Job trends highlight post-M&A analytic database investment

When I was messing around with Indeed.com job trends the other day I was struck by an interesting trend relating to the five recent major M&A deals involving analytic database vendors: Netezza, Sybase, Greenplum, Vertica and Aster Data.


netezza, sybase iq, greenplum, vertica, aster data Job Trends graph

netezza, sybase iq, greenplum, vertica, aster data Job Trends Netezza jobsSybase Iq jobsGreenplum jobsVertica jobsAster Data jobs

The trends aren’t immediately obvious from that chart, but if we break them out individually and add a black dot to indicate the approximate date of the acquisition announcement it all becomes clear.


(Note: scale varies from chart to chart)

While the acquisitions have accelerated job postings for all acquired analytic databases, Greenplum has clearly been the biggest beneficiary. Indeed.com’s data also explains why this might be: EMC/Greenplum is responsible for over 50% of the current Greenplum-related job postings on the site (excluding recruiter postings).

Greenplum had 140 employees when it was acquired in July 2010. Based on the hiring growth illustrated above, EMC’s Data Computing Products Division is set to reach 650 by the end of the year.

Netezza started with a much larger base, but IBM is expected to increase headcount at Netezza from 500 in September 2010 to a target of 800 by year-end. Thanks, no doubt, to Netezza’s larger installed base, IBM is responsible for just 7.7% of Netezza job postings.

This highlights something we recently noted in a 451 Group M&A Insight report: in order to make a considerable dent in the dominance of the big four, any acquiring company will not only have to buy a data-warehousing player but also invest in its growth.

While Vertica and Aster Data are both heading in the right direction, we believe that HP and Teradata will have to accelerate their investment in the Vertica subsidiary and the new Aster Data ‘center of excellence’ respectively.

HP recently told us headcount has grown about 40% since the acquisition (it wasn’t being specific, but Vertica reported 100 employees in January). HP/Vertica is currently responsible for 13.9% for Vertica-related job postings on Indeed.com

We had speculated that Teradata would need to similarly boost the headcount at Aster Data beyond the estimated 100 employees. Teradata/Aster Data is responsible for 24% of job postings for Aster Data.

But what of Sybase? While Sybase IQ also has a larger installed base, SAP/Sybase are responsible for just 6.4% of the Sybase IQ-related job postings on Indeed.com. The Sybase IQ chart illustrates some common sense investment advice: the value of your investment can go down as well as up.

3PAR’s rocketing valuation; belief, scarcity and little bit of luck

I had the opportunity to meet up with David Scott, CEO of 3PAR, the current belle of the ball in storage as the bidding war between HP and Dell continues to intensify (read our analysis of the deals for free by clicking here). Though discussion of any details concerning the acquisition process was strictly off limits, Scott provided some interesting color on why he believes the battle for 3PAR is taking his company’s valuation to unprecedented levels.

Actually, our conversation was a continuation of a discussion that we began over dinner at 3PAR’s analyst event in California a few weeks ago. During that discussion I asked Scott why 3PAR hadn’t yet been acquired; his response pretty much described the events that are now playing out. Scott believed there was in effect a Mexican stand-off taking place; multiple vendors would potentially be very interested in making a bid for 3PAR, but a fear of being outbid – and losing out – was holding them back. Thus, for the time being it was generally in all potential suitors’ best interests for 3PAR to remain independent.

Why Dell decided to break rank and shoot first is not entirely certain at this point — though HP losing its CEO may have been a trigger — and was certainly not on the menu for discussion with Scott. But the CEO was more forthcoming on the reasons for this fear of being outbid, which are rooted in 3PAR’s scarcity; ie the belief that there is no viable alternative acquisition target to 3PAR. The bidding war that has played out since Dell made its first offer would appear to support this. Why? Scarcity seems like a crazy assumption to make in an industry that is constantly spitting out new startups.

Scott’s reasoning for this scarcity has both demand-side and a supply-side dimensions, both of which have taken a couple of turns of the IT cycle to come to fruition. On the supply side; cast your mind back a decade, and the IT world was alive with the prospect that ‘xSPs’ (especially storage service providers and application service providers) would play a transformative role in delivering IT as a service; Cloud 1.0, if you like. What these xSPs required was a way of building these services on a scalable, secure and shared technology infrastructure. Unfortunately for SSPs such as Storage Networks, the infrastructure components to build such a stack were not available, and the entire model collapsed under the weight of having to build dedicated systems for each customer.

But the promise of the xSP model was also the catalyst for innovation at all levels of the IT stack. There was nothing inherently wrong with the model of IT-as-a-service – it was, and remains, highly attractive. What was needed was a new underlying architecture that could provide the required scale and flexibility as cost effectively as possible, such as blade servers, virtualization software and ‘utility’ storage.

Thus, as interest in the xSP model began to build, VC money started to flow into storage startups developing ‘carrier grade’ platforms; in particular Cereva Networks, Yotta Yotta, Zambeel and 3PAR. Only one of those companies managed to make a go of it; the rest succumbed to the same burst bubble that did for the xSPs. Cereva (which had raised almost $140m in VC funding) collapsed in 2002, Zambeel (which raised around $66m) closed its doors in 2003, while the assets of YottaYotta  (which took in around $100m) were eventually acquired by EMC.

