6 ms·

Download bots are a well-known secret of the iOS eco-system

56 points by akkartik 15y ago

alexlin 15y ago
I personally don't understand why "bot farms" are that "scummy" or unfair. It was possibly the best way for the average indie developer to stand out without a large wallet (to buy banners). This rash of a problem is even more aggravated ever since incent PPI was banned. Because of this, the era of rapidly acquiring users at scale on iOS is coming to an end. In it's place, developers will be forced to revert back to traditional ad buys in the form of banner ad networks with ridiculously "black 'box" targeting options. I don't know if this is a bad thing since developers will spend more time figuring out their LTVs through each respective acquisition channel, but you sure as hell better be able to throw a few thousand dollars here and there and expect it to never back out. All in all, Apple's app store policy should just read, "If you've solved discoverability on our app store, we're banning your method."
- jws 15y ago
  I personally don't understand why "bot farms" are that "scummy" or unfair. They fraudulently make the claim "lots of people like this". User's believe a "most downloaded" means other people like them have made a decision and picked this little grain of wheat from the chaff. Bot farming is fundamentally lying to your potential customers to the detriment of applications that should be on the list.
  - alexlin 15y ago
    And marketing as a concept isn't? Why aren't people rallying against Tapjoy when they're incentivizing people to download apps in exchange for virtual currency? As long as you pay $20-30k, you have enough downloads to breach the top 25. Or take banner ads. Zynga has the ability to throw $100k IOs at random ad networks and their new apps are instantly in the Top 25 shortly after launch. Does this reflect popularity/quality or is it just a competition between who has the biggest wallet? If Apple changed it's app store algorithm to function like the Android app store (that takes uninstall rates/ratings into account), this wouldn't be a problem.
- sokoloff 15y ago
  Really? You believe that, just because it was cheap for indies, that ranking an app more highly in the app store based on non-human downloads was beneficial to the human users of the app store? I'm no Apple fanboy generally, but this seems like they're making a choice based on what's good for their human users, and that choice benefits some the developers at the expense of others. (IMO, it benefits the ones who are competing "fairly", but that doesn't even matter compared to "is it good for users?")
  - alexlin 15y ago
    You're twisting my words. Are you really so naive to believe that if you make a great app that you're automatically going to climb to the Top 25? Try navigating the app store. Tell me how you discover apps. A great app that was released is never going to get anywhere unless a marketing budget has been allocated beforehand (unless you somehow win the lottery and create the next angry birds.) And lets talk about the "expense" of others. Did you feel like you lost anything when bots and incent PPI were allowed? Did you even know about it? The app store functions in a way that it's sort of like the survival of the fittest. If the people don't like an app, it's going to drop back down as fast as it climbed up. If Apple wants to create an ecosystem where indie developers can thrive, they need to change their ranking algorithm, solve discoverability for their developers, or stop banning the solutions people have been coming up with (incent PPI, bots.)
- ericflo 15y ago
  A Kantian ethicist might say that if you apply the first formulation of his categorical imperative you would find that it's immoral because if everyone were to do it, the whole system would stop working.
  - alexlin 15y ago
    That's not necessarily true. Bad apps will naturally drop. App ratings don't lie (at least to some extent.) You also have to account for the LTV of users. If it's not backing out (which is most likely for a bad app), they're just wasting their money.
- feralchimp 15y ago
  > It was possibly the best way for the average indie developer to stand out without a large wallet (to buy banners). How indie is indie? $5k - $15k may not be considered a "large wallet" by internet advertising standards, but it's a whole lot of money by "I just wrote a great app in my spare time and want people to see it" standards.
  - alexlin 15y ago
    It varies. But $5k means their app can take a trip to the top 25 and gain around 8k organic users/day. You know where $5k in shotty banner ads gets you? ~2k organics tops. Prove me wrong that $5k is a lot easier to scrape together than say... iAd's $25k minimum ad buy.
    - feralchimp 15y ago
      Very interesting to hear those numbers. Is that personal experience, anecdotal, or public data from someplace? Still, an explicit "sponsored app links" section would be much better than inflation through bot-pumping.
bprater 15y ago
