Full disclosure:  I worked for EMC under the Documentum brand until recently.

Dave DeWalt discusses why EMC is so successful in spite of the fact that most software company mergers result in failure in an article on SandHill.com.  It’s a great read for anyone interested in software M&A.

In the article, he discusses some of the inherent challenges faced by the $1.8B acquisition of Documentum in 2003 such as geographic disparities and differences in core competencies.

I have incredible respect for this team, but not solely due to my lengthy history with the company.  I have seen this team execute from the inside —- they understand how to build/sell software and how to buy/integrate software companies.  Rinse and repeat.
Link:  Making Software M&A Work

Apple has announced that it has released all 30 employees in their Bangalore office who were supposed to end up comprising 600 support and development staff in total. Offshore product development and support is not a model that works for all companies and environments, but surely the ramifications were known by Apple’s due diligence.  They did their due diligence, no?
The fact that Apple pulled the plug so soon into their offshore development and support efforts really says something about the management or the lack thereof. Did they learn something in the short two months that they were in operation that justified a show-stopper? There should have been no surprises going into this endeavor from an operational perspective. The numbers are clear and the factors of production are known.

It’s not like they didn’t have years of case studies available from high priced consultants to gauge customer reaction (and every other metric for that matter) if that was the reason for pulling out. It’s simply irresponsible to make such far reaching initiatives and then bail on the entire effort before really getting past the starting line. Shareholders should be inflamed for such high-school like management practices.

Symphony Technology Group announced recently that it was acquiring Hummingbird for $465M.  This is significant in that another ECM player is being taken off the market though Wall Street thinks that the offer amount was too low.

Hummingbird has steadily been losing market position against the mainstays over the years.  Symphony might be just the right company to turn things around.  If the acquisition closes, I wouldn’t be surprised if Symphony increased Hummingbird engineering resources with their enterprise development talent in India.  This is one company that really understands enterprise engineering.

It’s going to become increasingly difficult to be a standalone ECM player with customers buying more products from less vendors — unless you have a killer SOA play.  Many companies have tried this one form or another over the years, but I think it was always an issue of timing.  The foundation of SOA in the Content Management market makes as much sense as any other solution (if not more).  Maybe Hummingbird is the company that reinvents itself with SOA?

Dan’l Lewin writes a must read article for AlwaysOn entitled, “Is India At The Tipping Point?” He writes about the rise of the Offshore Product Development company (OPD) building “whole products for their clients at about one-third the cost.” It’s not just about building products though — it’s about domain expertise. The labor arbitrage game is a no brainer bet — you always double-down when the dealer has a 6 showing and you have 10.

Dan’l also says that “I believe that OPDs are the breeding ground of future software entrepreneurs and will impact the local software ecosystem.” I agree that this will be a future trend because I am doing it right now along with others. There were brief moments when I thought about setting up my own Indian subsidiary, but why bother with what amounts to minimal cost savings when the basic hassles of running a business won’t contribute to my core competency: building disruptive software. Plus, the bureaucracy involved for a Non-Resident Indian (NRI) to setup and own an Indian business is a nightmare. Running an HR department and having to bribe people to get an internet connection sooner than later is for the birds

. At the end of the day, the fully burdened cost of running your own show in India is comparable to using an OPD. So why subject yourself to pitfalls that could be serious show stoppers? Mitigate risk early on and give your startup a chance….

I had a conversation with a friend of mine recently about the cost of goods in India.  Some things are still a bargain compared to the West.  I routinely have dinner for under $2.  Lately, I have been topping dinner off with a Falooda for an extra 60 cents.  A rather complete lunch is served for about 40 cents in the office canteen.  A ride in an auto (a motorized rickshaw, not an automobile) that would cost upwards of $5 in NYC costs about 50 cents here.  Custom tailored clothing can be less expensive than the “ready-made” alternatives.
One of the most noticeable increases here in India is in the technology field.  With 15% year over year salary increases, it’s becoming increasingly difficult for employers to retain valuable employees.  Even when the country is crankin’ out 400,000 technology workers from schools each year, the supply is simply not enough to keep up with the demand from the West.

