I'm a creator, product manager and programmer based in London


Writings and musings about history, technology, development and anything else that seems interesting.

Building Foundbite: Seeding a Social App

This is the first in a series of posts I am writing about Foundbite and the lessons I learnt building a sound sharing social network - you can read more about the product here.

Apps that revolve around user generated content and community engagement have a "chicken and egg” problem when they are starting out. Without content they won't keep an audience's attention and without an active community there won't be any new content uploaded. Lessons learnt from other services that cracked this helped us a lot and led us to one of the best product decisions we made.

Social apps and two-sided marketplaces live and die by their ability to attract and retain users. You have moments to convince your customers of the worth of what you’ve built before they leave and, most likely, never give it a second thought. The problem for many products is that without users, and the content they contribute, they aren’t really that interesting or useful so, in turn no one will contribute any content and users won’t stick around. This situation is in keeping with Metcalfe's Law which states that the value of a network is proportional to the number of participants squared- when your network is very small it's not that valuable but becomes increasingly valuable with every new user.

The ability to keep your first users when the value of the network is most likely low, and your visions for it largely unrealised, is important. It’s a challenge but one that many companies have tackled before. As I see it, there are some key ways their products have managed to increase the perceived value of their product in the early stages, convince new users of its potential and to overcome the "chicken and egg problem"…

Narrow Your Audience

Launching a marketplace or community that appeals to a broad audience from day 1 is near impossible. If you try to please everyone with your first version you won't please anyone. It is far more effective to identify a single group who you think your app will resonate with and shape your launch strategy and product to them. It might not matter that you don't have much content if it is highly relevant and your users are able to easily connect with like-minded people, thereby increasing the value of the network to that user. Not only will this allow you to focus on your core customers but you'll also receive strong signals when something is or isn't working for them, information that you might not pick up on with a wider audience. Choosing a demographic you are familiar will also give you a big advantage.

Facebook initially launched to a tiny portion of what would be their total addressable market

Facebook initially launched to a tiny portion of what would be their total addressable market

Facebook's launch strategy is probably one of the best examples of this. They started catering to their fellow students at Havard and gradually expanded university by university during which you couldn't sign up unless you had a .edu email address.

Other notable examples include YikYak, a now defunct anonymous message board that also started on College campuses and AirBnB who launched in August 2008 to cater for the Democratic National Convention in Denver which had a hotel shortage and through which they got their first bookings.

Piggyback Off Others

These days if you're building a digital product your customers are going to be somewhere else online. Through third-party APIs it's possible to utilise other services to import content, market your product and allow people to find other friends who are also using your app, making things immediately less lonely. There's also the option of working with a prominent individual or company and using their audience to promote your service, this can provide a big initial influx of users, enough of which hopefully will stick around.

Meerkat used Twitter’s community and social graph to propel themselves up the App Store charts

Meerkat used Twitter’s community and social graph to propel themselves up the App Store charts

The selfie app Shots has since pivoted but got a huge amount of traction by teaming up with one of their investors, Justin Bieber, to promote the app to his many millions of fans on Instagram and Twitter. At the peak of Bieber-mania, I suspect many of their first users came to see more of what Justin was up to without initially worrying about the lack of content from other users.

Meerkat, now Houseparty, is a live streaming video platform that piggy-backed off Twitter's service. Initially you could sign in, find friends using the app and share your live stream via a link, all using Twitter. It was a strategy that worked brilliantly until Twitter decided to cut them off - always a risk when you're relying on any third-party service.

Provide a Utility

Providing your early users with features that help them achieve something easier/faster/better is a proven way to keep them coming back whilst your product is in its infancy. As usage grows you can then begin to introduce social features to the product and start fostering a community which should eventually prove a bigger draw than the utility and flip the product on its head. For many companies this initial tool stage, or "single player mode", can be incredibly short but if you’re not lucky enough to go viral, creating a valuable utility could make all the difference.

The ‘come for the tool, stay for the network’ strategy isn’t the only way to build a network. Some networks never had single-player tools, including gigantic successes like Facebook and Twitter. But starting a network from scratch is very hard. Think of single-player tools as kindling.
— Chris Dixon

Examples of companies that have leveraged this are plentiful: Pinterest started out as a simple way to bookmark images and links to create mood boards, Skype began by facilitating a video call between two people, Strava provided accurate run tracking features and AirBnb provided attractive property pages which users could then share to Craigslist where the standard of property pages was lacking (see also: Piggybacking Off Others!).

Fake It

Opening an app for the first time and seeing lots of posts uploaded by the same 5 people on your team can be massively discouraging. There are a number of ways you can give the perception to new users that they aren't alone or interrupting a private club and are instead joining a thriving community that they'll be more likely to contribute to.

It goes without saying that using this technique won’t work for all products and comes with risks. There is a very thin line between generating content to show off the potential of your service and outright deception. For me, the widespread practice of using fake profiles on new (and some established) dating sites that automatically like people to give the impression there are far more beautiful and available people using the app than there really are is squarely in the unethical region of "growth hacking".

Most famously Reddit's founders posted the kinds of links they wanted to see shared on the site using lots of fake accounts whose usernames were created using random name generators. They didn't use bots and there were no comments back then so no one missed out on any insightful discussions with "Joe Bloggs".

If you do use this technique make sure you account for it in your metrics. Don't go buying your own products to inflate your sales without telling your current and potential investors.