Part of this is Apple's fault. In general, Apple needs to look at some alternative ways to get apps in front of iOS user's eyes. AppStore has been mostly untouched over its lifetime. I frequently scan the "Featured" category and "Top 25", but I rarely go deeper. The curation could be much broader and touch individual interests, but right now, it doesn't. I would ask this to iOS developers: what would you like to see AppStore changed so that you might get additional exposure for your software?
- bobbles 15y ago
  As a pure iOS user (ie. not an iOS dev). I would like to see something that lets me hide an app from the app store entirely (and maybe provide a list of the apps I have hidden?) I end up just wading through lists and lists of apps I already know I don't want and arent interested in. Abilities to hide apps based on rules would be great too. For example, some reason chosen by me - so that when I go to download this app in the future, i might see something like "deleted this app because they added bad micropayments that ruined the game" Something like 'categorise by publisher' could be cool too, if a company is making consistantly good apps I would like to find more of them.
  - pagekalisedown 15y ago
    Being able to hide all apps from certain developers (crap peddlers) would be immensely useful too.
- vannevar 15y ago
  Apple has little interest in more equitable distribution of sales. Particularly in the early days of the App Store when they needed to attract developers, it made better headlines to have one dev making $10M than to have 100 making $100K. It was the illusion of a gold rush that helped grow the App Store so quickly. At this point attracting developers isn't much of a concern, but there's still little incentive to make app discovery easier. Apple only needs enough quality apps to populate their lists. The other 99.9% serve to pad the number of iOS apps for marketing purposes so that Apple can club Android over the head. For users it doesn't matter much that there are 500,000 iOS apps because there remains no really good way to find most of them. But that big number sure sounds good in commercials.
orbitingpluto 15y ago
The article states that the Android Market is much fairer. But from my experience bots are very much alive on the Android market. A competing app was released within one week of mine late last October. I managed 1000 downloads in a week, he managed 10,000 in far less than that. His Google +1s also went to 381 almost immediately. His current # of Google +1s more than two months later: 389.
- EtienneJohnred 15y ago
  This indicates absolutely nothing. Maybe the competing app is better than yours. Maybe the developer is better at marketing than you are. While I don't doubt that similar crap is going on in the Android market (likely on a much lesser scale since the stakes are so much lower), you can't simply chalk this particular case to "bots".
- alexlin 15y ago
  Bots don't exist on Android, but incent PPI does. It doesn't matter about the # of installs he gets because Android takes into account uninstall rates as well as your app ratings. If he's ranked higher than you, he's most likely hitting the factors better than you.
  - orbitingpluto 15y ago
    Why are you sure that bots do not exist on Android? The article did not deny their existence. As for that "more holistic" approach, Bots don't usually uninstall apps that they've downloaded. They just get re-instantiated (and possibly reused to further help pad results). Also if PPI was used, a lot of cash was blown to get 10k downloads and 381 Google +1s sandwiched into a manner of hours. That means the developer either has an inside track or lots of cash to blow and no thought to ever seeing a return.
- drivebyacct2 15y ago
  Which one? https://market.android.com/developer?pub=orbitingPluto https://market.android.com/developer?pub=orbitingPluto Sorry, don't know why this is upsetting to ask...
mikeash 15y ago
Is there any evidence for the bots theory besides the fact that some services provide a large increase in downloads without a corresponding large increase in app use? The article is pretty long, but that's the only piece of actual evidence I could find in it, as opposed to a whole bunch of hearsay and speculation. Seems to me that bumping downloads but not usage could just as easy happen with pushy ads targeting users with a healthy "ooh shiny" instinct, especially for free (or freemium) apps, so this really does not seem convincing to me.
- kimmai 15y ago
  It's not really hearsay. The top CEO quoted runs a company, Fiksu, that does user acquisition on behalf of Groupon, Gilt, Vh1, etc. Most large developers know about bots because they've used or tried them. They will say this off the record to reporters, but never on the record because they don't want to be banned by Apple. Also, Apple issued a warning to developers about using these services last week. Also: @Alexlin, who everyone is beating on in the comments, ran a user acquisition, marketing company for apps called G6Pay and runs his own mobile gaming company, which is part of YC. He actually knows his stuff about how to get to the top of the charts.
  - mikeash 15y ago
    The Fiksu claim still seems to be based entirely on downloads which don't correspond to use. They think it's bots, but they don't actually know, all they know is they got a lot of downloads from something that didn't use the app afterwards. For Alexlin, well, I'd be interested in knowing how he knows this too.

