentation and reach.
Creative matters.
Let Math Carry Some Of The Load
As direct marketers, we depend on math to make our living. It’s pretty simple: Campaigns and programs either work or they don’t.
The only way to determine success is math, which in the best case measures a program’s influence on sales or actual sales directly attributed. So if we all have to rely on math to make a case for success, why are we so reluctant to use it in the actual marketing execution?
Today, a lot is done manually that doesn’t have to be, including:
• Content testing. Use multivariate testing to determine the winner. A/B tests are nice and easily compare two or a few things.
A/B tests are very manageable from a setup perspective and sometimes are the only available option, even with some major email service provider (ESP) platforms. But to really get to the bottom of things and use the best creative combinations, you need to use a number of variations, not just a few.
Whether you favor a Taguchi-related method or another statistical model, the best practice is to use available services, such as 8 Seconds, to handle the math and determine the winner.
It is email. You could do it yourself, but why? It’s like being taught how to calculate standard deviation — nice to know, but there’s a function for that, and it’s more important to know how to leverage the results, not perform the calculation.
• Scheduling a campaign. My absolute least favorite best practice is a determination on which day is best to send a campaign. There is no best day or time to send a campaign that works for every company for all recipients. Many companies send nearly every day already and see positive returns from each send.
There are several good reasons to launch a campaign at a specific time. Among them are a very short sale window (same day) or when customer service is open, which is especially important if you’re doing some sort of account summary email because it drives a lot of call volume.
But for the most part, emails aren’t that time-sensitive, nor do they drive that kind of call volume. Let send-time optimization take over the decision of when to send email to someone within a specific campaign.
Many ESPs have some sort of send-time optimization technology. Use it. It’s generally worth it as long as it’s easy. If your ESP doesn’t offer it, AudiencePoint is an alternative.
But one can go further. There are several companies that utilize enterprise-level decisioning on when to send a campaign to an individual based on all the data available to them.
• Subject lines. Why are we still coming up with a single subject line? As an industry, we still don’t spend enough time on the words that draw people into our emails.
Subject line testing should be a part of every campaign. If it’s not, you’re bunting on the campaign. It’s just not that hard to do.
Add to this the machine learning that’s now available. Using a company like Persado, one can have 16 subject lines generated automatically, whose results drive a 17th new subject line designed to drive opens.
Segmentation Should Draw On More Than Email Data
A lot of email segmentation is based on data from email campaigns, such as opens and clicks. It makes sense; if you don’t have an integrated marketing database, this is the only available information — and it’s better than nothing.
But it’s not a best practice. True best practices in today’s world of advanced analytics and technology include:
• Using offline transaction data. If you know that someone recently transacted in a store or tends to buy from you on a particular cycle, that information should trump any arbitrary deliverability rules regarding inactivity.
It’s true that they may not use the email address that you have for them, but shouldn’t you take a shot?
If you’re concerned from a deliverability perspective, use a cleansing service to test the mailability of that address to reduce risk.
• Using site data. This is not 101-level segmentation, nor does it refer to the use of site behavior triggers that seem to be all the rage these days (and that are great for their intended purpose). Many companies have connected their site data with that of known customers and prospects — but have done nothing with it.
The best practice is to deliver dynamic content using site behavior data to select who is in a campaign or what someone sees in an email. Doing this requires delving into your data deeply and understanding the relationships between the known data and site data.
But once applied, the rewards can be twofold: more revenue with better efficiency in selection.
• Using environmental data. One of the easiest opportunities for email marketers is to leverage environmental data to personalize the message. By leveraging one of many available services (e.g., Movable Ink, LiveClicker, PowerInbox), the email marketer can determine operating system and relative location (via IP).
Whether the application is to show a local map, reinforce a local sports team, adjust creative for weather, show a particular app bazaar or some other treatment, environmental data drives insights into mass-buying patterns. Maybe we haven’t gone to bat for the budget to use it, but it allows us to leverage location and technical capability to provide an advanced message.
Forget Email
We all like email. Yay for us. But during my career, I have found that as much as 60+ percent of your list can be inactive with your email program for a year or more, yet still be likely to interact with you on the Web.
The best practice today is to use that email list for targeting and lookalikes in other media and channels; the most common are Facebook, Twitter and Google.
The basic point here is that you can create reach and frequency on your subscribers outside of email. It takes a lot to do it right, like ensuring the security of personally identifiable information (PII), avoiding creative fatigue and nailing measurement, but it’s a huge opportunity to connect with your customer base.
The Presentation Layer
Math and technology are great, but you need to bring art into it, too.
No matter what goes on behind the scenes, people react to what is being said, sent and presented to them. That comes in the form of words and images.
The words and images you select matter. Their placement matters, as well. The best practice is not the “reverse Z” nor keeping everything above the fold. Those are helpful guides.
The best practice is being clear on what you want someone to do and making sure that the creative facilitates that action. I wish it were as simple as “make your emails responsive” or “use the same navigation as the website,” but it’s not.
Companies that want to increase their revenue from email need to gain insight on what makes people buy versus what makes them buy from that brand. That needs to be applied to email and matched with an individual’s stage in the purchase funnel.
Best practices are important, but ultimately, we all do what we must to hit our business goals. And in the case of email marketing in the 21st century, that means thinking beyond what currently amounts to mediocre practices.
The most successful email marketers are redefining what’s best, using more innovative tactics to drive better results for their clients.
コメント