Rich-text Reply

Ideas to get VP's and other business stakeholders to buy into a culture of optimization testing?

jpfotoz 12-23-15

Ideas to get VP's and other business stakeholders to buy into a culture of optimization testing?

At my company we seem to have a fear of testing.  The fear is that the testing will cause a loss in lead generation and thus is not a good idea.  Any ideas on how to get them to jump in the testig pool and see what happens and put their fears aside?

Optimization Hero!
baconipsum.com - because everyone need
a little bacon in their text!
Level 6

robertchan 12-23-15
 

Re: Ideas to get VP's and other business stakeholders to buy into a culture of optimization testing?

The best way in my opinion is to demonstrate value. One way I've gone about demonstrating value has been by getting our properties' CPMs and creating custom goals in Google Analytics that I then tie into an experiment. I have then integrated Google Analytics with Optimizely to show how testing has the potential to increase conversions by creating an analysis spreadsheet with the values reported.
Robert Chan

Experimentation Hero
MartijnSch 12-24-15
 

Re: Ideas to get VP's and other business stakeholders to buy into a culture of optimization testing?

Tell them that the risk of not testing is bigger than with testing. It depends a bit on how much volume you'll have but that usually makes things more clear. To prove this point it sometimes work just to show them a couple of tests that a random company has done before so you could have them guess what the winner is. Chances are very high that half the team disagrees with each other which will already make your case more easily and because of that most people will chose the wrong variant. Making it clear that most of the times that they're making the wrong decisions by not trusting data.
scleveland 01-06-16
 

Re: Ideas to get VP's and other business stakeholders to buy into a culture of optimization testing?

I lean to agree with what MartijnSch had said.  That the risk of not testing is bigger to not test.  Our sites aren't perfect.  If we think that everything we do is exactly what our customers want, we are simply foolish.  You can always improve, you can always do better.  If you don't try anything, yes you probably won't hurt the business, but you will never grow the business.   At least with testing, you have unbias testers that are telling you what they like and don't like based on their behaviors.

 

One of the things we did when we first started testing and getting our VP & CEO on board was, track everything.  That way we could see the big picture of what was being affected by any changes. 

 

Plus we met about what tests we wanted to run and weighed the risks.  For example if we were running a test on our cart area.  At first it's kind of scary b/c that's how we make money, if we deter people from there, sales suffer.  So we scaled back the amount of traffic, knowing that it would take longer to get a statistical winner.  But at least they were willing to do it.  We also did smaller tests with smaller changes to get them used to the process, reporting, etc.  When they started to see positive results, or even when they were down how easily it was to stop a test.  They were more likely to do more tests.

 

Ultimiately, everything came down to the data and results.  The biggest things they wanted to make sure was that our site conversion didn't drop.  I always had to to know where our site conversion was at, and if it dropped due to the test to stop it and rethink.

Re: Ideas to get VP's and other business stakeholders to buy into a culture of optimization testing?

Oh no! That must be frustrating. The key phrase you used is "jump into the pool" of testing. Maybe instead of asking them to jump, you can convince them to dip a toe in.

It sounds like they really are afraid about loss and control. You can address those in different ways.

Minimize potential for loss...
Losing leads can be a very big deal, so it's important not to minimize that fear. One idea is to solve something that everyone agrees is a problem already, rather than trying to improve what they already believe to be working well (or are afraid to risk losses on). For example, if everyone agrees that a particular channel of traffic is performing poorly relative to others, you may be able to run a test there, to see if you can get a lift in results.

Provide confidence in controls...
Try to get an example of the scenarios they are afraid of. Referencing those scenarios, explain the different ways you have to control who sees the experiment, how you can turn it on or off, the kind of data you have to make decisions, etc.
---------------------------------- - - - - - ➳
Vanessa

Revere.ai Monitor your Optimizely experiments