56 points by akkartik 15y ago

I personally don't understand why "bot farms" are that "scummy" or unfair. It was possibly the best way for the average indie developer to stand out without a large wallet (to buy banners). This rash of a problem is even more aggravated ever since incent PPI was banned. Because of this, the era of rapidly acquiring users at scale on iOS is coming to an end. In it's place, developers will be forced to revert back to traditional ad buys in the form of banner ad networks with ridiculously "black 'box" targeting options. I don't know if this is a bad thing since developers will spend more time figuring out their LTVs through each respective acquisition channel, but you sure as hell better be able to throw a few thousand dollars here and there and expect it to never back out. All in all, Apple's app store policy should just read, "If you've solved discoverability on our app store, we're banning your method."

jws 15y ago
I personally don't understand why "bot farms" are that "scummy" or unfair. They fraudulently make the claim "lots of people like this". User's believe a "most downloaded" means other people like them have made a decision and picked this little grain of wheat from the chaff. Bot farming is fundamentally lying to your potential customers to the detriment of applications that should be on the list.
- alexlin 15y ago
  And marketing as a concept isn't? Why aren't people rallying against Tapjoy when they're incentivizing people to download apps in exchange for virtual currency? As long as you pay $20-30k, you have enough downloads to breach the top 25. Or take banner ads. Zynga has the ability to throw $100k IOs at random ad networks and their new apps are instantly in the Top 25 shortly after launch. Does this reflect popularity/quality or is it just a competition between who has the biggest wallet? If Apple changed it's app store algorithm to function like the Android app store (that takes uninstall rates/ratings into account), this wouldn't be a problem.
sokoloff 15y ago
Really? You believe that, just because it was cheap for indies, that ranking an app more highly in the app store based on non-human downloads was beneficial to the human users of the app store? I'm no Apple fanboy generally, but this seems like they're making a choice based on what's good for their human users, and that choice benefits some the developers at the expense of others. (IMO, it benefits the ones who are competing "fairly", but that doesn't even matter compared to "is it good for users?")
- alexlin 15y ago
  You're twisting my words. Are you really so naive to believe that if you make a great app that you're automatically going to climb to the Top 25? Try navigating the app store. Tell me how you discover apps. A great app that was released is never going to get anywhere unless a marketing budget has been allocated beforehand (unless you somehow win the lottery and create the next angry birds.) And lets talk about the "expense" of others. Did you feel like you lost anything when bots and incent PPI were allowed? Did you even know about it? The app store functions in a way that it's sort of like the survival of the fittest. If the people don't like an app, it's going to drop back down as fast as it climbed up. If Apple wants to create an ecosystem where indie developers can thrive, they need to change their ranking algorithm, solve discoverability for their developers, or stop banning the solutions people have been coming up with (incent PPI, bots.)
ericflo 15y ago
A Kantian ethicist might say that if you apply the first formulation of his categorical imperative you would find that it's immoral because if everyone were to do it, the whole system would stop working.
- alexlin 15y ago
  That's not necessarily true. Bad apps will naturally drop. App ratings don't lie (at least to some extent.) You also have to account for the LTV of users. If it's not backing out (which is most likely for a bad app), they're just wasting their money.
feralchimp 15y ago
> It was possibly the best way for the average indie developer to stand out without a large wallet (to buy banners). How indie is indie? $5k - $15k may not be considered a "large wallet" by internet advertising standards, but it's a whole lot of money by "I just wrote a great app in my spare time and want people to see it" standards.
- alexlin 15y ago
  It varies. But $5k means their app can take a trip to the top 25 and gain around 8k organic users/day. You know where $5k in shotty banner ads gets you? ~2k organics tops. Prove me wrong that $5k is a lot easier to scrape together than say... iAd's $25k minimum ad buy.
  - feralchimp 15y ago
    Very interesting to hear those numbers. Is that personal experience, anecdotal, or public data from someplace? Still, an explicit "sponsored app links" section would be much better than inflation through bot-pumping.

bprater 15y ago

Part of this is Apple's fault. In general, Apple needs to look at some alternative ways to get apps in front of iOS user's eyes. AppStore has been mostly untouched over its lifetime. I frequently scan the "Featured" category and "Top 25", but I rarely go deeper. The curation could be much broader and touch individual interests, but right now, it doesn't. I would ask this to iOS developers: what would you like to see AppStore changed so that you might get additional exposure for your software?

