4 ms·

I have seen the results of these A/B tests closely on a major consumer site and I can tell you it definitely hurts the business to make cancelation really easy.

TheCoelacanth 1y ago
How do you A/B test your company's reputation as being difficult to cancel? You can't exactly serve up different word-of-mouth to different users.
- mrguyorama 1y ago
  Believing that "reputation" actually matters for American businesses is laughable. Craftsman tools are STILL riding the reputation they had half a century ago, despite being made out of the cheapest chinesium and losing their impressive warranty stance. The American consumer has demonstrated an absurd inability to consider past events as useful information to predict future results. Things continue to enshittify because the 3 consumers who recognize that quality is going down are vastly outweighed by the increase in consumption by the rest of your market. Kitchenaid still sells plenty of mixers that die after a year. Hell, American car brands are still successful businesses even though they have made only a few reasonably competitive vehicles since the 50s. Disney and Netflix are still making plenty of money despite making it difficult to share accounts.
  - frollogaston 1y ago
    The "Toyota tax" is entirely based on reputation, and Amazon and Costco both benefit a lot from customer trust.
  - gosub100 1y ago
    Comcast
- const_cast 1y ago
  This is the danger of data-driven decision making. You can only gather a very, very small subset of all data. So now you're basing your decisions off of a tiny picture, so you end up with sometimes strange conclusions. Conclusions that, intuitively, make no sense. But the data says so, so I guess that's what we do.
  - reverendsteveii 1y ago
    I mean, doing it wrong is a danger in any process. Data driven decision making should do every meaningful thing it can to ensure that its small subset of data is representative of the whole picture. If you're careful about being random you can do that pretty accurately with a shockingly small subset of data.
    - const_cast 1y ago
      In my experience, data-driven decision making is typically a reverse engineered process. Typically the decision is already made, usually based on a combination of experience, ego, and emotions. Then you have to find and structure data to justify it to other people. You can find data to justify almost any decision.
      - reverendsteveii 1y ago
        How are you sure this isn't a data-driven opinion?
porridgeraisin 1y ago
Yep... I've been in a meeting where we were shown the result of moving a cancel button's position on the page to a more crowded place so it would be noticed less. It actually works people click on it less. I couldn't believe it. Thankfully, the feature got vetoed and cancelled (the end result was really visually horrendous).
lostlogin 1y ago
How do you A/B test the OP when they won’t sign up due to a perception that they can’t cancel?

TheCoelacanth 1y ago

How do you A/B test your company's reputation as being difficult to cancel? You can't exactly serve up different word-of-mouth to different users.

mrguyorama 1y ago
Believing that "reputation" actually matters for American businesses is laughable. Craftsman tools are STILL riding the reputation they had half a century ago, despite being made out of the cheapest chinesium and losing their impressive warranty stance. The American consumer has demonstrated an absurd inability to consider past events as useful information to predict future results. Things continue to enshittify because the 3 consumers who recognize that quality is going down are vastly outweighed by the increase in consumption by the rest of your market. Kitchenaid still sells plenty of mixers that die after a year. Hell, American car brands are still successful businesses even though they have made only a few reasonably competitive vehicles since the 50s. Disney and Netflix are still making plenty of money despite making it difficult to share accounts.
- frollogaston 1y ago
  The "Toyota tax" is entirely based on reputation, and Amazon and Costco both benefit a lot from customer trust.
- gosub100 1y ago
  Comcast
const_cast 1y ago
This is the danger of data-driven decision making. You can only gather a very, very small subset of all data. So now you're basing your decisions off of a tiny picture, so you end up with sometimes strange conclusions. Conclusions that, intuitively, make no sense. But the data says so, so I guess that's what we do.
- reverendsteveii 1y ago
  I mean, doing it wrong is a danger in any process. Data driven decision making should do every meaningful thing it can to ensure that its small subset of data is representative of the whole picture. If you're careful about being random you can do that pretty accurately with a shockingly small subset of data.
  - const_cast 1y ago
    In my experience, data-driven decision making is typically a reverse engineered process. Typically the decision is already made, usually based on a combination of experience, ego, and emotions. Then you have to find and structure data to justify it to other people. You can find data to justify almost any decision.
    - reverendsteveii 1y ago
      How are you sure this isn't a data-driven opinion?

porridgeraisin 1y ago

Yep... I've been in a meeting where we were shown the result of moving a cancel button's position on the page to a more crowded place so it would be noticed less. It actually works people click on it less. I couldn't believe it. Thankfully, the feature got vetoed and cancelled (the end result was really visually horrendous).

lostlogin 1y ago

How do you A/B test the OP when they won’t sign up due to a perception that they can’t cancel?