NetApp Makes A Move ... Sort Of

The twittersphere was all abuzz this morning.  Netapp agreed to acquire Bycast, a 60-person startup.  The press release is here.

Part of me is fascinated with any acquisition in our industry.  Deals are complex things, and have so many angles to them.  And, being at EMC, we've seen more than our fair share over the last decade or so -- including getting an inside look at many we've passed on.

One of my friends in M&A has a clever saying -- "the best deal is the one you don't do".  

That being said, if you're expecting an unreasonably negative response from me, I'm going to disappoint you.  Like all deals, there's some good points and weak points.

To Start With ...

The "long tail" of information storage is becoming increasingly more important.  There's more data, it's being kept around longer, it has to be accessible, there are more rules and regulations about how that should be done, etc.

Certain industries (e.g. helathcare) are feeling the pain now.  Others will follow.  For lack of a better descriptor, I'll use the term "digital archive' to describe this space, although that's woefully imprecise.

If you look at the storage market strategically, you certainly want some sort of play in this space.  Indeed, vendors like EMC are making multiple plays -- more on that later.

Requirements Are Different

Information that wants to be archived is deemed to be important - -otherwise, you wouldn't be interested in saving it, would you?

That implies a suite of characteristics that must be met: reliably stored, rules to manage how things are stored, recoverable if corrupted, access control and security, maybe even audit trails.

Geography is a key aspect of most serious solutions.  Information is better protected as more distance is involved; frequently creators and users of digital archives are in multiple locations.

Once the mandatories are in place, we turn to the cost side of the equation.  

Storage efficiency is first and foremost -- single instancing, spin-down, compression.  Strangely enough, ordinary dedupe isn't usually a strong play in most digital archives -- frequently, the information in each element is relatively unique, for example.

The operational side isn't much of a concern -- until things start getting big and complex.  There's a big difference between a few terabytes in two sites, and dozens of petabytes scattered across time zones.

And, finally, there's application integration.  Digital archives are usually part of a workflow that puts information into the archive, and gets it out again as part of a larger business process.

Not everyone needs the same thing, of course.  The needs of a small local hospital, for example, might be very different than, say, a regulated insurance provider.

Where Bycast Fits In

Bycast offers a software layer that does some of this good stuff -- but not all of it.

It runs as an application, presents a file system with some HTTP extensions, and uses a wide range of storage that the customer might already own.

If you're a small IT shop (back to the hospital example), this sort of approach can be attractive -- re-use existing assets, meet the basic requirements, low startup costs, etc. -- get something in place and move on.

But as the requirements get larger and more sophisticated, this type approach gets far less attractive for many.

Where Things Start Breaking Down

Now, a lot of this is going to sound negative, but it's not meant to be -- different customers have different requirements -- that's what makes the world go 'round.

The first pain point is likely to be management -- all of the storage arrays used in this type of solution present themselves as distinct points of management.  

That's not an issue when you have two.  It's an issue when you have 20, or 200.  More effort, more people, more cost.  Hard to avoid with this sort of solution.

The second pain point is likely to be compliance and auditing.  More stringent regulations have this notion of "chain of custody" which are difficult to meet with layered (vs. integrated) approaches.  

How do you prove that no one tampered with the data in court?

The third pain point would be metadata support.  Part of the appeal of a true object-based model is its ability to exploit integrated metadata to manage the information better.  In any file-based approach, the metadata has to be kept externally.

Applications populate the metadata fields for the objects they create.  The metadata is used to automatically drive policy regarding placement, retention, access control, security, protection, etc.

Again, not being able to do this isn't a problem for a small environment with a single use case -- like health care records in a smaller hospital.  You have one or two apps, and everything gets treated the same -- they're all health care records, for example.

This becomes more of a problem when you have one shared service that's serving multiple needs across the organization.

The NetApp Angle?

So, why do I think NetApp did this deal?  More tactical than strategic, in my humble opinion.

