So here’s the thing.
I don’t really have an issue with testing, tests, experiments etc regarding Fiverr. Sure, some are annoying, others are mildly useful and many of them cause massive upsurges in the amount of (sometimes positive) debate on the forum. Great.
However, I am beginning to wonder just how useful the tests can actually be at times. After all, when you test something, you need to be able to see the cause and effect of any results or changes that occur. Put simply:
If you increase a fee for buying to check how it affects overall sales, but at the same time have bugs which stop people ordering, change the algorithm to show more Pro/more New Sellers, and cut the marketing for “normal” gigs - how can you get any kind of definitive result from the test? How do you know which change caused which effect?
Comparing to other months or even the same period in the previous year is misleading as there have been tests ongoing for well over a year now so it seems to me that the baseline is not even there to measure against.
Surely it would be better to structure tests as one at a time, for a fixed period, preceded and followed by a “fallow” period where there isn’t any testing at all. Surely that would give better insight to how the change being tested affected things?