I had the pleasure of teaching a new case at HBS yesterday on foursquare that I co-authored with Professors Tom Eisenmann and Mikolaj Piskorski as part of Tom's new course "Launching Technology Ventures". Foursquare executives Dennis Crowley, Naveen Selvadurai and Evan Cohen were kind enough to allow us to interview them in preparation for the case, which framed some of their current key strategic issues and looked back on the choices they made in the early days to draw pedagogical lessons of lean start-up best practices, building a platform business, network effects and running monetization experiments.
The foursquare team was consumed this week with SXSW preparations, but we were fortunate to have as class guests Charlie O'Donnell, who wrote the original blog post on foursquare that got many in the community excited about the company, and Andrew Parker, who was an associate at Union Square Ventures at the time of their Series A investment.
As I did with the class a few weeks ago when Fred Wilson visited, I asked the students to pull out their phones and tweet throughout the class. You can see the rich "dialog behind the dialog" here, using the Twitter hash tag #hbsltv. Here were some of the takeaways I had from the class discussion framed around three major questions I posed to the students:
1) Why did foursquare succeed as compared to the same founder (Dennis) in a similar venture (Dodgeball) in a different era and as compared to other teams pursuing LBS services in the same era?
The students concluded that the context around a venture matters tremendously – that smart phones, the explosion of apps and social networking all were important enablers that allowed foursquare to succeed at this particular moment in time. At the same time, the foursquare team was incredibly skilled at applying lean start-up best practices, specifically:
- Product-obsessed founders: both Dennis and Naveen were consumed with the product. Always interacting with users in bars and over Twitter, thinking less about strategy, analytics and monetization and focusing more on a great user experience.
- Hunch-driven: they had deep domain knowledge and didn't need outside studies or market research to guide their prioritization. One of the key takeaways that both Charlie and Andrew emphasized to the students was to be power users in whatever area of focus they choose to develop those instincts.
- Minimum viable product: they didn't wait years and years to perfect the product but instead got it out there to solicit user feedback.
- Modest burn: the company only raised $1.35 million in its series A financing and kept the burn rate at less than $100k per month to make he money last. Dennis wrote a great post at the time of the financing that showed just how product obsessed he was, even after taking the seed money. There's no bravado or BS – just a list of the great features they're going to roll out as a result of having the extra capital.
2) What was the magic of the foursquare system that drove rapid adoption that so many other consumer Internet companies fail to achieve?
- Game mechanic - students really honed in on the playfulness of the service, both the entertainment value and the addictive nature of competing for badges and mayorships.
- NYC launch – the fact that the service started in such a perfect venue gave it great advantage – a highly concentrated, very social community.
- VC validation - having Fred Wilson invest and promote the company helped provide it credibility with an insider crowd that may have provided some strong tailwinds.
- Win-win for all constituents – unlike many services, the students understood a key insight about foursquare: the local merchants make the service. The fact that merchants are so incented to promote, discuss and reward consumers creates a positive feedback loop that transcends the power of a consumer-only service.
- Online – offline combination. Another aspect of the magic of foursquare is that it is not an online only service. In fact, the ability to drive consumers to actually walk into local venues is a special dimension of the service. As one student pointed out: "Facebook tells me what my friends are doing. foursquare tells me where they are and where I can meet them." This is a unique and powerful aspect of the service.
3) Once a company achieves product-market fit and starts to scale, how do their priorities, and burdens, shift?
- Raising money, scaling the team. A rich discussion ensued about what it means to raise big money. When foursquare took $20 million in venture capital at a reported valuation of $100 million, suddently they had transformed the company from a lean, product-obsessed start-up to a company that would need to generate tens if not hundreds of millions of dollars in cash flow to justify a billion dollar valuation. A product-obsessed management team suddenly had to transition to become an operational scale management team.
- Monetization. Consumer Internet companies have to decide when they begin to monetize – as part of the lean start-up experimentation or only after they achieve enough scale to attract partners and advertisers. But it's not a binary decision. Foursquare has run monetization experiments from the beginning, but to justify the big valuation they will have more pressure to show real financing results, perhaps at the expense of the user experience. It takes a strong founder to resist that temptation (think Jesse Eisenberg playing Mark Zuckerburg in "The Social Network", sneering: "No advertising. Advertising isn't cool.")
- Vision/Becoming a platform. What does the company want to be when it "grows up"? To be a generation-defining company and enter the ranks of Facebook and, arguably, Twitter, foursquare needs to evolve from a great application into a platform. But becoming a platform company requires a whole different approach and set of priorities. Do you build out your own features or expand your APIs and invest in supporting third party developers to build applications to your platform. One of the students had coincidentally tried to work with the foursquare API to develop an application and complained that it was very rudimentary and limiting relative to the Facebook and Twitter API.
The verdict? I ended the class by polling the students – who would buy foursquare stock at a $200-250 million valuation (my very rough estimate of the current trading on the secondary market) and who would sell? One third of the students were buyers at that price at the end of the class. Two thirds were sellers. One student pointed out in a tweet that the voters were unfairly negatively biased because only 10% of their classmates had even tried the application and, besides another tweeted, 3/4 of HBS students apparently wanted to sell Amazon short in 1998! Another student tweeted that if there was even a 3% chance that the company could be a $10 billion company, it was worth buying at $200 million. Now there's a future venture capitalist in the making!
Thanks again to the foursquare team for letting us write the case and adding to the HBS community's intellectual capital.
Jeff: many thanks to you for organizing and delivering a terrific learning experience for my students, to Charlie and Andrew for visiting and sharing their insights, and to the team at foursquare for letting us study the company.
In our course, we’re studying the challenges of scaling a startup, and as you outline above, foursquare raises some great questions on that front. Mining the tweet stream you reference, here are some additional issues:
– When does a hunch-driven approach need to give way to a data-driven approach? A hunch-driven approach, fueled by Twitter, email, and face-to-face contact, worked well when foursquare’s founders were similar to early adopters in terms of behavior and attitudes. As one student said, in this context, ‘market aware’ does not necessarily mean ‘market research’. But will this be continue to be true when evolving the service to meet the needs of mainstream users?
– Most foursquare users share location data with fewer friends than they connect with on Facebook, Twitter, etc. This raises the question of how network effects behave as foursquare scales. As penetration grows, will users feel pressure to share with marginal friends in ways that clutter and diminish the experience? In other words, are there diminishing returns to network effects? Can list management and privacy control tools address this concern?
– Can foursquare monetize local merchants without building a big and costly local salesforce?