Stop speculating; expert opinion is great though must be challenged

I have a client that we are working with on a range of projects and websites.

Through A/B testing, we have started on a value learning cycle and journey on one of the websites we are working on and whilst we don’t know exactly why certain things are working or not working, at least we know that they are working or not and can start to make some educated guesses about what to do next.

Anyway, we started working this morning on one of the other websites the client owns, one of which, we have no previous experience working with.

Fast forward six hours and I am sitting in a quick afternoon review of some of the new landing pages we have designed for the client and I’m offering ‘expert opinion’:

Open forms convert better!

Users will not consume that content!

The typography needs work!

People need price-points.

Nobody uses scrollers.

Which might be true and true on other websites, though how do we know until we test? A/B test.

I argued that the client would not want to spend (I said "waste") budget testing assumptions and expert opinion the client felt held fairly true and which I also felt held fairly true.

Though I was overruled by a majority vote at the meeting and rightly so:

1. It was not as if I was commenting on poor design or bad practice; the designs were great and I was merely offering ‘opinion’.
2. My opinion was based on past experience, though on different websites and clients.
3. No two websites are the same. Sure, open forms (as an example) do generally convert better than closed forms though not always.
4. We need a base-line and not one based on hypothesis or opinion; we need to test, prove and eliminate.
5. And that is a justified and useful spend of PPC or other traffic budget.

 

Teams and businesses need one, reliable baseline on which to base their work and arguments. No balls can be up in the air because they’re too holy to be tested or challenged.

And however much of an expert you might be, it doesn’t matter. We have to test because we need to categorically know.