Tuesday, September 30, 2008

Larry Whistles Past the (Cloud) Graveyard

Larry Ellison recently unleashed a tub-thumping tirade against cloud computing covered by Ben Worthen (with further comments from Daya Baran, Giva Perry and Dan Farber) . Here is a quote from Larry:
The interesting thing about cloud computing is that we've redefined cloud computing to include everything that we already do. I can't think of anything that isn't cloud computing... The computer industry is the only industry that is more fashion-driven than women's fashion.
Now as usual with big whoppers told by people in fear of their checkbooks, Larry's rant has an element of truth. There is much cloud lipstick being sloshed on many a barnyard animal these days. But behind all the buzzwords, a basic shift is occuring - the wholesale outsourcing of core business applications.

Yes, this is a return to timesharing, but this ain't your grandfather's timesharing. It is interesting that Larry was a huge supporter of thin client computing when it threatened his enemy Microsoft but is now a detractor of the son of thin client when it threatens his own rice bowl.

Forrester research has a simple chart that illustrates Larry's dilemma - he paid $30B for a collection of low growth businesses, making Oracle is the Computer Associates of the new millennium (and we know where this strategy got CA).

Larry's rant is an extraordinary example of whistling past the graveyard. Oracle's huge transformation over the last 10 years has been from an infrastructure company (databases & middleware) to an applications company (ERP, CRM, SFA ect). Now, just as this transformation is completed, along comes an infrastructure that will obsolete all the applications Oracle just got done rolling up.

No wonder he sounds ticked - you would be too if you just spent more than $30B (see Oracle acquisitions chart below) on a bunch of wasting assets that are going to be shafted by the Cloud/SaaS shift just as Siebel's market share was eviscerated by SalesForce.


The dead giveaway is that Larry's rant focuses on the metric most important to him - profitability - while ignoring the metric most important to customers - value. Larry knows that customers are deserting him in droves, what really angers him is that he can't make as much money in the SaaS world.

Friday, September 26, 2008

Why Open Source Is Euro-Chic

Larry Augustin recently wrote about the differences between how Europe and the US view the open source software market. His comments came after attending the Olliance Think Tank conference in Paris this week (tough assignment, that).

He identified a number of differences between how Europe and the US view open source. For example, he gives the primary European reason for adoption open source as wanting to avoid vendor lock-in, while the primary reason for adopting open source in the US is cost.

While recognizing the differences, I think that the open source business model is tending to converge on the dual-license approach that I outlined in the Silverado Rules for Open Source Software. So what accounts for these regional differences?

I believe the biggest single factor to explain these differences is the underlying attitude towards Microsoft. In the US, where Microsoft is seen as a successful, if aggressive, competitor, adoption open source is a business decision, that is to say based on cost and ROI.

In Europe, where Microsoft is seen as a malignant force, adoption of open source is seen as a political decision, that is to say based on and ideology that trumps business considerations. Viewed this way, the differences described by Larry between the two geographies are very consistent, particularly if you just insert the word Microsoft at strategic points.

For example, on the first observation, the primary reason for adopting open source in Europe is to avoid [Microsoft] lock-in; while the primary reason in the US is cost, as Microsoft lock-in per se is not seen as a bad thing.

At the end of the day, open source is not so much an ideology as an evolution in how software is developed and distributed. For WaveMaker, open source is a great way to accelerate adoption of our free JavaScript download for building web applications.

Wednesday, September 17, 2008

The Revenge of Zaphod BeebleBrox - Why 2 In a Box CEOs Never Work

Matt Asay broke the news that the talented COO of SpringSource, Neelan Choksi, left the company recently just as SpringSource announced 250% annual growth. Savio Rodriguez expanded on the theme and wondered why an exec would leave a growing company.

I have no particular insight into Neelan's situation, but I have a lot of insight into the difficulties of managing a startup with a CEO and a COO (at SpringSource, Rod Johnson is the CEO and Neelan was the COO). I have found that splitting the CEO/COO in a small company, also known as "two in a box" is pure trouble.

Two in a box means putting two people in charge of essentially the same job. You don't need an MBA to see that there is great potential for mischief here. The inevitable conclusion of two in a box is always the same … and then there was one.

