How Paypal and Reddit faked their way to traction
When “faking it before making it” actually works
There were several factors that contributed to YouTube becoming the #1 video sharing service on the web. But a lot of initial adoption was driven by the fact that it had pirated content hosted on it. If you wanted to watch the latest episode of Lost for free, YouTube was your best bet: no queued downloading through torrents, just stream it from the server. Kim Dotcom noted how pirated content was driving YouTube’s adoption and figured that seeding some of that could unlock traffic for MegaUpload. It definitely worked. But playing with pirated content, somebody was bound to get burnt. MegaUpload went under when it was alleged that the pirated content was, unlike the YouTube case, a deliberate part of the platform’s strategy.
Kim Dotcom’s mail to his crew, which ultimately damned MegaUpload, offers a rather brute-force but effective tactic for seeding platforms. Platforms and two-sided markets are difficult to kick off, in general. The problem gets a lot more complicated when both consumers and producers need to be on the platform simultaneously. For a service like Yelp, the consumer side can gather traction independent of merchants coming in. But for marketplaces like eBay and communities like Quora, the consuming side (buyers, question askers & readers) and producing side (sellers, question answerers) need to be on the platform from Day 1 for interactions to spark. (These may not be distinct groups of people.) This is because there is no inherent value in the standalone product for a user (eBay, Quora) without the participation of users from the other side who create complementary products (goods for sale, answers). When users come to the platform, they need to get an indication of activity.
Some community owners, as in the case of MegaUpload, solve this by creating a false aura of activity using a variety of methods. (Who would have tried MegaUpload were it a veritable video ghost town?) First-time users get the impression that the platform is already in business and stay on. Over time, the user base grows and the platform sustains activity without having to fake it.
There are typically three ways of going about this:
Seeding and Weeding
Dating services simulate initial traction by creating fake profiles and conversations. Users who come to the site see some activity and are incentivized to stay on. Marketplace sites may also show fake activity to attract buyers and sellers.
In the book, Paypal Wars, Eric M. Jackson talks about how PayPal grew a base of sellers who accepted PayPal by creating demand for the service among buyers. When Paypal figured that eBay was their key distribution platform, they came up with an ingenious plan to simulate demand. They created a bot that bought goods on eBay and then, insisted on paying for it using PayPal. Not only did sellers come to know about the service, they rushed onto it as it already seemed to be getting popular. The fact that it was way better than every other payment mechanism on eBay only helped repeated usage.
A platform is useless without complementary products. Marketplaces, especially, are dead without sellers posting on them. To solve the chicken-egg problem, some marketplaces create fake supply to attract buyers. Services marketplaces put up fake projects to show activity. Steve Sammartino talks of how he seeded Rentoid.com by essentially buying the initial items himself and renting them out (though he refers to it as “Inventing Demand”, when actually he was seeding supply).
There is, of course, a method to the apparent madness.
Use attractive seeds
When you’re faking it, you might as well offer the best. MegaUpload knew that Collective Soul’s music videos would drive more traffic than some home video of a puppy snoring in a bathtub. Dating services know what men want (who doesn’t) and seed the network with photos of Latin American models with eclectic interests (not that the latter matters).
Encourage behavior you want to see on the site
Reddit co-founder Steve Huffman admitted that the link-sharing site was initially seeded with fake profiles posting links to simulate activity. The key was that the links being posted were the kind of content the founders wanted to see on the site over time. Reddit not only gathered steam, the initial content also attracted people who were interested in similar content and created a culture of good content in the community.
If you’re inventing supply or demand and don’t want to do it all in-house, incentivize an initial set of users to spark greater activity than they otherwise would. Amazon Mechanical Turk may not be a bad place to start at if what you’re looking for are users who can churn out some activity for cheap.YouTube’s contests are another example of getting initial usage sparked with artificial incentives.
Scalability rocks! I mean it’s the whole point this industry is what it is. But during the early pre-viral, pre-engagement phase of a community, non-scalable methods may help kick start things. Quora admins answered a lot of questions themselves at first. Brian Chesky talks about how they got the first users on AirBnB themselves literally going door-to-door.
Sometimes, it helps to learn manually what works and then automate it, instead of automating and testing a hundred times.
More importantly, seeding initial activity helps you to experience the product as a user. Invaluable education that!