You’ve touched on one aspect of how we calculate it. We certainly bias it towards recent sales.

However we don’t document it publicly, and I’m not going to here, principally because we reserve the right to tweak the formula as to how font popularity is exposed and utilized. It’s a tricky thing to compare the popularity of an expensive 20-weight text family that sells a handful of copies, with an inexpensive display font that sells in the hundreds per week. Yet even within a genre and a price bracket, there’s no single answer to the question of the best formula to use.

I’d be interested to hear how other distributors do it, and even how foundries determine their own “bestsellers”. How much cash did this family generate this week (or other period)? How many weights have been sold? How many sales did this family play a part in? Are last month’s sales part of this font’s current popularity? If so, what about the month before last? If so, to what extent? Are the families as organized by the foundry or by the distributor the best aggregation units for popularity scores? Perhaps we should build a grid with hundreds of controls so everyone who sells a font can appear high up on some list...

proprietary “popularity” calculation

interesting concept.

—————
Paul Ducco
Graphic Designer - Melbourne

Any idea how many/ how much sales it takes to get to #1?

~~~~~~~~~~~~~

...from the Fontry

Try for yourself. Or ask Jos :) Nothing wrong with not revealing ranking formulae, surely?

Nearly every site ranking anything has a proprietary popularity calculation. MyFonts doesn’t release that information for the same reason Google isn’t telling us how to get our websites to show up #1 in the results. I’m not sure why this surprises anyone.

While I certainly agree that under some circumstances popularity ranking systems should not be publicized in great detail as that would lead some to try and game the system to their own advantage (as I’m sure a few will try to do anyway), it’s not unfair to expect a very generalized explanation of the guiding principles of such a formula. We aren’t talking about democracy or the Academy Awards here, after all.

Summarizing what L.P. said, MyFonts’ formula is biased toward recent sales, and there may be other factors. Not very detailed, but at least it’s no longer mysterious. That maybe isn’t all that a lot of people would want to know about it, but it is sufficient to assure that they aren’t pulling numbers out of their butts while still not providing enough information that someone could artificially skew the results, at least not without significant personal expense.