The Website Testing Debate – Can You Test Too Much?
The other day, someone asked me a provocative question. For all of the emphasis we place on tracking, testing, and tweaking to determine what makes websites convert the most visitors into purchasers, members, clients, and customers for life, how much testing is too much?
First let me say: if you’re serious about growing your business through online marketing, you need to be tracking, testing, and tweaking.
In deciding how to present our messages, we must continually study industry best practices, then adapt them to the needs of our audiences. (Note I didn’t say, “pick your internet marketing hero and do exactly what they do.”)
The way we learn what converts the best is by monitoring visitor statistics on our websites, monitoring visitor behavior on our sales letters, tracking opens and click-throughs on the e-mails we send out, and doing split tests.
So, how much testing is too much? Are you testing too much?
The question should not be whether or not you are testing too much.
The correct question should be whether you are testing the stuff that’s worth testing.
Here are some common errors I see when it comes to testing:
- Testing just for the sake of testing. What are you measuring? What is the goal? In what way do you expect the results to improve your website conversions and grow your profits? Do you know? If not, what value will the test results have for your business?
- Testing trite things that don’t really matter. I’ve seen marketers invest enormous resources to determine whether having a headline that’s #CC0000 or #FF0000 (which are two VERY SLIGHTLY different shades of red) has an impact on conversions. Meanwhile the headline they were testing was horrible and did nothing to appeal to the reader. FIRST, get a headline that WORKS, then paint the town red.
- Testing things that don’t matter YET. Really, do you NEED separate adtrackers on all 32 order links on the sales letter (and to hinge your sanity on making sure your webmaster submits tedious weekly reports on them), when you haven’t yet figured out why average visitor time on your 14-page sales letter is 23.8 seconds and the two sales you’ve ever made were both over-the-phone?
- Running tests that won’t be supported by your website. It’s great to experiment with various ezine subject lines to improve open rates. But when they click through to the ezine to your website, will there be any reason for them to stay on your website? Do we have THAT nailed down first?
- Way too many tests going at once. When you walk, how do you get from here to there? One step at a time. If you’re building muscle mass, how do you do it? One rep at a time. Losing weight? One pound at a time. See a trend?
- Using too many tracking softwares at once. Then, you’ll wonder why the results seem inconsistent. That’s because every software uses different metrics to determine what the casual observer would think is the same thing. What matters most to YOU, NOW?
When deciding what to test, ask yourself this.
What knowledge, if you had it in hand, would move your business forward the fastest, with the least amount of effort?
Is your testing calibrated to give you this information, or is it too much?
Here’s To Getting The Results That Move Us Forward!