It’s taken us 4 years to scale Web Summit from 400 attendees to 20,000 and a bunch of physicists have played a big part.
Back in 2010, 3 international journalists showed up, this year it will exceed 1,200. Investors is up from 4 to over 800. Exhibitors from 3 to over 2,000.
With each passing year an increasing number of people ask how has it grown so fast? When I share the answer it’s almost never what people expect, so here it is:
Our growth has been largely propelled by data science. Or more correctly, in my view, network science. While conference companies hire event managers, we hire physicists with PHDs in areas like complex systems and network analysis. They then apply that knowledge and understanding to the task of creating and optimising real life networks. After all a conference is a network, albeit a momentary one. We love stuff like Gephi, NetworkX and Datasift, and algorithms like eigenvector centrality, Force Atlas and Fruchterman-Reingold. But what does that even mean?
Put it this way: While conference companies typically hire experienced event planners, we hire computational physicists, applied statisticians, engineers of all shapes and sizes, some folks who know a thing or two about machine learning and AI, and then some awesome front end developers.
While traditional conference companies fret over manually curating seating plans, compiling speaker lists and handpicking invites for networking events, we approach the challenge from a technical and mathematical point of view. We build algorithms that take into account who you are and who you might benefit from being on a pub crawl with or at a table with or in a meeting with.
In other words, we “engineer serendipity” at the scale of 20,000 attendees. Put another way, the people at your table or on your pub crawl at Web Summit are neither a random collection of attendees nor a manually curated group of attendees, but rather the product of algorithms.
In truth, there’s an invisible (engineer led data-driven) hand at work before, during and after Web Summit helping improve serendipity. And slowly we’re expanding that approach to other conferences around the world. Back in Dublin at Web Summit, that engineer-led data driven approach pervades every aspect of what we do: who we focus on getting to Web Summit, who we focus on connecting at Web Summit, and who we help connect post Web Summit. And that approach, that invisible hand, as odd as it sounds at first, seems on balance to work.
Arguably no technology conference in history has grown faster. Somehow we’ve achieved that growth with no background in the conference industry and no resources to speak of, and all from a pretty peripheral location called Dublin. But what we lacked in experience, funding and location, we’ve made up for by building software and using data.
But that doesn’t mean we’re anywhere close to really truly disrupting the conference industry. Software might be eating the world, slowly disrupting industry after industry. But in our minds, we’re only starting to barely nibble at the giant and antiquated industry of conferences, events and trade shows. And that industry isn’t a multi-billion dollar global industry, but rather a trillion dollar industry.
In the years to come, if we don’t manage to take a tiny slice out of the the big old conference industry cake, then I’m sure many other tech startups will. It’s just a matter of time. No industry, no matter how big, no matter how seemingly ill-suited for software and data science, is immune to disruption over time.
See you in Dublin, and if not at one of our conference in 2015.
If you like lots of data and like code drop me an email: firstname.lastname@example.org.
We’re always looking for talented people across a huge range of fields including computer vision, complex systems, network analysis, machine learning, data architecture and more. We also need PMs. Or if you’d just like to work on some interesting non-obvious challenges, you’ve got my email.