I once worked on an infrastructure-critical financial application, that was super time sensitive as in, when the user clicked “calculate”, it was critical the they got the results back ASAP.
We worked really hard to ensure that the app would work 100% of the time, as well as that the underlying infrastructure would leave plenty of headroom to be able to a ridiculous amount of queries at once, and still return the correct result in less than 100 ms 95% of the time.
Then, we were asked to artificially insert a 3-5 second sleep function, because focus groups revealed that people had a hard time trusting results that appeared faster than that…
Although, later on, we discovered that, instead of doing the calculations all at once, we could show the user the intermediary results as he was typing in the information. Then when he had typed in the last number, the fact that the result was already on screen would be less jarring, not to mention he could double check the math as he went along.
I once worked on an infrastructure-critical financial application, that was super time sensitive as in, when the user clicked “calculate”, it was critical the they got the results back ASAP.
We worked really hard to ensure that the app would work 100% of the time, as well as that the underlying infrastructure would leave plenty of headroom to be able to a ridiculous amount of queries at once, and still return the correct result in less than 100 ms 95% of the time.
Then, we were asked to artificially insert a 3-5 second sleep function, because focus groups revealed that people had a hard time trusting results that appeared faster than that…
Although, later on, we discovered that, instead of doing the calculations all at once, we could show the user the intermediary results as he was typing in the information. Then when he had typed in the last number, the fact that the result was already on screen would be less jarring, not to mention he could double check the math as he went along.