The Future of App Stores (and the history of video)

I grew up watching movies.  Mostly on video tapes, which came in hard plastic cases from a place called Flicks & Pics a few blocks from my house.  Flicks & Pics was something of an institution in my neighborhood; it was a mom-and-pop rental joint with about twenty thousand VHS tapes in stock, organized by genres like “Action,” “Noir,” “Cult,” “Classics,” and even “Stinkers.”  On Tuesdays you could rent any movie in the store for a dollar, and the owners refused to charge for movies that starred Brooke Shields.  I watched Akira, The Maltese Falcon, Strangers on a Train, and Blood Simple on Flicks & Pics tapes.  A Taxing Woman, Beverly Hills Cop, Drunken Master 2, City Lights–the list goes on.  You selected a video by finding its box on the shelf and taking the tag hanging under it to the register; if no tags were present, all copies were already rented.  This system meant that even minor films for which only one copy existed still got the same amount of shelf space as the latest summer blockbuster. Some films even bore small stickers with comments from employees, inviting you to take a chance on a feature you might have never otherwise considered.  It was always easy to find something new at Flicks & Pics.

Then, sometime in the early 1990s, Blockbuster arrived in town.  I visited it shortly after it opened and was impressed by its large size and volume of titles.  But upon closer inspection I noticed something strange: Blockbuster always had twenty or thirty copies of the latest hit, but every other film in the store seemed to be some Grade Z schlock that I’d never heard of before.  The oldest films in the store were made around 1982; not only did they not carry High and Low, they didn’t even have a foreign section!  Instead they focused on a narrow selection of big budget flicks and padded the rest of the store out with soft porn and “films” like Ninja III: The Domination. “Who rents this crap?” I wondered.

I later learned that Blockbuster’s business model operated on deals with the movie companies.  In exchange for discounts for the big-budget, heavily marketed films, Blockbuster agreed to supplement its stock with direct-to-video movies that had been produced with almost no budget.  The theory, I guess, is that the big titles would draw people into the store and eventually they’d rent some of the schlock.  Since the schlock was so cheap to produce, Blockbuster’s guaranteed purchase of a certain amount of it was a good deal for the movie companies.  And for Blockbuster it allowed them to fill their catalog at low cost while ensuring that they always had films less than a year old in stock (this is called, apparently, “depth of copy”).

Blockbuster (and its twin, Hollywood Video) rolled into towns across America and pretty much put an end to local mom-and-pop video stores.  Several in my town gave up the ghost after the big B arrived.  But Flicks & Pics survived.  Small as it was, it catered to a crowd that was looking for films they couldn’t find at Blockbuster.  What gave that neighborhood video store staying power was the way it was curated; the selection of films, complete with staff recommendations, made it a goldmine for people who were ready for something a little more challenging than Red Shoe Diaries.  Flicks & Pics had “breadth of copy;” while not every film was good, there was always something new to find and the chances of renting a real train wreck were pretty low. When the store finally went out of business in 2007 (long after the local Blockbuster had become an empty, derelict building), it was because the VHS format was no longer viable.  Advancements in technology dealt the blow that the big chain stores could not.  When the store closed there was a sort of community vigil in the parking lot.

The lesson here is applicable, I think, to modern app stores.  We have a problem very similar to that faced by video stores in the 1990s: the volume of apps is now so immense that an unfiltered catalog is impenetrable.  iTunes and Google Play both do their best to surface the content of their catalog to users, but they don’t have the virtual shelf space to promote every high-quality app.  They probably don’t even have a good way to identify high-quality apps (especially since the Apple review board apparently spends in inordinate amount of time trying to ensure that no screenshots containing male genitals are approved).  Users have become wary of spending money on apps they haven’t tried because the crap apps are right there in the database along with the good ones.

