Chris Lynch, a partner in the technology group of Boston-based venture capital firm Atlas Ventures, spends his days scouring the tech landscape for the best and brightest big data startups. And, as the former CEO of Vertica Systems, the big data startup that was acquired by Hewlett-Packard in March 2011, he knows exactly what to look for.
Lynch recently spoke with CRN about the big data landscape, vendors that seem ahead of the curve, and what it all means for the IT channel. Here's a glimpse into the interview.
CRN: What do you look for when investing in a big data startup?
Lynch: I think there are three primary challenges that have to be overcome in big data: scale, security, and simplicity. So I look for companies that are addressing one or all three of those. There are over 100 startups in Boston in the big data space currently, so it is crowded. I look for investment pieces that make sense, where [companies are] addressing an unsolved problem in the market. Probably most importantly, I look for strong entrepreneurs that have the will to do it, the inspiration to do it, and the commitment to do it, because it's really hard to start a company in tech [with] all the consolidation of the bigger companies out there -- you basically are doing it against all odds. So it takes a special kind of person who isn't daunted by that. I really look for quality people who are focused on areas where I think there is an opportunity because the incumbents aren't going to go [there].
CRN: Why has Boston become such a hub for big data innovation?
Lynch: First and foremost, we have 5,000 computer science graduates every year in Boston, so that creates a lot of raw material. We also have a high concentration of machine learning students in the area, and there are 35 universities in 10 square miles in Boston. So there's a unique advantage that Boston enjoys.
I also think you have a legacy of strong enterprise companies that have been born in this area. When you look at big data and the origins of it, a lot of this movement is centered around guys like Jit [Jitendra] Saxena, the founder of Netezza. You have Netezza, you have my old company Vertica, you have Cognos, you have Essential Software, all these companies in the area and the next generation of entrepreneurs with Greenplum/EMC. All of these people are sort of the next generation of entrepreneurs.
CRN: Which vendor incumbents do you consider to be ahead of the curve with their big data offerings?
Lynch: IBM, by far. They are the most advanced [with big data] and enjoy the most success. I'd follow with EMC. I'd guess I'd say IBM, EMC and Oracle in that order.
CRN: What is it about those three companies that make them stand out?
Lynch: Significant investment in the space. They are investing in people, and acquiring technologies and putting it all together, where I think other people are sort of dabbling in big data and made maybe one move. But the problem is, it's such a ubiquitous space, and it's big, so you really need to build out an entire stack to deliver real solutions to customers. And those three vendors, particularly IBM, have been the most aggressive about building out an entire organization around big data and building out a complete portfolio of solutions to deliver to customers. IBM, in particular, has probably outspent all the others combined.
CRN: HP is now selling its Vertica solutions through the channel. What advice do you have for HP VARs, or VARs in general, trying to jump on the big data opportunity?
Lynch: It's hard for channel partners to do much currently with HP Vertica because they are not providing an entire [stack] solution that they can sell. I am confident Meg [Whitman, HP's CEO] will reignite investment activity and acquire the other assets needed to provide partners with a full solution. The other thing I am hopeful for is message clarity to partners regarding the myriad of offerings HP has and the various channels used to get to market -- there has to be a clear story around their efforts to sell SAP HANA, Microsoft SQL Server and Vertica. Again, I expect this is being worked on or will be soon under Meg's leadership. It's an important issue because for the channel, it's a significant investment up front in big data training, and you can't make that commitment as a reseller unless you know that your partners are serious to being in the business and are committed to the channel. So that's one challenge there, but I think the more fundamental one is having a suite of products. So for instance there are no BI tools, no visualization tools, no ETL tools. All of these are required with a database to help you solve problems.
Conversely, if you look at IBM, IBM owns multiple products across each of those categories. So a partner has a large palette of potential solutions to deliver to customers. At the end of the day, that's what keeps the channel alive -- adding value on top of these products for customers.
CRN: So partnering with vendors that touch every piece of the big data story is where VARs should place their bets?
Lynch: At the end of the day, [in the] channel, your job isn't to push products for a vendor -- it's to deliver solutions to your customers. So work with the companies that provide the best and broadest suite.
CRN: Do you think we'll start seeing more big data startups opening their doors to the channel?
Lynch: There is absolutely an opportunity in big data, and it's now, and I'd be in front of it if I was a reseller. I think there's a shortage of knowledge and know-how and that always creates a void in the marketplace that the channel can fill. So I would encourage them to engage early, and build up the expertise.
In terms of small companies, I think it's similar to the maturation of every new tech sector. Even in networking, when it started, Cisco and Wellfleet sold primarily direct. ... And then over time, ultimately they all embraced the channel. I think the same will happen here. And I think the forward-looking channel guys are guys that embrace the technology proactively and sort of pick and build their portfolio that complements one of these bigger guys, like IBM, because none of them have all the solutions. They certainly can't innovate at the rate that the market can; that's why acquisitions exist. The companies are simply ... at their maturity level, they just haven't gotten to developing channel programs. If I was a VAR, I would go to best-of-breed companies in each of these sectors and proactively try to deal with them.
CRN: Is big data a priority only for big companies?
Lynch: It's very much a universal problem. I can give you an example. There's a Boston-based company called InsightSquared. ... They sell analytic insights to small and medium-sized companies in a SaaS model. They basically bring big data analytics to these companies and have been very, very successful and grown very quickly.
PUBLISHED APRIL 8, 2013