This whole idea of "negative testing" actually gave me a lot of food for thought.

The philosophy behind user experiments (A/B, A/A and all the variants in between) is usually that you want to try out new features assuming that those features will have a *positive* impact on the user experience. That is "positive testing".

For example, you have a UI with a date picker made of multiple <select> elements. You believe that a calendar-like date picker will be easier to understand and use? Just wrap it up an A/B experiment. If more users are likely to go ahead with the process after being shown the new picker over the old one, you have proved your case - and improved the user experience.

"Negative testing" goes in the opposite direction. You want to provide users with a different experience, under the hypothesis that that new experience will be *worse* for the user - for example, by removing popular features, hiding relevant content, or draining the battery. And you want to monitor how users change their behaviour in this new environment.

The proponents of this approach apparently argue that the collected data provides insights that eventually make the overall experience better for the majority of the users. But is this really a price worth paying?

Would you appreciate it if your car's speed was suddenly capped while driving on a highway, or if the connection to the brakes was temporarily interrupted, so the producer of your car can collect better data on how drivers react under conditions of stress?

If we wouldn't accept this in the real world, why should we accept it in the digital one?

The purposeful crafting of a worse experience for some customers, no matter the goal, is the absolute negation of one of the founding principles of a capitalist market - that the customer is always right, and it should be the mission of a business to be in good faith and to provide a good user experience.

And the fact that Facebook can get away with literally draining the battery of your phone for their experiments (and even much more than this) is the negation of another founding principle of a capitalist market - that in a competitive market users should be able to just walk away to another comparable competitor if their user experience deteriorates.

In an ideal market, the only sensible response to the question "how do users change their behaviour if my app starts draining their battery?" should be "users will stop using your app and switch to a competitor". The fact that this is not the most likely outcome of their experiments proves that our economic system is deeply flawed.

@blacklight

Hayward said, "I have never seen a more horrible document in my career."

I have.

I worked for a company that did software for the insurance industry. One insurance company wanted a program which would be triggered when a policyholder's claims exceeded a certain dollar amount. The program would generate an Excel spreadsheet which had a checklist of things they would do to deny and delay claims.

Evil. Pure evil.