Who would dare to blaspheme what is arguably the ultimate technique for optimizing one’s website copy? Surely Joanna wouldn’t let a statement like “testing ain’t all that” simply slip through in a guest blog post… And it couldn’t come from someone who earns a living helping clients run a successful testing program… could it?
It could. And it does.
Split testing is often lauded as a panacea for website optimization.
I mean, you get to (cue sinister Transylvanian voice) experiment on your visitors… test your theories… and ultimately see the impact of your copy changes on visitor behavior – using statistics to confirm that the difference is real!
Yes, testing is amazing. And no, I’m not going to try to talk you out of testing completely.
But testing ain’t all that.
Instead, stand by as I make a case for an equally powerful – and typically much faster technique – for optimizing your website copy…
Inside This Post
Fact Is, Testing Has Many Limitations
Testing is a terrific way to validate assumptions you may have about your visitors… but it can also take awhile to get there. See, testing requires traffic. It requires patience (and no peeking at the results!). It requires a testing tool and proper set-up of that tool. And it often requires careful interpretation of the results…
In my testing travels, I’ve seen startups and large companies run tests to inform their visitor theory (in other words, to learn what makes their visitors tick).
For example, you might test a new headline that focuses on your most popular product feature. The results of the test can tell you whether or not that feature (hopefully worded as a benefit!) resonates with your target audience…
And if you have a bunch of cool product features, you might choose to run a headline test using 5 variations… with each headline focusing on a single benefit of your offering.
Whether you run 5 headline variations as an A/B/…/F test, or run 5 separate A/B tests sequentially, it’s a perfectly valid way to learn about your website visitors.
But here’s the rub: Unless you have sufficient traffic to support testing those 5 headline variations, you may be watching the data roll in for quite some time. Weeks or even months may pass – all in the pursuit of informing your theory about what product features resonate best with prospective customers.
I Don’t Want To Give You The Wrong Impression Here
There is absolutely nothing wrong with split testing to learn about your visitors, but it’s typically not the most efficient way to get inside your visitors’ heads.
There are other ways to learn what will resonate with prospective customers. Ways that don’t require a ton of traffic or time to execute. In fact, you can start tomorrow and have answers within days.
To find out what will resonate with your website visitors, look to your existing customers. Learning what led to your customers’ decision to buy from you will inform your visitor theory more quickly (and I’d argue more reliably) than any series of split tests. More specifically, you want to understand why your customers chose you over all other competing solutions (or chose you over doing nothing at all).
How can you learn from your customers?
(It’s Not as Terrifying as It Sounds)
There are a couple of ways to go about interviewing, neither of which requires you to be a trained interviewer. You just need to know what questions to ask.
OPTION 1: Interview Your Customers Directly
Whether it happens on the phone or face to face, this is a fun and effective way to inform your visitor theory. Like usability testing, where as few as 5 participants can reveal 80% of the problems on a website, you can learn a TON about your customers from a handful of discussions… and then apply what you learned to persuade your website visitors that they too should buy from you (like the rhyme?).
OPTION 2: Interview Your Customers Indirectly
By indirectly, I mean interviewing your customers without speaking to them. The most common approach is an email-driven survey. But I hate the word ”survey”. Surveys are cold and impersonal, and they have abysmal response rates. Nobody wants to take a survey, but people like to be interviewed. So call it an “interview” when you invite your customers to respond.
Interviewing Is Easier & Faster Than A/B Testing
(Especially with These Qs to Guide You)
Compare the interview approach to split testing… where each test should be designed to reveal 1 thing about your visitors (remember that split tests are univariate, meaning you test a single variable per test)… a well-designed customer interview can generate a ton of useful insight, all from a single exchange!
Below is a sample 12-question interview that has worked well for us. We’re sharing it in the hope that it’ll kick-start your decision to give this approach a try.
You can use it as is, or modify it to suit your business. Try it out with some customers on Skype, or turn it into an email interview using your favorite survey (ugh!) tool.
1. Thank you for choosing us! How long have you been using PRODUCT/SERVICE?
- I’m on a free trial
- Less than 2 months
- 3-6 months
- 7-12 months
- 1-2 years
- More than 2 years
2. How long ago did you first realize you needed a solution similar to what you’ve purchased from us? What was going on in your world that caused you to go looking for a solution?
3. More specifically, what problem were you trying to solve?
4. After realizing you needed solution, what did you do or where did you go to begin the process of getting it?
5. How did you first hear about us?
- A colleague/co-worker
- A friend/family member
- A search engine
- One of our partners
6. What are the top three (3) reasons you chose us over others like us?
7. What one (1) feature were you MOST EXCITED to use when you got started with PRODUCT/SERVICE?
- Feature 1
- Feature 2
8. What one (1) feature do you find yourself USING MOST?
- Feature 1
- Feature 2
9. What 3 words would you use to describe the benefits of PRODUCT/SERVICE to your friends or co-workers?
10. Have you ever used one or more of our competitors?
11. If you don’t mind, please tell us the name(s) of the competitor(s) you’ve used.
12. What do you think we can learn from our competitors?
That’s it! If you use the questions in a phone interview, you can simply ask the questions – don’t worry about reading out the full set of responses on questions 1, 5, 7, and 8.
Shoot to complete 5-8 direct interviews or 12-15 email interviews. You’ll need more email interview completions because the open-ended responses will be less thorough.
Once you’ve completed a series of interviews, look for themes in your customers’ responses. Themes will give you a level of confidence in turning the interview responses into new website copy. Also look for language that you can “steal” from your customers and use directly on your site.
If you like – remember, we still think testing has its place in your optimization efforts – you can always validate what you’ve learned from your interviews in a split test. 🙂
The trick (and time saver) with this approach is this:
You’re not experimenting your way to that learning.
You’ve taken a shortcut by engaging with your customers.
Think I’m crazy… like a fox? What kinds of questions or techniques have you used to learn what makes your visitors tick? Leave your advice below!