Questions? (877) 643-7244
FutureNow Post
Friday, Aug. 13, 2010 at 1:04 pm

Top 10 Things To Do When Waiting for GWO Results

By Brendan Regan
August 13th, 2010

We all know that waiting for test results is exciting, but somewhat tortuous. Our minds run wild, asking things like: Why is this taking so long? Why won’t Google Website Optimizer (or any other testing platform) declare a winner already? What does “statistical confidence” mean, anyway?

You experience this anxiety, FutureNow clients struggle with it, and we analysts stress over it, too. You’re not alone.

The key point is that you can’t give in to the temptation to end tests early! Ending a test without getting statistically valid results is the hugest waste of money! It’s the Anti-Optimization. I know it’s difficult, but “patience is a virtue” in conversion rate optimization just like in life.

To help you be patient, I’m presenting this handy list of things you can do while patiently waiting for your test results to fully validate.

#10: Figure out ways to avoid the forever test.

#9: Learn about statistical significance and why it’s important to let your testing platform reach it.

#8: Stare out the window. Seasons are changing, the breeze is moving gently through the trees, etc.

#7: Take advantage of the patience of strangers. Guess which test variation won on other people’s tests.

#6: Study your web analytics. Don’t bother guessing how your test will turn out. Dig into your data to gain insight that will drive your next tests.

#5: Set up another test. If you don’t have half a dozen tests running, you probably have “room” to be running a few more. Nothing eases the anxiety of waiting like having other tests to worry about ;)

#4: Build the next test. The best way to optimize your conversion rate is to have the next test already built and waiting (variations and alternate pages can be built and stored locally or in your “dev” environment). The second you end one test, you should be ready to roll out the next test.

#3: Calculate the value of a recently-ended test. If you got an improvement from a recent test, have you calculated an estimated dollar (euro/yen/etc.) value of the result? How many more sales/leads/subscriptions will your winning variation net you going forward? I’m pretty sure your boss and teammates would want to know.

#2: Kick off a persona project so that you can understand your prospects’ motivations and plan smarter tests.

And now, dear readers, the #1 thing to do while waiting for test results to come in…

#1: Take a break, already! You work hard, so spend a few moments relaxing…still online, of course :) Perhaps enjoy a giggle about parents on Facebook or laugh at raunchy, profane [YOU'VE BEEN WARNED] texts from last night?

Add Your Comments

Comments (23)

  1. I was shocked to see that Google expect you to have 1000′s of conversions to best use Multivariate testing!

  2. Great post – I normally do a mixture of getting the next tests ready and spreading the news upwards to the big chiefs.

    You can obviously learn from tests with a clear winner – but you can also learn from inconclusive tests. Maybe the inconclusive idea was good but poorly executed.

  3. What is really interesting sometimes is the version that starts out as the leader may end up being the loser in A/B testing. So it is very important to wait tests out.

    Plus the simple fact you are tests means you are way ahead of the other guy. Most people are still not testing.

  4. We did a couple of tests before but make sure you have a good sample size before making any conclusions. Make sure that the time period are roughly the same to make sure that no seasonal factors are in place.

  5. I really like #1 tip. That is really the best thing you can do while waiting. Why not take the time to visit fun sites or social media sites, you never know, you might be unconsciously gathering ideas already for your next test.

  6. An important thing to do while testing is also figuring out how to present the test results to the rest of the organisation. Yes – it is cool to found out that a change suddenly increased you conversion, but if the organisation doesn’t get it, the test might be worthless

  7. Here’s the most important thing to do while waiting out a test:

    let go of the test

    Up until now you’ve put a lot of ego into the new version. You really hope that it’s going to be a winner – it’s been your idea, your baby. You wouldn’t be trying it if you didn’t want it to succeed.

    Now that the test has started you’ve got to forget the ego bit. Step back. Let the dice fall where they fall and trust the numbers.

    This can be hard.

  8. @John Hyde: Oh…so…true! I recently spent too much time on a Variation page for a test. This caused me to check the results obsessively. And of course, my beautiful variation BOMBED. The only saving grace was that we “failed quickly,” which Avinash K. would have been proud of :)

  9. Great advice. Always good to let things take their course. Once you’ve made your efforts…let the test results be what they are, whenever they come back.

  10. I actually got a big chuckle out of your one listed item to do, and that was stare out the window the seasons are changing, that is not only funny but also oh how true. Many times Myself I have caught myself doing that very thing after putting alot of work into a website, thinking it’ll perform, we stare out the window for a days and wonder why we don’t get the results, excellent post.

  11. I try to take the time to read, I sit down away from the computer and drift off into a new book. Soon enough you have your results and some new knowledge.

  12. Nice list but when waiting for GWO results i like studying.
    There is always something new to learn.

  13. I’m with you on this one. They have way too much private info. I am switching to Mozilla. Wait, someone is at the door. . . It’s Eric Schmidt and the Google police. Quick delete my post! I was just kidding.

  14. I like #5: Set up another test. If you don’t have half a dozen tests running, you probably have “room” to be running a few more. Nothing eases the anxiety of waiting like having other tests to worry about

    nothing kills time better then concentrating on the next project as they say a watched pot never boils

  15. I’ll have to agree with Scaricare. There is always so much more to learn about affiliate marketing, so while waiting I’ll either look something up or read a new book about marketing – I recently purchased “Cashvertising” which is supposed to really great.

  16. Affiliate marketing can essentially be a lifesaver. Especially if you have a large group of business people to work with.

    Honestly–simply going to networking meetings to try to improve referrals and relationships with those that could potentially refer you business has saved a lot of companies I know.

    Especially if you live in a large metro area, every business right now is willing to build relationships for the sake of higher earnings.

  17. Good article, sometimes while waiting I almost forget that I am doing the test because it is just taking so long. We need to also think firts whether to do the test at all knowing the duration.

  18. I try to take the time to read, I sit down away from the computer and drift off into a new book. Soon enough you have your results and some new knowledge.

  19. “#8: Stare out the window. Seasons are changing, the breeze is moving gently through the trees, etc.”

    That is a great way to pass the time. I also recommond watching paint dry or watching grass grow!

  20. We did a couple of tests before but make sure you have a good sample size before making any conclusions. Make sure that the time period are roughly the same to make sure that no seasonal factors are in place.

  21. It’s better to take a break of at least a couple of days until the next test. Otherwise, you’re going to get burned out.

  22. sometimes while waiting I nearly overlook that I am managing the check because it is just taking so long. We require to furthermore believe firts if to manage the check at all understanding the duration.

  23. I do #6 all the time. I prefer to study my analytics and find actinable insights for my next test.

    Good post

Add Your Comments

 

Print this Article
Share

More articles from Brendan Regan

Marketing Optimization Blog
FREE Newsletter Sign-Up
send it once every: