As an SEO copywriter, you’ll get better results, more traction, more visibility, and more conversions if you use a testing and data analysis process to track and improve the results of your writing.
You’ll build a stronger portfolio, keep your clients happier, and, without putting too fine a point on it, it’s good for your bottom line.
So, read on, and find out how you can leverage data to improve your copywriting.
Testing
It all starts with data collection. You have an hypothesis, or you want to try something new? Put it to the test.
Here’s how it works.
Let’s use a simple example. Say your CTA (call to action – a request of your readers to engage in a particular way) is for a newsletter signup. Should your button say “Join Me” or “Sign Up”? Which of those two is more likely to get someone to give you an email address?
Try both!
Set up your page so that every user will see one of the two buttons, but not both, and, after thousands of users have seen their buttons (either clicking or not clicking) and you’ve got a data set of sufficient size, chart your results and see which button got more clicks.
It’s that simple.
It can be as simple as a pie graph, or as complex as a context-heavy report (maybe “Join Me” worked best for mobile users on iPhones, while “Sign Up” was better for desktop, Android, or more conservative users.) Just let the data tell its story, and use it to inform your next move.
This has the happy bonus of bypassing the sticky pitfalls of what Optimizely (a testing platform) calls “HiPPO Syndrome.” See, marketing as an industry lends itself a little too easily to personality cults, so if Jane Q. Moneyton, marketer extraordinaire, expresses an opinion, that view makes the rounds almost as gospel, and the data be damned. HiPPO Syndrome describes the undeserved weight given to the HIghest Paid Person’s Opinion. The data can’t be swayed, and it can’t lie. If the test was fair, the results will speak for themselves.
Implications
The actual in-practice, day-to-day stuff is pretty simple, really. Just incorporate a more data-driven approach to your benchmarks and metrics, and you’re good to go.
The weird part comes in the ways people conceptualize what we do (and you’ll be explaining it to a lot of clients who share a sort of root misconception). The idea of “a writer” hasn’t changed much in the past century, in the popular conception. Between Wilde, Gibbon, and Hemmingway, we’ve covered the dry wit and romanticism, the prolific need to document and transmit information, and the excursus on the human condition.
As a writer in the Information Age, you need to add scientist to the mélange. For the first time, writers are publishers, marketers, and analysts, all at once.
As a digital marketer and SEO copywriter, I’m always thinking about numbers. I think about the relative stats for different potential keywords, or the conversion rates and page-view durations of one post over another. My writing is informed and shaped by the data I collect from how my work is received.
But I hope you’ll see that this is just another way of looking at an age-old fact. If you write stuff people will want to read, more people will read it, and you’ll be asked to write more stuff. It’s just another way of tracking and measuring the traditional relationship between author and audience. (Audience, by the way, is different from viewership. Viewers see. An Audience hears.)
Pitfalls
Testing has a nasty habit of lending itself to bad-practice. It’s too easy to implicitly defend mediocrity by watching it outperform other mediocrity. Bad writers bang out a few half-assed copies and pit them against each other. Good writers, though? Good writers pit their best against their other best to see which one is best for a particular audience.
A rule of thumb? Testing should be used to find out which is best for a specific group. Everything that makes it into the testing phase should be best for someone.
Where Do We Go from Here?
Before you can test content with audiences, you need some basic groundwork.
You need a forum, like a website, to publish content. You need enough traffic to make up for sampling errors. It’s not useful data if two people came and one clicked a button.
You need a bare-minimum of engagement before you can say anything useful.
So how do you get traffic?
You can check out our blog post on Off-page SEO for more on this, but here are some quick tips:
-Use social media
-Blog consistently
-Guest blog and interlink as much as possible
-Don’t shy away from a PPC model to get the word out
So, there you have it.
Don’t wait – get started today. If you’re not sure about split-testing, try something out. Add a button for a newsletter, and see what happens with the traffic on your own site.
And find us on Facebook or Twitter to let us know how it goes!