It’s all very reminiscent of the late 90’s when technology workers in the States were jumping ship from the mainstays and heading to the .com companies in droves.  In India, workers usually have to remit 30 days notice to their current employer, but a prospective employer can acquire that notice period for some agreed sum.

The level of experience in India compared to the States is noticeably different.  Indians with vast amounts of experience are usually already in the States.  5 Years of development experience here and one is branded as a “Senior Engineer,” but it’s just the nature of the scale that exists.  A Sr. Engineer who is in search of a new job will likely have between 3-6 offers soon after being declared a free agent depending on various factors of experience and geography. Employers must be quick to make an offer knowing that the market is extremely competitive and driven largely on one thing we all understand:  personal wealth.

At the current rate that salaries are growing, within 5-7 years India will be very close, if not on par, to American software engineering salaries.  The value paradigm will shift from a model built primarily on economics of using offshore resources in developing nations to one that stresses the growing domain experience of Indian developers.  I met an Indian company recently that knows as much about the US Health Insurance Portability and Accountability Act (HIPAA) as any US healthcare software vendor that I have ever met.  But it’s not just related to US laws and regulations — they understand our industries and they’re starting early.

I walked up to my 9 year-old cousin who was sitting at the computer busily spouting off seemingly technical jargon in his quasi-English accent.   The 9 year-old was teaching the 7 year-old HTML — hand coded.  He wasn’t using any slick Integrated Development Environment (IDE) — he was using Notepad.  And he knew more HTML than the geek who was standing behind him with his jaw on the floor.

A VC friend of mine who focuses on the consumer space recently quipped that the word “repository” means “proprietary” to him. To a certain extent he’s correct in that the back-end data models of repositories are always unique by vendor, but a repository might be a necessary evil in the enterprise.  Without repositories, user-behavior can be left completely unchecked.  Yeah, I know —- one can setup security here and there, but that requires a human to do something and security is not all encompassing.  Let’s face it:  most users have complete disregard for corporate assets — they could care less.
User-behavior in regards to applications in the enterprise can be a double-edged sword.  Every developer of enterprise software applications is worried about user-experience.  An application that requires drastic alteration of the usual point-an-click semantics of a Microsoft Office application can be quickly relegated to the “it’s too hard to use” pile, rendering the application and the investment dead weight.

Without managing (and mandating) user-behavior in the enterprise with a repository, a process is all but completely shot.  Forget about requiring users to follow procedures when they depend solely on generic tools like Windows explorer, Adobe Acrobat, and Microsoft Word without repositories on the back-end to enforce business rules.

As an example, how do you force a user to perform version control for all modified Word documents so that you can see the delta between all the revisions?  Without a repository, you can only hope that a user follows unenforceable rules regarding file naming conventions when they save newly modified files in a directory tree somewhere.  Oh, and good luck forcing them to save things in the right directory.
Vendors who try to sell you on “we don’t alter user behavior” are really saying, “you can continue to let the users do whatever they want.”  Repositories are a necessary evil indeed.

I’m in Coimbatore, India for a few days after Bangalore to catch-up with some relatives before heading North. I was finally able to upload some photos after being broadband deprived for the past few days.

Pictured above is Pete Deemer, CPO of Yahoo, talking about agile project management at BarCampBangalore held at the Yahoo Bangalore offices on April 21st.

Pictured above is Chad Dickerson who also works for Yahoo talking about “Yahoo Hack Day.”

Pictured above is an old school wiki.

More photos from BarCamp here.

(Full disclosure: I worked for EMC/Documentum until recently.)

I have blogged about enterprise search before here, here and here. By the way, what is with the blogging trend where people link to themselves using the famililar “here, here, and here” moniker? And many bloggers seem to be “musing” about something or another when we previously just “thought” about something.

