Item — HN Reader

You might be interested in the concept of 'satisficing'. A product needs to be significantly better to replace an existing product that is satisfactory.

Your point is slightly different though and can be seen with M16 still being the assault rifle of choice in the US military even though it's not as good as other products by most measures. That's more the IBM effect where the agent deciding which option to buy has nothing to gain by challenging the status quo. If IBM fails, it's not your fault. If you choose something else and it fails, it's your fault. Success won't be recognized either way so the obvious choice for a purchase is IBM.

2 comments

It's interesting because the difference between the jerrican and existing cans were lots of small affordances that add up: fewer tools needed and easier to move around. In the end, soldiers will find a way to put gas in vehicles, so there's never a problem from the strategic view even though several small efficiencies can add up to a big difference both for a single person gassing up a tank and then multiplied by the number of tanks that are involved in warfare.

When OS X came out about 20 years ago, the UI was so much better than Linux desktop environments of the time. I think that is a relatively uncontroversial view, but try fitting the differences between them into a checkbox-style acquisition table and they both use the basic WIMP interface; the obvious differences (icon dock, global menu bar) are if anything something that people moving from Linux to Mac dislike about it, yet those same people will tell you that the overall experience is better.

I'm imagine sunken costs fallacy applies too with the costs being tooling, supply lines, etc. Changing from smooth tin cans to stamped steel cans with internal liners does require a substantial investment in new tooling.