5 ms·

Why Google has 200m reasons to put engineers over designers

• pavlov 12y ago
  In the world of the hippo, you ask the chief designer or the marketing director to pick a blue and that's the solution. In the world of data you can run experiments to find the right answer. "Hippo" is an acronym for "highest paid person's opinion". Clever, but are we to believe that this dichotomy represents the whole spectrum of decision-making processes? The Achilles' heel of data-driven approaches is that setting up the experiments can constrain the possible solutions too early. When you test 50 shades of blue, the answer you get will always be a shade of blue. A design-driven process could give you solutions that go beyond tweaking one value in CSS -- but that research won't be something that you can stick into an Excel spreadsheet.
  • fsk 12y ago
    If you know statistics and do it correctly, you can reduce the chance of a false positive.
• monoid 12y ago
  While I believe in testing, I doubt that the new shade of blue is solely responsible for an extra $200m a year in revenue. Even if all the testing was done at the same time, the period with the alleged revenue increase was obviously later and during a whole year. So during that year a lot of other things including external market situations influenced the revenue change. Meaning there is no clear link between the shade of blue change and the revenue increase. Ofc a change of blue worth $200m / year makes a better headline, even if it's rubbish talk.