Meet LeWeb 2012 winners Qunb: “a supermarket for numbers”
We’re living in the age of big data. That’s the gospel, and it might even be true. But as Jean-Baptiste Théard and Cyrille Vincey noticed in their professional life as a consultant, it’s still damn hard to find good data. It’s even harder to find data that’s actually usable. And where there’s frustration also lurks opportunity, they thought, so they started Qunb, a startup that wants to bring useful, usable and dare we say pretty data to your fingertips.
WE WON #LEWEB! YEAH! WE. ARE. ROCKING. DATA!!!!
— qunb (@qunb) December 6, 2012
Jean-Baptiste Théard (cofounder and Qunb’s “data nerd”): “A supermarket for numbers, is what we’re trying to create. A one stop shop to find numbers which you can also use in a very easy way.”
Jean-Baptiste Théard and Cyrille Vincey met each other when Jean-Baptiste was working as a consultant in Cyrille’s boutique consultancy firm, he explains. “We were both strategy consultants, also doing some operational stuff. We needed data every day of the week, and we were constantly frustrated by how hard it was to find what we needed. We also realised that thousands of people are frustrated by that. That’s how the idea came up to make a business of it.”
Creating a big data business: how to dream big while being lean
Actually creating a data business was less easy than having the idea, he says. “We sometimes joke that we’re three startups in one: the first is the team that has to acquire the data – by grabbing data all over the web but also by building partnerships with data firms to feed our platform. The second is the back office, where they’re doing a lot of big data crunching with Hadoop and stuff. And the third is more business intelligence oriented: those are the guys trying to make sense out of the data and present them in a nice way.
In fact, their background as consultants was a handicap, Jean-Baptiste admits. “We weren’t familiar with the lean startup ideas – at all. That led us down some rabbit holes that, in retrospect, we shouldn’t have gone into (laughs). Cyrille was an entrepreneur, but no web entrepreneur. We knew no one in the tech scene, and we never even built a website – it was quite a switch. From working with big private equity firms to building a site – that was actually a tough nut to crack.”
“At first we thought we would just raise a lot of money to crack the problem, because that’s how we were used to working, see? We thought we’d raise three or four million dollars and then start building. Unfortunately, that’s not how startups work, as we found out. We even moved to the US for three months, trying to raise money by showing a prototype that we built ourselves. It was pretty easy to get access to vc’s, but getting money from them was another thing entirely. We had no track record as a tech entrepreneur. They just wouldn’t bet on us. They said: come back when you have a working prototype.”
“When we realised we weren’t gaining traction this way, we’d spent almost a year. Actually, we were saved by season 2 of Le Camping (the Paris based accelerator where the docTrackr team also went, ed.). Everything moved very quickly after that – we got funding and started building a team.”
Qunb exited LeCamping with 500 000 euros in funding, that they raised from 2 business angels and a family office, and a clearer understanding of what kind of business they wanted to make.
Big data: everything is tougher than you think
Still, while Big Data is sizzling hot as a topic and as a business trend, being a data startup is a lot harder than you might expect, says Jean-Baptiste.
“Everything about data is harder than you imagine. Everything. Hunting data is very tough. Then, you have to do data cleaning: also tough. Our first idea was to present really pure data – no errors, always interoperable, with very complex semantic tech behind so that every data point would be enriched. Well: that wasn’t realistic. So we decided to move faster and postpone the data enrichment for later.”
It sounds like switching gears from big corporate environments to startup was the big issue for Qunb, I suggest. Jean-Baptiste agrees. “We were used to how big SAP projects move, and that’s what we were used to – we weren’t afraid of building a big project before we executed (laughs). While being a startup is all about doing a fast prototype, iterate, we really had to learn that.”
“We also weren’t used to the management – how you build a team with no funding and no track record in the business. That has been tough as hell. In the end, we settled on yes, we’ll iterate fast and do prototype, but on the other hand, we do want to keep our ambitions – which for us was in our core DNA. Because we figure this ambition leads us to people who are also ambitious. We’re not trying to build an app with a 1000 users – we really aim at hundreds of thousands of users. ”
“At the same time, we want to be humble, realistic and opportunistic. Realistic because we do things step by step, based on user feedback. And opportunistic – whenever we get a business opportunity, we will at least dig into it to see if it’s valid. The advantage of being in big data is that it’s a hot field, and a lot of large enterprises have big data projects. That’s starting to be a business for us, big companies coming to us with their data to analyse it against our data. There are some very specific marketing questions that you can’t answer with only outside data, but you can’t solve them with only internal data either. But with Qunb, combining internal and external datasets, you can do it with a couple of clicks.”
“That seems like it will become a business that can at least sustain us for the next couple of months, so that’s nice.”
Qunb user experience and user interface: the next frontier
Giving Qunb a little spin gave me the impression that it’s easy for a user to become overwhelmed with all the data, I say. There is so much of it that you can imagine visitors to just give up.
Jean-Baptiste Théard: “That’s one of the things that we need to address. We actually have two main concerns at this time: one is not having enough content to deliver a nice experience. The other is the search results. It lacks some context. We’re working on adding that now.”
“UX and UI (user experience and user interface, ed.) are a big concern for us. We worked with an ergonomist to get it right but we’re still not happy with it. The thing is that you just don’t look for data the way you look for an internet page. It’s important for us to understand the exact mental steps that people go through when searching for data.”
Who are your main competitors? Wolfram Alpha?
Jean-Baptiste Théard: “No, not really. The difference is that they have a closed database, meaning they select curated results. We don’t, we offer results from different providers. We give you the choice. Also, they put a lot of energy into the natural language capabilities of their search engine. We are not able to do that, so we are more about refining your query as you go along.”
“Our biggest competitor is actually a firm from Iceland, who started a year or so before us: Data Market. They are evolving to a domain centric platform, starting with energy. We’re still broader than them. They focus on the user side – we focus on the provider side. A lot of data firms are still selling unprotected PDF’s. We could bring them the tech to add a link under each table where you can access the clean, usable data. In the next months, we’ll follow a technology approach rather than a market approach.
It’s not unthinkable, says Jean-Baptiste, that this strategy of offering technology to data providers might become a pivot point for Qunb.
Qunb is a semifinalist at LeWeb and will be among 16 startups who pitch at the LeWeb startup contest on December 4, 2012. Three finalists will pitch to the entire LeWeb audience on December 6. Good luck, guys!
Receive our top stories in your mailbox every day!
Sign up here:
Know more about this subject?
Powered by Facebook Comments