Netscape founder's web BI investment targets IPO float
GoodData CEO on common ground with eBay inflater
Roman Stanek has backed the "right thing" at the right time twice during his career in tech. Is he now onto his third?
In 1997, the Czech-born entrepreneur got behind a team of local programmers writing a Delphi-like Java IDE, called NetBeans.
Java had only just been invented by Sun Microsystems' James Gosling and the air was ripe with possibility; everybody felt Java was a big deal, but nobody was sure where to use it.
On the hunt for some good Java tools to help stake its claim in the gold rush, Sun bought Stanek's NetBeans company in 1999 for an undisclosed sum. History will record that Sun blew it with NetBeans, leaving others to prosper on Java.
Stanek's next venture was Systinet in 2001, building a registry to help keep track of companies' assets. It was the service-oriented-architecture bubble, with people figuring out how applications could be "mashed" together and turned into services that discover each other. Application testing and monitoring specialist Mercury Interactive bought Systinet for $105m in 2006 and HP bought Mercury soon after, for $4.5bn.
"My personal goal is to take the company to IPO - that could happen in a three- to four-year time frame" – Roman Stanek
Stanek's latest muse is business intelligence as a service. He's now chief executive of GoodData, a company he founded in 2007.
Unlike his first two ventures, GoodData is a keeper and he has no plans to sell – not immediately, anyway. Instead, Stanek is talking in terms of going public, the current fad in exit strategies among Silicon Valley start-ups. We've had LinkedIn and Pandora so far this year bringing relief to early-stage investors with Groupon supposedly next.
"Unlike NetBeans and Systinet, where there was no plan to make them independent companies, there's a chance GoodData could become a large BI company," Stanek told us during a recent interview. "I have no plans to exit in the next couple of years. My personal goal is to take the company to IPO – that could happen in a three- to four-year time frame."
BI is a saturated market. Never mind the existence of incumbents like IBM, Oracle and SAP there's plenty of new players playing exactly the same card as GoodData. They are arguing that customers of the incumbents want relief from the cost and complexity of using the big stuff.
GoodData provides a set of operational dashboards, reporting and data warehousing. Like many other companies, GoodData claims it can deliver these at fraction of the cost and of the complexity of rivals. But unlike some in BI, GoodData comes on top of a multi-tenant web-based architecture. It also adds in extract transform and load, and it connects to your existing data sources.
Stanek is selling mostly through partners. Initially, GoodData was just another BI hopeful approaching customers directly, a process that in the enterprise meant 18-month procurement cycles. That's not very helpful if you're a CEO with entrepreneurialism in your blood at a four-year start-up working towards IPO within the next four years.
Now Stanek's working with partners who are doing the selling or getting the traffic online: twenty per cent of business comes from OEMs and 40 per cent via Salesforce. Stanek claims more than 100 large "direct customers" and 2,500 "indirect customers" who are OEM customers.
The company is selling 25 BI applications that plug into analytics from Salesforce, Facebook, Google, Microsoft and SugarCRM. GoodData BI tools are also now available through Twilio, so users of the service can pull up interactive dashboards that display their voice and SMS usage rather than having to use complex tools or slum it in Excel.
The change in strategy took place last summer. "We realised that selling BI sucks," Stanek said. "If you go to a customer and sell from scratch the velocity of the sale is so slow you become a traditional BI player. We had to come up with a model that let us go much faster.
"This is a completely different approach... sales cycles are much faster. This lets us sell to business people – not just the traditional BI audience."
Backing this approach is former Netscape wunderkind Marc Andreessen, whose venture fund Andreessen Horowitz recently led a $15m round of funding for GoodData. It followed earlier seed funding from Andreessen Horowitz a few years back; GoodData now has $29m.
To an extent, Andreessen and Horowitz are bucking the preference among VCs for investing only in web and cloud companies by going with GoodData. Web and cloud companies promise quick growth based on consumer uptake and, therefore, a reasonably rapid return on investment or rapid exit through an acquisition. Salesforce and Google have been the preferred buyers in recent years.
Today, Andreessen has become one of those helping float the Silicon Valley tech bubble of overblown investments and valuations. He helped persuade Microsoft's chief executive Steve Ballmer to pay over the odds for the loss-making web telco Skype, getting Microsoft to take it off eBay's hands.
Ballmer paid $8.5bn and for a company that Andreessen and co held a $1.9bn stake in. Andreessen is thought to have more than tripled his own $50m investment in Skype and is now feted by The Economist as the man who took on Microsoft at Netscape – his browser was ultimately crushed by Internet Explorer – and is now stirring things again.
The Andreessen factor
Andreessen's philosophy is simple: "Software is eating the world". He therefore sees innovation as a permanent force and says he reckons that networking and storage are about to experience the same changes that ripped through servers a few years back and that paved the way for Facebook and Google. Translated: expensive, proprietary boxes will be replaced by low-cost commodity systems, creating the conditions for a new generation of companies.
Interestingly, Andreessen – like Stanek – spent time at HP: the computer-maker bought Andreessen's post-Netscape start-up Opsware for $1.6bn in 2007. Opsware specialised in server and network device provisioning, configuration and management.
With that in mind, when it comes to GoodData, Andreessen is simpatico with the idea of a long (ish in Silicon Valley terms) build to IPO rather than a pulling another NetBeans- or Systinet-style rapid turnaround after a few years, or even another eBay.
"They [Andreessen and Horowitz] speak about it [the investment] as GoodData has the potential to become a $1bn company," Stanek said. ®