29

Fuck Optimizely.

Not because the software/service itself is inherently bad, or because I don't see any value in A/B testing.

It's because every company which starts using quantitative user research, stops using qualitative user research.

Suddenly it's all about being data driven.

Which means you end up with a website with bright red blinking BUY buttons, labels which tell you that you must convert to the brand cult within 30 seconds or someone else will steal away the limited supply, and email campaigns which promise free heroin with every order.

For long term brand loyalty you need a holistic, polished experience, which requires a vision based on aesthetics and gut feelings -- not hard data.

A/B testing, when used as some kind of holy grail, causes product fragmentation. There's a strong bias towards immediate conversions while long term churn is underrepresented.

The result of an A/B test is never "well, our sales increased since we started offering free heroin with every sale, but all of our clients die after 6 months so our yearly revenue is down -- so maybe we should offer free LSD instead"

Comments
  • 9
    This. Somebody explain this to gaming companies too please.
  • 2
    @RememberMe
    The problem is, that in "gaming" (actually, just casual gaming, but lets forget the remaining 0.1% of the market), gambling mechanics really just work best.
    Depending on what you sell, it even "just works" outside "gaming". Almost always it works at least better than any high-quality approach.

    Shitty candyman tactics and dark patterns aren't overtaking the web because they don't work.

    Society will adapt at some time. But right now, ads work, dark patterns work, cheap psycho tricks work, and popping newsletter subscription dialogs over the content works too.
  • 3
    @RememberMe after being a game dev I honestly just wish 90% of that industry explodes... It's so full of bs
  • 0
    @Oktokolo With games it's very visible indeed, the whole Skinner box routine is pretty sickening.

    But in a lighter form it's also present in websites and other products: The main "thing" (Let's say a bicycle rental search engine) gets tweaked every few days. It looks flashy and modern, although way too busy with redundant elements, due to all the little experiments which are trying to get you to click on a result. Then you go to your account page, and it's an anachronistic bootstrap relic thrown together without any design going on.

    The problem with this?

    In the short term you might get people who visit your site to actually click on the result page and make a reservation, netting you a commission.

    But people won't FEEL any "wow, that was an amazing, smooth working beautiful website" warmth in their belly.

    You see this a lot with travel websites, which lure customers in with predatory tactics and have great conversion rates — but poor aftercare and zero brand loyalty makes those websites extremely replaceable.
  • 0
    Bro, just A/B test, just put that lil candy next to the sign up form. Users are more inclined to click the huge "GIVE ME CANDY" button if they see an extra lil candy cane on the side. Just do it bro FFS!
Add Comment