[darushimo]

I think you're right here, but the qualifier "all things being equal" can't be taken lightly. Rarely if never are "all things equal." I also don't think a shift in the market context is necessary. We already have a market place where people who feel tricked or duped will have a lower opinion of the company doing the duping. dark patterns have a bad edge to them too--very few of them, when noticed by the user, leave the user actually feeling good about how they were duped.

This kinda reminds me of the talk PG gave at 2008 start up school titled "Be Good." (http://www.paulgraham.com/good.html or https://www.youtube.com/watch?v=q7K0vRUKXKc) He outlines many ways why it's not merely ethically appropriate to be good, but also a strategically superior approach to running a startup. Ease of use, simplicity, trustworthiness--these are important qualities of a biz's reputation, both internally and in public perception. And Dark patterns tend to harm that reputation.

Edit: Addendum--obviously there are monopolies or monopoly-like situations where reputation isn't important.

[dragonwriter]

> I also don't think a shift in the market context is necessary.

If the dark patterns empirically work, then a market context shift is necessary.

> We already have a market place where people who feel tricked or duped will have a lower opinion of the company doing the duping. dark patterns have a bad edge to them too--very few of them, when noticed by the user, leave the user actually feeling good about how they were duped.

If they had enough of a "bad edge" that they didn't actually work to increase the returns realized by a business--then they wouldn't be an issue, as there'd be no incentive to use them.

[PeterisP]

1) Dark patterns do empirically work, that's the whole point.

2) Market won't magically shift - they work based on homo sapiens basic psychology, and that won't change so soon. However, dark patterns can (and are) reduced by making them illegal - effective consumer protection / truthful advertising laws can mitigate them. For example, if 'I agree' opt-out boxes are legally considered invalid, then it makes sense to use opt-in subscriptions; similarly for other [mis]representations.

[jleader]

Sometimes the "bad edge" effect is cumulative over time. This can lead to a new entrant to the market suddenly gaining a lot of market share, because so many people have gotten fed up with the dominant player. So maybe the dark patterns work well, until suddenly they nose-dive, because everyone's switched to the less annoying alternative.

A/B testing only tells you what works today, not what's going to keep your customers satisfied in the long run.