Ratio of sellers to buyers here

[misscrystal]

From their sample of gigs they used in their analysis, they counted 531,841 buyers.

That’s not (or wasn’t the total number on the site) but if what they say is correct and the counted it correctly that, with the count of sellers they did, should be enough for a reasonable estimate of the percentage of sellers vs buyers (or sellers to buyers ratio). Though it also probably doesn’t take into account sellers who are also buyers.

I’m only saying what they said in their analysis of 2016. I can’t give an up to date figure and I can’t actually check and it would most likely be against the terms of service. If Fiverr ever gives me permission I could probably count them though.

they counted 531,841 buyers.

How do they know how many buyers??? How many buyers have I gotten in the past month? Can you see that somewhere on my gigs that I’m not aware of?

How much have they paid? Is that apparent?

[uk1000]

they counted 531,841 buyers.

How do they know how many buyers??? How many buyers have I gotten in the past month? Can you see that somewhere on my gigs that I’m not aware of?

How much have they paid? Is that apparent?

They crawled the site looking at 10 categories and 103 subcategories and their software did that for 10 days (looking at as many gigs it could find in that time - while likely not doing it too fast).

eg. their software probably went to the Fiverr homepage and looked at all the links on that (or if there is a sitemap they could use that file/files) to look at first page of each category. Then the software could go through each page of the categories/subcats and for each of those, load the gigs for each of those.

For each gig it could keep a list/total of buyers and sellers, where the buyers were obtained from the reviews (obviously it would only be able to see the buyers who had left a review) and the seller would show on the gig page.

While it wouldn’t show the exact payment like I said, it would just go off the reviews and gig prices and gig package prices if it had packages. So the revenues wouldn’t be totally accurate.

But the revenues aren’t important for finding out the number of sellers vs buyers.

[misscrystal]

They crawled the site looking at 10 categories and 103 subcategories and their software did that for 10 days (looking at as many gigs it could find in that time - while likely not doing it too fast).

eg. their software probably went to the Fiverr homepage and looked at all the links on that (or if there is a sitemap they could use that file/files) to look at first page of each category. Then the software could go through each page of the categories/subcats and for each of those, load the gigs for each of those.

For each gig it could keep a list/total of buyers and sellers, where the buyers were obtained from the reviews (obviously it would only be able to see the buyers who had left a review) and the seller would show on the gig page.

While it wouldn’t show the exact payment like I said, it would just go off the reviews and gig prices and gig package prices if it had packages. So the revenues wouldn’t be totally accurate.

But the revenues aren’t important for finding out the number of sellers vs buyers.

You can’t tell how many buyers each seller has nor what each buyer paid, certainly not by looking at reviews. However if you want to believe that report that’s your decision.

You don’t know the number of buyers nor what each one paid but that’s ok, let’s pretend it’s possible anyway. :roll_eyes: Let’s pretend it’s possible to have software that crawls each seller to see how many order they have each day, or each week. Maybe I’m wrong, it’s not important.

[uk1000]

You can’t tell how many buyers each seller has nor what each buyer paid, certainly not by looking at reviews. However if you want to believe that report that’s your decision.

You don’t know the number of buyers nor what each one paid but that’s ok, let’s pretend it’s possible anyway. :roll_eyes: Let’s pretend it’s possible to have software that crawls each seller to see how many order they have each day, or each week. Maybe I’m wrong, it’s not important.

Not the exact number but you can get quite a good estimate. It should be easy to tell if someone is a seller (they have a gig) and whether someone is a buyer (they’ve left at least one review). Not all buyers will leave reviews but for a reasonable estimate of percentage of sellers vs buyers it should be good enough. There may be other ways to see if someone is a buyer only if they haven’t left a review (if it’s possible to retrieve users by eg. userid) and see if they have any gigs or if there’s something else in the public page that shows that.

[timothywcrane]

One thing that is becoming clear to me, both in the data sets and the thread, is that little attention has been paid to what a buyer and seller by definition really is. If we are talking about actual sales, this would be totally different than the ratio of buyer to seller defined accounts. This latter account ratio is increasingly harder to gauge as in order to do opposition research as a seller is to “log in as a buyer”. Let us actually define the problem a little more, and then do some more digging.