|One||This is intel direct from people I know at iTunes: it's little to do with downloads, a lot to do with impressions (clicks on the show, posting reviews, liking or disliking the other reviews). Also: you'll zoom up the charts if there's a very sudden uptick in action on your show, ie, if you post a brand new show it goes from zero subscriptions to non-zero, an infinate increase, so a lot of new releases have quick but shortterm appearances high in the charts.|
|Two||Downloads per day|
Newness (when did the show launch)
(I also believe there's a kicker like "rate of growth" on some of the factors above)
|Three||Spikes in iTunes interaction.|
Interaction = new subscribes or streams. Perhaps ratings and reviews, although I'm skeptical. I actually think ratings and reviews are a red herring that are simply correlated with more iTunes interaction (because how could it not be?). But maybe they count.
Spikes ≠ automatic downloads to people who are subscribed. Thus, if you have 100k subscribers and on publish day only those 100k listen, you will not have had a spike. But if 100k subscribers get their episode, and another 100k decide to stream/new subscribe, you have spiked. Even 1-2k in new interaction will put you near the top of the charts. (I have seen this.)
This system would account for
a) new shows that we know have relatively small numbers being on the top of the charts
b) old shows that we know have big numbers often being off the charts. c) New press // iTunes promotion moving things up the charts.
It would also mean that shows that are consistently near the top (TAL, Serial, Radiolab, Wait Wait...) are consistently seeing INCREASES in interaction, not simply high subscriber rates. I cannot verify that. It also means we should stop equating high iTunes placement with total listenership.
It's not the Billboard charts.
But cool new idea: add those billboard chart arrows that show whether something has been moving up or down.
|Four||This might be super obvious, but it definitely seems like speed of growth is factored heavily.. and also maybe something to do with the number of downloads per number of episodes posted... That usually seems to be the reason that some weird podcast will shoot up to number one in their very early weeks and then drop off in a couple weeks. That also might help explain why This American Life has remained so dominant for so long as they only keep a few episodes up at any given time. Another thing to consider is that iTunes allows you to link directly to the iTunes directory listing of a podcast or episode... not sure how often people use that, but I'm curious how those links might fit into their weightings. Anyway, rooting for you and the success of Hot Pod 2.0 from DC! - Mickey|
|Five||Some measure of surge in listening is probably the most difficult to understand. Downloads this week over downloads in an average week, or something.|
|Six||I know some people believe that Apple has offline listening metrics that it doesn't share. Like, even if you download the mp3 and listen with no Internet connection, that file reports back to iTunes at next connection with info on whether it was played. I'd like to see this investigated.|
|Seven||Momentum (lots of downloads in short time periods), new-ness (particularly for major publishers like NPR, WNYC, Gimlet, Panoply, etc.), promotional consideration (often in association with newness)|
|Eight||I run a music podcast based around episode releases. We normally rocket into the top 15 and have reached as high as 5th place in the music charts and always fall out of the top 200 between episodes. Recently I watched the charts over a 2 hour period and watched us jump up and down spots from about 18th to 5th so it appears that it's very touchy around download spikes.|
To me this says recent downloads are a heavily influencing factor. We are relatively new so I wouldn't be surprised to see us stabilise over the next 6 months, but to be honest who knows?
A big problem with the iTunes charts though is that it only accounts for downloads via Apple media. 70% of our downloads come from outside Apple.
|Nine||how regularly you post new episodes, if there's a sudden percentage increase in the amount of listens, if your artwork is nice.|
|Ten||Here's what I wrote on this a little while ago!|
|Eleven||Number of downloads|
Number of subscribers
|Twelve||The number of reviews, subscriptions, plays and downloads|
|Thirteen||In the past it has been a combination of new subscribers|
last 24 hour + 48 hour + 72 hour
At one point the system could be gamed by subscribing and unsubscribing over and over again in the same day
|Fourteen||Revenue for Apple in some form or another|
|Fifteen||Apple as already said that new subscription initiations are a big factor that is weighed. Translation: promote your shows using the link to them on iTunes and they will appear higher on the charts. The Disney people did that - as did the religious right/ neo conservatives...|
|Sixteen||New & Noteworthy are shows that iTunes have personally selected to feature in that section.|
What's Hot is an automated random selection of shows that have a larger than normal change like an influx of reviews or subscriptions for example.
The Top 150 lists are based on the total number of NEW subscribers a show has had in the last week with a weighted average for the last 24, 48 and 72 hours.
|Seventeen||Depends on which chart you're looking at. New and Noteworthy is a curated list, but they look at popularity metrics (subscribers, plays, reviews, etc.). |
From a bit of insider knowledge, and having a lot of our podcasters move up and down in these charts this is what I can tell you. The lists themselves use to be entirely driven by those three metrics, with each one being much more heavily weighted depending on the chart; some seem to drive rankings in the categories, while other metrics seem to drive the top-100 more.
Seems like it was a relatively unsophisticated algorithm, so I wouldn't be surprised if Apple working on it, especially since they launched iTunes Connect. That could account for while we're seeing lots of fluctuations.
btw - there are a few things Apple changed with iTunes connect that are beyond just cosmetics: ability to redirect your iTunes listing, manually refreshing your feed, additional validation features, ability to turn podcasts on and off, and an easy way to delete podcasts. Surprisingly, this wan't available before. I did a quick write up on it a few days ago on Medium - https://medium.com/create-buzz/submitting-your-podcast-to-itunes-using-itunes-connect-22b95b38068d#.cytrgujap
Sorry for the late reply. It's gotta be heavily weighted towards momentum --> namely, if you got more downloads this hour than last hour, how many you've gotten over the past 24 compared to yesterday.
I've seen articles about it factoring in the amount & score of ratings and reviews (duh). But I'm leaning towards/hoping that momentum is a key factor