Planet Quark

When Quark released the technology study of the new DesignPad app, we were absolutely flabbergasted when we saw the App Store Ratings the next day:

#1 in its category in France and Germany, in France even #2 for all free apps.
So before the popular apps of the large social networks, shopping & auction apps and news apps.
And similar results in many other countries for that day.

Of course I (naively?) thought the download rank is just ranking the daily absolute downloads per app.
Sure, for search ranking Apple (like all search engine providers) must use an ingenious formula that Apple – like Google – changes over time:
http://techcrunch.com/2012/06/29/looks-like-apple-has-changed-its-app-store-algorithm-again/

But for download ranks? (which has an obvious effect on the success of your app)

When after the first hype the ranking slowly dropped again and by looking on the absolute numbers I saw some discrepancies of this theory:

(App Store Ranking France for DesignPad)

So I began to wonder how Apple might calculate this, maybe taking absolute numbers or a weighted history into consideration?

Of course and totally understandable there is no public information available for their algorithm, as otherwise it would be open to manipulation. However when searching the web, I came across an interesting study that claims that they have cracked the code for ranking (not Search Optimization). I can’t be sure whether this is the right code and personally I have no insight into Apple’s algorithm, however I thought it is interesting to share this theory:

(Theory of weighted algorithm for Apple’s App Store download ranking)

So the theory is that the downloads of the day is the most influential factors, however the downloads of the past three days are also factored in. You can find the original slide-deck here: http://www.slideshare.net/misteroo/how-to-market-your-app (slide 25)

Popularity: 7% [?]