bobbles 15y ago
As a pure iOS user (ie. not an iOS dev). I would like to see something that lets me hide an app from the app store entirely (and maybe provide a list of the apps I have hidden?) I end up just wading through lists and lists of apps I already know I don't want and arent interested in. Abilities to hide apps based on rules would be great too. For example, some reason chosen by me - so that when I go to download this app in the future, i might see something like "deleted this app because they added bad micropayments that ruined the game" Something like 'categorise by publisher' could be cool too, if a company is making consistantly good apps I would like to find more of them.
- pagekalisedown 15y ago
  Being able to hide all apps from certain developers (crap peddlers) would be immensely useful too.
vannevar 15y ago
Apple has little interest in more equitable distribution of sales. Particularly in the early days of the App Store when they needed to attract developers, it made better headlines to have one dev making $10M than to have 100 making $100K. It was the illusion of a gold rush that helped grow the App Store so quickly. At this point attracting developers isn't much of a concern, but there's still little incentive to make app discovery easier. Apple only needs enough quality apps to populate their lists. The other 99.9% serve to pad the number of iOS apps for marketing purposes so that Apple can club Android over the head. For users it doesn't matter much that there are 500,000 iOS apps because there remains no really good way to find most of them. But that big number sure sounds good in commercials.

orbitingpluto 15y ago

The article states that the Android Market is much fairer. But from my experience bots are very much alive on the Android market. A competing app was released within one week of mine late last October. I managed 1000 downloads in a week, he managed 10,000 in far less than that. His Google +1s also went to 381 almost immediately. His current # of Google +1s more than two months later: 389.

EtienneJohnred 15y ago
This indicates absolutely nothing. Maybe the competing app is better than yours. Maybe the developer is better at marketing than you are. While I don't doubt that similar crap is going on in the Android market (likely on a much lesser scale since the stakes are so much lower), you can't simply chalk this particular case to "bots".
alexlin 15y ago
Bots don't exist on Android, but incent PPI does. It doesn't matter about the # of installs he gets because Android takes into account uninstall rates as well as your app ratings. If he's ranked higher than you, he's most likely hitting the factors better than you.
- orbitingpluto 15y ago
  Why are you sure that bots do not exist on Android? The article did not deny their existence. As for that "more holistic" approach, Bots don't usually uninstall apps that they've downloaded. They just get re-instantiated (and possibly reused to further help pad results). Also if PPI was used, a lot of cash was blown to get 10k downloads and 381 Google +1s sandwiched into a manner of hours. That means the developer either has an inside track or lots of cash to blow and no thought to ever seeing a return.
drivebyacct2 15y ago
Which one? https://market.android.com/developer?pub=orbitingPluto https://market.android.com/developer?pub=orbitingPluto Sorry, don't know why this is upsetting to ask...

mikeash 15y ago

Is there any evidence for the bots theory besides the fact that some services provide a large increase in downloads without a corresponding large increase in app use? The article is pretty long, but that's the only piece of actual evidence I could find in it, as opposed to a whole bunch of hearsay and speculation. Seems to me that bumping downloads but not usage could just as easy happen with pushy ads targeting users with a healthy "ooh shiny" instinct, especially for free (or freemium) apps, so this really does not seem convincing to me.

kimmai 15y ago
It's not really hearsay. The top CEO quoted runs a company, Fiksu, that does user acquisition on behalf of Groupon, Gilt, Vh1, etc. Most large developers know about bots because they've used or tried them. They will say this off the record to reporters, but never on the record because they don't want to be banned by Apple. Also, Apple issued a warning to developers about using these services last week. Also: @Alexlin, who everyone is beating on in the comments, ran a user acquisition, marketing company for apps called G6Pay and runs his own mobile gaming company, which is part of YC. He actually knows his stuff about how to get to the top of the charts.
- mikeash 15y ago
  The Fiksu claim still seems to be based entirely on downloads which don't correspond to use. They think it's bots, but they don't actually know, all they know is they got a lot of downloads from something that didn't use the app afterwards. For Alexlin, well, I'd be interested in knowing how he knows this too.