As the only remaining player in this new generation of high-end storage platforms, Scott says 3PAR was in a unique position. Perhaps even more crucially, these failures meant VCs were now loath to invest in high-end storage startups; even if the next “3PAR killer” came along, it would have struggled for funding. Instead, VCs turned their attention to startups targeting the mid-range storage market – LeftHand Networks, EqualLogic, Compellent, Pillar — which was growing much more quickly than the now-slowing high-end space.

Scott admits 3PAR came under pressure to target the mid-range space more aggressively (and it did release smaller versions of its InServ arrays), but the company’s core efforts remained on the high-end, with a continuing focus on direct-, rather than channel-based, sales. Scott and his team remained as convinced as ever that ‘utility’ computing was real, and would eventually pay dividends via 3PAR’s scalable storage platform.

In particular it found traction with the next generation of service providers –such as managed hosting providers and telcos – that, subscribers to the cloud model attest, will collectively host the vast majority of the enterprise IT workloads of the future. Indeed – and this is where the demand-side argument comes in – the post-recession reality for organizations of all types and sizes – from financial services giants to local government offices – is that they are looking for more cost effective methods of running their IT processes.

These service providers differentiate themselves on quality of service and cost, and the only way of achieving this – according to Scott – is through best of breed IT infrastructure. Scott and co have made much of the fact that seven of the ten largest service providers by revenue are 3PAR customers, and we’re sure this point is not lost on HP, Dell or any other would-be acquirer.

Of course, with hindsight it’s easy to make the facts fit a story, but we’d note that 3PAR‘s own strategy and messaging has scarcely changed since day one. 3PAR has always targeted ‘utility’ computing, and has stuck with the term as the rest of the industry dispensed with what to them was just the latest buzzword (for proof, see the first research report (451 clients only) we wrote on 3PAR, back in 2002). Indeed, for 3PAR and Scott, delivering IT as a utility is an integral part of its proposition; it gets to the core of why the company believes it is different, and why (at least) two giants of the industry are prepared to pay well-over-the odds to own.

Bridging the “storage-information” gap

When Nick first unveiled this blog last month he rightly noted ‘storage’ as one of the many categories that falls into a capacious bucket we term ‘information management.’ With this in mind he reminded me that it would be appropriate for the 451 Group’s storage research team to contribute to the debate, so here it is!

For the uninitiated, storage can appear to be either a bit of a black hole, or just a lot of spinning rust, so I’m not going to start with a storage 101 (although if you have a 451 password you can peruse our recent research here). Suffice to say that storage is just one element of the information management infrastructure, but its role is certainly evolving.

Storage systems and associated software traditionally have provided applications and users with the data they need, when they need it, along with the required levels of protection. Clearly, storage has had to become smarter (not to mention cheaper) to deal with issues like data growth; technologies such as data deduplication help firms grapple with the “too much” part of information management. But up until now the lines of demarcation between “storage” (and data management) and “information” management have been fairly clear. Even though larger “portfolio” vendors such as EMC and IBM have feet in both camps, the reality is that such products and services are organized, managed and sold separately.

That said, there’s no doubt these worlds are coming together. The issues we as analysts are grappling with relate to where and why this taking place, how it manifests itself, the role of technology, and the impact of this on vendor, investor and end-user strategies. At the very least there is a demand for technologies that help organizations bridge the gap – and the juxtaposition – between the fairly closeted, back-end storage “silo” and the more, shall we say, liberated, front-end interface where information meets its consumers.

Here, a number of competing forces are challenging, even forcing, organizations to become smarter about understanding what “information” they have in their storage infrastructure; data retention vs data disposition, regulated vs unregulated data and public vs private data being just three. Armed with such intelligence, firms can, in theory, make better decisions about how (and how long) data is stored, protected, retained and made available to support changing business requirements.

“Hang on a minute,” I hear you cry. “Isn’t this what Information Lifecycle Management (ILM) was supposed to be about?” Well, yes, I’m afraid it was. And one thing that covering the storage industry for almost a decade has told me is that it moves at a glacial pace. In the case of ILM, the iceberg has probably lapped it by now. The hows and whys of ILM’s failure to capture the imagination of the industry is probably best left for another day, but I believe that at least one aim of ILM – helping organizations better understand their data so it can better support the business — still makes perfect sense.

What we are now seeing is the emergence of some real business drivers that are compelling a variety of stakeholders – from CIOs to General Counsel — to take an active interest in better understanding their data. This, in turn, is driving industry consolidation as larger vendors in particular move to fill out their product portfolios; the latest example of this is the news of HP’s acquisition of Australia-based records management specialist Tower Software. Over the next few weeks I’ll be exploring in more detail three areas where we think this storage-information gap is being bridged; in eDiscovery, archiving and security. Stay tuned for our deeper thoughts and perspectives in this fast-moving space.