If NetApp was looking for an advanced technology buy around object-oriented storage with associated cloud delivery models through service providers, there are far better companies to go buy.  

If, however, they were looking for something rather more modest, that had paying customers, and could be sold "as is" without any serious integration on their part, they made the right choice.  They get some nice incremental revenue, and can at least make a claim that they have something -- anything -- in this arena.

The EMC Angle?

Fair disclosure -- EMC has been playing in this sandbox for many years, and with many offerings.

Centera provides a compliant object store that meets the most stringent regulatory requirements.  It offers an integrated management model, geographic distribution capabilities, robust security / compliance / auditing capabilities, as well as a vast library of application and layered integrations.

It's been very, very successful in this space, with many of the same sorts of customers that Bycast has sold to.  It does not, however, run on top of any old storage :-)

Some important Centera concepts found their way into EMC Atmos storage, which brought a serious multi-tenancy model, a robust geographic-aware policy engine and an architecture suitable for service providers offering shared services.  Not to mention insane scalability.  It too is finding some strong fans out there.

On a more pragmatic note, Celerra's FMA (the virtual edition) does much of what Bycast software does, and in much the same way -- as a storage-agnostic software appliance.  It too has found a strong following with customers looking to build more modest digital archives.

The Bottom Line?

Frankly, I was expecting a bolder move from NetApp.  Maybe I was even hoping for one.  But, let's face it, technology acquisitions at NetApp have historically been less than successful, so I can understand the rationale to play it safe.

NetApp gets the opportunity to sell a bit of kit into Bycast's installed base.  That's nice.  But when they compete in the open market, they'll be up against some very stiff competition -- from EMC as well as others.

Game on.

Comments

[ChucksBiggestFan]

1 out of 5 stars.

One of those posts where you have to say something because you are the nominated guy at EMC. I make the extra effort to say this because your blog is somewhat vacant in terms of valuable, informative content for your target audience.

This is just another infomercial for EMC to harass the competition and talk vaguely about how their products solve customer pain but lacking in specifics.

I do find it ironic that you are sledging the competition about unsuccessful product acquisitions. Perhaps you would like to articulate EMC's very long history of acquisitions for the audience? I am thinking a deep and open financial analysis on the Return-On-Equity for each and every one of those acquisitions. Perhaps your M&A buddy can document that information for you and you can summarise it here? As a long-suffering shareholder, I am keen to understand how many of the good investments are weighed down by bad ones. i.e. How could the EMC EPS be improved by more prudent purchases? At least my NetApp shares continue to impress for all the right reasons.

Credit where it is due though, you did refrain from an outright attack on NetApp and Bycast. More a thinly veiled one.

Looking forward to rating your next post.

[Chuck Hollis]

Hi, whoever you are -- thanks for the comment.

I run a "full disclosure" policy here -- you need to disclose your affiliations, if not your real name, if you expect me to post your comments. We get a lot of vendors masquerading as something else here.

If you read through my posts collectively, you'll generally see an honest effort to provide "valuable, informative content" for my target audience. At least, that's what I hear from people. Some posts do this better than others -- this individual post was simply industry commentary, nothing more.

A lot of people weren't able to put the Bycast acquisition in context. I had context, so I shared it. Unfortunately, I happen to work for a competitor, so the bias should be obvious to all.

As far as EMC's string of acquisitions, they're generally held up as very successful in the tech M&A community. I believe very few vendors can match our track record. At some point, I might get around to writing it all down, since it's been an amazing journey.

People thought us crazy when we acquired VMware. They thought us crazy when we acquired RSA. Indeed, the rationale of so many of our acquisitions doesn't become apparent for quite some time.

We do a very thorough study post-acquisition of each and every acquisition, both on a standalone basis, but also in the context of integration and synergy. However, we don't share that information publicly -- and neither does anyone else, as you probably are aware.

As far as relative share prices, you're free to invest as you see fit. Personally, I like our long-term prospects, but everyone's entitled to their opinions.

I look forward to being rated!

-- Chuck