The volume of apps has made visibility the #1 problem for developers to solve.  The rise of free-to-play isn’t just about new monetization schemes–it’s a way to lower the user’s perceived risk and thus get more people to give the game a chance.  All the social media integration in games lately serves as another way to get noticed outside of the stores themselves.  These sorts of alternate approaches are increasingly necessary because even with promotion by Apple and Google, high quality paid games aren’t being purchased anymore.  You might have built the Citizen Kane of mobile apps, but it’s unlikely to gain traction while wedged between You Got Served and Battlefield Earth on the shelf.  We need a filter.  A way to more effectively separate the wheat from the chaff.

Blockbuster’s filter was pretty simple: let the market(ing) decide.  The chain would observe which films did well at the box office and then ensure that they had enough to satisfy the high customer demand for those films immediately upon the video release.  Depth of copy: they focused on having a large number of a few recent hits because that’s what customers generally wanted.  And of course, customer demand was driven by marketing campaigns paid for by the production companies.  Money goes into marketing, marketing drives demand, demand drives investment by places like Blockbuster.  And for everything else, Grade Z apparently sufficed.

The App Store today is admittedly kinder than Blockbuster’s model.  Apple and Google make valiant attempts to promote content that they think is high quality, and this helps offset the necessity for heavy marketing.  It’s still possible for an app or game without mega marketing dollars behind it to become a hit.  But as the volume of apps increases, and as users become increasingly risk averse, mega marketing is becoming more important.  Promotion by Apple and Google is unreliable; developers with the cash are investing heavily in marketing to avoid being drowned in the stream of new releases.  As in every other medium, marketing works.  Marketing, or including an already well-marketed brand, seems to be the only way to make a paid game viable on the App Store today.

But the problem with the marketing-driven model is that it favors the developers who have giant money bags to throw at marketing.  If a huge marketing spend is required for success, only companies that are already rich will be successful.  The app stores have been a democratizing force, especially in the game industry, because suddenly small developers and independents have found themselves on equal footing with the big guys.  If Google Play and the App Store stay the way they are now, I think that the large companies will smother everybody else.

But what if we had something a little bit more like the model employed by Flicks & Pics? A breadth of copy model, where the entire catalog–not just the boxes in the window–is selected with a certain audience in mind.  Valve’s Steam seems to prove that this model can work and be highly profitable, but there is a big difference between what Valve does with its store and what Google and Apple do with theirs.  The audience for games on Steam is extremely well defined: they are PC gamers who like Valve games.  From that you can extrapolate which types of products are likely to do well and which are likely to fail and thus apply a content filter fairly easily.  Apple and Google, on the other hand, must address everybody who has a phone.  Which is to say pretty much everybody in the world.  That’s a lot harder filter to design.

Apple and Google do promote apps that they select based on some secret criteria.  But hand-picking content isn’t going to cut it in the long run because the process just doesn’t scale.  Back when people browsed the web with Netscape, Yahoo was little more than a directory of web pages.  A phone book of links, organized by hand into categories and sub-categories that could be easily clicked through.  Yahoo’s viability fell through the floor when Google came out because its hand-built index only represented a small portion of the web.  Google, on the other hand, could search for anything.  You know what happened to those two companies after that.

We need a an app store that operates like Flicks & Pics did, a breadth-of-copy approach that is designed to introduce new content to specific types of users.  But gone are the days when this is a task that can be performed by hand.  We need a system that suggests different apps to different users, one that considers quality metrics other than raw downloads.  A way to ensure that users can find apps that marketing has got them excited about, and a way to expose those same users to apps that have no marketing.  A system that exposes unknown gems to enough people that they have a chance to become a hit.  If Steam is a store for PC-gamers-who-like-Valve-games, we also need a store for young-adults-who-like-mysteries and middle-aged-parents-who-enjoy-travel and children-who-like-outer-space, etc, etc, etc.  Simple categories don’t cut it.  Top 10 lists don’t cut it.  There’s so much content on the app stores now that we need smarter systems to browse it.

Perhaps Apple’s purchase of Chomp will eventually cause the App Store to evolve into something smarter, something more dynamic, something less controlled by marketing alone.  I sure hope so.