Interesting. Why?

To coincide with an upcoming event or promotion.

I hear you; there’s only one chance to get the test live in time for the promo.

If it isn’t ready in time, we can’t launch it at all - and we spent all that time and effort on it (#sunkcostfallacy).

And it shouldn’t have taken this long anyway (#planningfallacy).

Is it possible to run a successful promotion without the test? Probably. But is it possible to run a successful promotion if we launch a broken test?

We haven’t hit our target number of test launches this month.

“Number of tests” is a terrible success metric. Assuming we’re stuck with it, though, can we scrap this experiment and do something with button colors instead? 😉

We wanted to launch it last week, and it’s still not live.

So frustrating. But for all the same reasons we don’t launch on Fridays, we shouldn’t launch in a rush. We’re frustrated now, but think of how we’ll feel if we ship a broken experiment.

We do everything in a frickin’ hurry.

Thanks for your candor. You shouldn’t be testing. And I have somewhere to be …