Focus on Quality

Rather than relying on sheer quantity of content to impress, you could choose to focus your efforts on including less but higher quality content. Although time consuming to produce it has the advantage of showing off the full potential and capabilities of your service as well as giving users some ideas about what they could use the service for. It’s worth carefully considering what you choose to post as it sets the tone for your early adopters and how they behave. Whatever you post you’ll probably end up with a lot more of it and the kinds of people it resonates with.

Quora, the Question and Answer site, started out with all it's staff answering questions helping their users get good answers to their first question and, as mentioned above, Reddit seeded the site with the kind of links the founders were interested in seeing. AirBnB (again!) went around taking photos of as many apartments as they could in New York to improve the quality of listings and leading to a 2-3 times the number of rentals. Rather than filling Tinder with fake profiles, Whitney Wolfe, the then CMO and now Founder of Bumble, tackled this problem by visiting sororities (Narrowing Your Audience) where she would get them all to sign up before going to a neighbouring fraternity and getting them all doing them the same which ensured there was a smorgasbord of young, single and attractive people using the service from the off.

Do Things That Don't Scale

By now you may well have read Paul Graham's influential essay "Do Things That Don't Scale". It's a fantastic discussion on the scrappy, manual tasks and techniques founders use to delight their early customers, until the product takes off at which point doing these kinds of things become unfeasible. Katerina Fake, one of the co-founders of Flickr, likens this stage to being a good host at a party where you take the time to greet everyone, take their coat and try to ensure they have a good time and don’t leave after their first beer.

As you can probably see from the anecdotes I've mentioned so far, from AirBnB taking photos of their customers apartments and Whitney Wolfe going to college frat and sorority houses, most of them involve putting in a lot of legwork. Steve Sammartino of online rental marketplace Rentoid even went as far as to go out and buy all the products his first users wanted to hire so he could rent them to them, a time consuming and initially expensive undertaking.

One big advantage of greeting a user in the comment section or curating and posting interesting articles is that there is the opportunity to use the product you've created as your users would. This dogfooding can be an invaluable experience, giving you the opportunity to notice bugs early and get a feel for the product you’ve created.

How we solved (or tried to solve) this problem at Foundbite

Launching on Windows Phone, Microsoft's late entry to the mobile operating system race, was both a great decision for us and a terrible one. On one hand they helped us to get the product off the ground with funding and support but it was a bad place to build a social app. As we've seen from Metcalfe's Law in order to grow the value of a network you need to be able to connect your new nodes to other nodes in the network. For social networks on iOS or Android that often just involves using a "find my friends" feature but whilst tied to Windows Phone (Foundbite was exclusive for 1 year) most of the users coming to the app didn't have anyone else they knew using the app or, in many cases, Windows Phone itself. For social networks reliant on these valuable close connections and viewing updates from friends this might have proved to be fatal but on reflection we made some good product decisions that helped to keep our core user base of devoted “Foundbiters” interested and enjoying their experience despite not having a social network-sized audience. We knew very few of the anecdotes and techniques I’ve summarised above, and as I've written about previously we didn't completely solve this problem, but I think we stumbled across a combination of them that worked well for us.

The Foundbite Windows Phone app showing the featured feed

The Foundbite Windows Phone app showing the featured feed

Early on we decided that we didn't want to present new users with a blank feed when they opened the app or before they followed anyone, that's deeply uninspiring. Instead we built a featured foundbite feed which all users, logged in or not, could see (Fake It!). The featuring process pushed a new foundbite uploaded by someone using the app into the feed of every other user every 3 hours (we didn’t want to inundate users). These curated foundbites were interspersed between those from people you were following, much like Twitter ads. The feed ensured there was something fresh and interesting to view every time you opened the app (whether you were just curious, or if none of the new people you were following had uploaded anything new). The diverse nature of these curated foundbites, from protests in Ukraine, to football in Rio, helped show off the app and what it was about (Focus on Quality!).

As part of the process I sat down every morning to see what had been uploaded the previous day and selected those that would be published in the next 24 hours (Do Things That Don’t Scale!). Although time consuming, it was a fantastic glimpse into how people were using and reacting to our service. Users whose foundbites were featured were notified to show we were valuing their contribution, hopefully encouraging them to post more. In an ideal world this process would have become unscalable and impossible for me to do due to the sheer number of new uploads but alas that wasn’t to be (though that can bring another set of problems).

Focusing on promoting these quality contributions had one drawback which was mentioned a couple of times: some felt the quality was too good and therefore felt discouraged from posting their own. Reflecting on this, we could have sought to make the app seem less daunting though our situation doesn’t seem that different from many others areas of the internet where 1% of users will actively create new content and other 99% largely lurk (the 1% rule).

Of the other techniques I’ve spoken about above, we didn't really have a key demographic in mind that we felt we should particularly focus on (a mistake in retrospect). We also chose to not focus on providing a utility though there was one user who was an AmTrack employee who recorded audio notes with pictures of wagons he was inspecting – fascinating!

In retrospect the featured feed was the best product decision we made. It was the main reason people were still coming back to us when we were exclusive to Windows Phone. Without it, the vast majority would have simply opened the app to a blank feed and an empty friends list and closed the app for the first, last and only time. There’s a lot I’d do differently if I were to build Foundbite again but the featured feed would be one of the first things I’d keep.

Thanks for reading. I'd be really interested to hear your thoughts on these, and if you think there are any other areas or good examples I've missed out.

Many thanks also to Stevan Popovic for helping me edit this post.