But what does it really mean?
Well I guess most People assume I might refer to “cheating” or “black hat” techniques, that’s why I always follow it with the slide “Always take the straight and honest road”.
That’s a core, non-negotiable value of our team but I won’t be talking about that today :)
What does Whatever Works really mean?
Let me take a step back in order to explain it. We try very hard to make most of our decision based on hard data. To do that, every time we make a new decision we define the situation, state our assumptions, design an experiment and test it.
Gosh – this is complete theory crap – so i`ll share an example to help those of you that are committed enough and haven’t fallen asleep by reading the theories above.
SituationWedding related Services (Photographers, DJ, Catering, Flowers, Sweets etc) are one of the most important categories on AtYourService.
AssumptionsI think our landing page really sucks for people who are getting married.
I mean look at it. No mention of weddings except one word in the Category “Events & Weddings”, lots of Handymen and other weird looking cartoons.
I say we change it and build something completely different – something that looks more like a wedding website – something more like this!
HypothesisYup. This is beautiful and will convert like 10 times better. That’s what all the good sites are doing, so while we are at it , let’s create one such page for each important Category of our site .
Now that’s how most teams operate. Someone (usually the one with the bigger title or the louder voice) makes up a theory, convinces the team and off they go for changes. In most cases (as in the above case) the theory sounds very convincing so the team buys in the idea and starts working on the changes.
That however, is not how the AtYourService Team works
The above situation is a real example, as you can probably imagine by the landing page. But before rushing into changing our whole site we look for evidence that supports our Hypothesis. And to do that we design experiments
Experiment or A/B TestOne of our biggest source for people looking for Wedding Services are Google Ad Words. Let’s redirect half of them to the new page, and half to the old page until we have significant visits (this is a topic for another day) and see if the new page is better.
What does better mean? In our case it’s pretty simple! If users post more Jobs after seeing this page.
So we run the experiment (even though of course we were sure that the dedicated landing page would do better).
But oh… Wait… Really?
How can this be. I mean a dedicated landing page must always convert better than a general one. At least that’s what the theory says. But evidence is clear in our case
Cost per Converted Click is €23.42 for the General Landing page and €39.34 for the dedicated landing page. And click conversion rate is 1.65% vs 1.1%.
Whatever WorksSo that’s Whatever Works means. Every person in a team has (and rightly so) an opinion of what is the best option. Teams can either decide democratically, hierarchically or by actually trying to understand what is the best option.
As a team we choose the latter.
We do a lot of Analysis and A/B testing and if you people enjoy this type of post I will continue writing about our experiences, and about real dilemmas we have.
The reason I am writing a post is to begin a nice talk about best Practices for startups so please write your comments (especially if you disagree), share and let’s help build our ecosystem :)