Google’s most recent foray into the enterprise will probably be on par with their other endeavors: they’re always late to the game, but then they usually dominate by half-time. Federating disparate databases and applications as part of the search paradigm for business users is at least 7-8 years old. The major Content Management vendors have been doing this for ages: IBM (acquired Venetica), Documentum (acquired Ask Once), FileNet, Autonomy, Interwoven, and the list goes on. They all index each other’s content in addition to most ERP and RDBMS applications. They all have some creative ways of presenting clustered results from a federated search.
What I find interesting is that enterprise search has crawled its way out of the traditional realm of the Content Management vendors and found itself squarely in the hands of new storage infrastructure players like Scentric, Kazeon, and Abrevity — though I can’t say that I completely agree with their approach.

Google will likely popularize the technology in general. It’s obvious that enterprises are all acquiring applications that have their own unique data models that results in a myriad of challenges from a usabilty perspective. Enterprise search is one way of tackling the problem that presents the user with a very familiar interface: a single search dialogue.

So the acquisition rumors some months ago were half true. Atlanta based JBoss is being acquired by Raleigh based RedHat for $350M with a $70M earnout. It makes a ton of sense for RedHat to own JBoss in my opinion. They’re both based in the South and both are fundamentally open-source initiatives, but more importantly they target a very similar enterprise customer.

In my opinion, it makes more sense for a vendor like JBoss, that primarily is known for their application server, to be owned by an OS infrastrastructure vendor like RedHat than by Oracle.  RedHat’s stock price today indicates that the analysts are bullish on the acquisition news as well.  The investors should be happy with the exit terms, considering they only invested $20M.

The most recent $8M from Sequoia puts them at $11.5M total.  YouTube recently announced that they were limiting the running minutes on videos uploaded to the site in order to reduce the amount of copywritten content.

If I was an IP attorney, I would be incredibly concerned with the legal ramifications of making money on the backs of copywritten content.

To be in your early 20’s and to be reportedly selling your company for $2B is simply awesome. If Facebook gets anywhere near that: congrats to the management team and the investors.

Here’s a link to a picture of him with his ethanol powered Chevy Tahoe.
Link:  Vinod Khosla’s Tricked Out Chevy Tahoe

Vinod is at it again. Besides his own VC firm, he’s also urging Californians to endorse clean energy. Turning agricultural waste into fuel makes good sense to me as does reducing the country’s dependency on foreign energy sources.

Mr Khosla is particularly enthused by “cellulosic” ethanol, a highly efficient way of making fuel from agricultural waste. President Bush touted this new technology in his recent state-of-the-union speech, suggesting that it may come to market in six years. In typically impatient form, Mr Khosla wants to halve that gestation period. Anyone who spends time with him is liable to be hit with his well-researched but mind-numbing PowerPoint presentation on ethanol—unveiled with the affection that some men reserve for pictures of their grandchildren.

Link: A healthier addiction

I can usually rely on SandHill.com for thought provoking business, technology, and venture capital articles, but a recent article by Guy Smith falls short of the usual.

While I admire his ability to be creative in regards to what might constitute an enterprise mashup, the reality is that the engineering costs alone might far exceed what a customer would pay for a finished product.

I have obsessed endlessly about enterprise mashups.  Customers love what can be done, but the fully burdened cost of amalgamating disparate interfaces is expensive.  And you can’t just expect to create an application that becomes the “new interface” or console.  That’s an even harder battle to fight.  Dashboards by themselves are not interesting enough to rush out and develop mashups.

That brings up the old question of value.  What value can your core application add that’s not being supplied through another vendor’s API?  If you’re application is doing little more than tying a couple of APIs together, then neither is it that interesting nor is the barrier to entry in whatever you’re doing all that great.

For enterprise mashups to work, they have to be multi-dimensional in my opinion.  In other words, you create middle-tiers that exposes new functionality from what’s gained in mashing multiple APIs together, create your own interface on the other end and then add value to other applications that normally wouldn’t have any knowledge whatsoever about the mashed entities.  (More on this later…maybe.)

Matt McCall writes via Tim Draper via Marc Hedlund:

These are not tech specific by the way.

link:  Entrepreneurial Proverbs