Unless you’ve been stuck in a Wi-Fi or 4/5G deadzone—this rules out Mount Everest, by the way—you probably heard that Google last week acquired Looker for a cool $2.6 billion.

For perspective, $2.6 billion is roughly the equivalent of 1,056 Holzhauer Units® (HU)—with a single HU pegged at $2,462,216, or the total winnings of Jeopardy! champion James Holzhauer during his 32-game run. That’s not far off from Qlik Technologies’ $2.8 billion market cap (1,137HU) prior to its acquisition by a private equity firm and its subsequent delisting from financial markets. It’s nearly twice Cloudera’s market cap ($1.4 billion, or 569HU)—but nearly equal if you factor in the $1 billion+ in cash Cloudera generated over several VC funding rounds. (What’s a couple of hundred million dollars here or there?) It’s three times rival Domo’s market cap ($814 million, or 330HU)—although both Cloudera and Domo, along with MapR (privately held, 114HU in total VC funding), have had especially hard-knock runs of late. So what is Google (300,948HU) getting for its $2.6 billion investment?

• Looker is a BI and analytics platform—with a twist. I’ll say more about this below. Right now, let’s skip to the next bullet point, which helps set up how Looker does things a little bit differently. Hint: Looker isn’t your best friend’s sister’s boyfriend’s brother’s girlfriend’s BI and analytics platform. Nor does it pretend to be.

• Looker does data virtualization—with a twist. Looker uses data virtualization (DV) technology to abstract access to distributed data sources and services. In this way, Looker presents a unified (logical) view of business data—irrespective of a data source’s system or platform context, access APIs, or physical/geographical location. DV is neither a new nor a unique idea, of course, but—unlike DV infrastructure offerings, which used to be marketed as data warehouse replacement technologies—Looker’s DV happily runs against both operational and analytical systems.

• Looker is footless and schema-free. Looker can run in data warehouse-less mode, i.e., directly against operational applications or OLTP databases. It can also consume data from a data warehouse or similar analytical platform. Unlike a conventional analytical platform, however, Looker describes itself as footloose and schema-free: it doesn’t use a conventional data model. Nor does it require that someone (IT, the self-serving analyst) design a presentation layer or create business views to permit BI-like access. Looker doesn’t even do slicing and dicing in the conventional, OLAP-y sense: it supports row-level access to data, permitting ROLAP-like generation of virtual cubes.   

• Looker is a self-service BI and analytics platform—with a twist. Looker’s data access underpinnings and user interface (UI) elements can be configured via a human-readable mark-up language called LookML. (LookML is less XML-like than YAML-like.) Looker describes LookML as analogous to a flexible data modeling layer. It claims that business analysts and similarly savvy users can use LookML to prepare their own data extracts and build their own interactive dashboards. They can also prepare data extracts and build interactive dashboards for non-expert users.

• Looker long ago learned to stop worrying and to love IT. Think of Looker as a BI semantic layer that sits on top of business data. This data can be distributed—i.e., living in remote OLTP database systems and services. It can be centralized, e.g., living in the equivalent of an on- or off-premises data lake, operational data store, etc. In either case, Looker-the-BI-layer simplifies three key problems: (1) single sign-on for data access; (2) the elimination of multiple (and often redundant) point-to-point connections between individual self-service users and data sources; (3) metadata management, data lineage-tracking, and other traditional data management priorities. Looker pitched itself as an IT-friendly alternative to traditional IT-driven (and data management dominated) data warehouse practices. It was designed to be installed and managed by IT, customized—and, in key respects, designed—by self-service users. 

So much for what Google gets. As to what the combined Google + Looker might mean for the rest of us, I’ll defer to my friend Donald Farmer, a principal with Treehive Strategy.

“With their pending acquisition of Looker, Google could be one of the leaders in business analytics. In fact, by some measures, if they integrate Looker effectively, they should be the world's largest BI vendor by far: there are perhaps 50 million or more websites using Google Analytics to understand traffic and user behavior.”

Donald’s take is appropriately nuanced. To those who know him, he’s the maestro di color che sanno, after all. But Donald gets at something that’s at once odd and irresistibly fascinating about the BI-analytic megafauna: even the largest of these creatures shares space in a single ecosystem with a staggering variety of smaller creatures. Before there was Google-the-world’s-largest-BI-vendor there was Microsoft-the-world’s-BI-behemoth. And before Microsoft there were, arguably, Oracle, Arbor, and others. Notwithstanding the size, clout, and aspirations of these mega-vendors, the BI and analytics ecosystem—la cosa nostra, as I call it—adapted accordingly. Farmer predicts something similar will happen this time, too.

“If Google make the Looker acquisition work effectively and become[s] a huge player in the BI space, who suffers? Would you be surprised if I said no-one? Or at least, none of the major players. I expect independent vendors will compete with and complement Google much as they do with PowerBI today.”

Donald doesn’t address the question of whether—at $2.6 billion—Google overpaid for Looker. With more than $100 billion (or 41,216HU) in cash on hand, this seems like the canonical definition of a moot question. And Looker does have a few other things going for it, at least from Google’s perspective. In an article I wrote several years ago, I noted that Looker is a distinctively developer-friendly BI and analytics platform. “[I]ts users tend to be start-up (or …  “born-of-the-Web”) type companies,” I wrote, adding that “its use in the Web world … is much more pervasive” than in traditional enterprise-type BI/analytic deployments.

In this sense, Looker seems like a happy match for Google. And Google, for its part, seems ready to morph into the spitting image of Microsoft BI—or, at any rate, of Redmond’s pre-Azure BI stack, circa, say, 2011. As my friend Mark Madsen pointed out to me today, Google’s own product moves in advance of its acquisition of Looker aren’t just reminiscent of but, er, phonetically similar to Microsoft’s: how else can one explain Google’s “Big Query DTS” data integration service? Did Google intentionally ape the acronym of Microsoft’s seminal SQL-server-based data integration facility? In addition to this, you’ve got Big Query BI Engine, a server component that’s akin to SQL Server Analysis Services. Plus, Looker gives Google a self-service discovery environment that’s at least notionally analogous to PowerBI. Google also boasts not one but two SSIS-like cloud services: Big Query DTS and Cloud Data Fusion. Whether the rest of la cosa nostra realizes it or not, there’s a new behemoth in town. 

A new boss, even. Chances are, it’ll be the same as the old boss-behemoth.

Apropos of the events of this last sennight, I went back and reread a piece that I wrote in November of 2007, written just after IBM acquired Cognos, capping a major consolidation cycle in the classic BI era. I quoted spokespeople from Information Builders, MicroStrategy, and, of course, SAS Institute who assured me that consolidation in the BI/analytics space was actually a good thing for their companies. Much of this was the brave face one puts on in times of crisis. Looking at it this afternoon, however, I realized something: each of these vendors is still with us—and still independent. In the cases of IBI and SAS, this isn’t surprising. Both companies aren’t just privately held, they’re inextricably associated with the charismatic executives who own controlling stakes in them: Gerry Cohen and Jim Goodnight, respectively. Neither man is likely to relinquish control of a company he’s spent his entire life building.

In any case, something that Russ Cobb, then and now an employee of SAS, said to me then resonates to this day. It has to do with innovation. And, yes, it could be dismissed as so much marketing bravura. But I think the subsequent career of SAS itself has borne witness to Cobb’s claim. “We’re innovative. Unlike [Oracle, SAP, and IBM], we aren’t going to get distracted. We’re going to keep innovating.” Selah—and Amen. So it shall be and so be it.

Image credit: https://cloud.google.com/