The whole thing reminds me of Zaphod BeebleBrox, the two headed character who was always arguing with himself.

Here's how the Zaphod BeebleBros brainstorm pitched by the board: "this will be great! Zaphod here (the CEO) will be the external spokesperson for the company, while Beeblebrox (the COO) will be the internal, get it done guy."

Needless to say, it never works quite this way. Even when you have two really strong players (which I believe is the case at SpringSource), it is extremely difficult in a small company to keep from stepping on each other's toes. Here are just a few of challenges in a two-in-the box startup team:
  1. Who's vision? Inevitably, there are differences in vision between the CEO and COO and these differences always come out at awkward times.
  2. Maybe I'll get a better answer from mom. Any exec who gets an answer from datd they don't like can always try asking mom.
  3. The buck stops over there. Particularly for difficult decisions, there is a natural tendency for each exec to palm off some of the tougher calls to their counterpart.

Wednesday, September 10, 2008

Where have all the 4GLs Gone?

Brad Feld recently wrote a blog post entitled, What Happened To The 4GL? In it, he describes the difficulty of working with today's web development frameworks:
This summer I spent some time playing around with Google AppEngine.... It didn't take long before I realized I needed to really understand how to program in Python to do anything.
People often ask whether Ruby on Rails is a competitor to WaveMaker. My response is that WaveMaker is for people who don't want to learn a new language to build web apps, RoR is for people who do.

In Proustian fashion, Brad looks at the new cloud computing tools in hopes of recapturing that sense of power that comes from working with a well-designed 4GL:
When I was playing with Google AppEngine, I kept waiting for the 4GL "aha moment." That's the moment I had using Clarion and Access where I realized how easy it was to do certain things. That moment never came with Google AppEngine - the deeper I got, the more confused I got.
This is the elephant under the table in web programming today - it is no longer possible for mere mortals to build basic business applications. Even formerly technical guys like myself and Brad are intimidated by the bookshelf worth of O'Reilly books you need to read just to get started with web development.

WaveMaker has a very strong 4GL pedigree and is funded by Mitchell Kertzman (of PowerBuilder fame) and Roger Sippl (of Informix 4GL fame). Our belief is that the time has come for a 4GL web development solution.

Ten years ago, there were loads of ways to build client/server apps easily - PowerBuilder, MS Access, Lotus Notes, Filemaker. None of those tools have made the leap to the web, leaving a huge market vacuum.

It has been so long since there have been decent tools for non-expert developers that people have literally forgotten what a 4GL even looks like. The best proof of this is the many comments on Brad's post by people who think that coding frameworks like Ruby on Rails and Django are reasonable substitutes for a 4GL.

This is not to slight Rails and Django, just to say that these products are targeting making hard-core developers even more productive. They are absolutely not appropriate for visual developers who don't want to do any programming in the first place.

Wednesday, September 03, 2008

Startup reality check: launching versus landing

While startups always make a big deal about introducing new products, the real birth of a company comes not when you launch your first product but when you land your first big customer.

Here's the difference: it only takes a demo to launch a product, it takes a business model to land a customer. Thus the scariest time in a startup's lifecycle is the period between shipping the product and closing the first big customer deal (this is period that the VCs call "proving that the dogs will eat the dog food").

In true ClueTrain Manifesto fashion, product launches are largely a vendor driven activity, that is to say mostly wishful thinking and hand waving about your new, improved, bright shiny thing. Getting a customer to dig deep and pony up 7 figures for your bright shiny thing puts you in a whole different league.

Launching answers the following questions:
  1. Can we give a good 5 minute demo? Silicon Valley is littered with technology in search of a market.
  2. Can we sound like we know who might want to buy this in the abstract?
Landing a big customer answers much more interesting questions:
  1. Is my product solid enough for somebody else to bet their job on?
  2. Does my product solve such a big problem for that person that they are willing to bypass much more established, safer vendors to bet on an unknown startup?
  3. Is my business model credible enough that another company will bet their business on our being around 5 years from now?
WaveMaker launched our visual Ajax development tool in February. Yesterday, WaveMaker announced a 7 figure deal with KANA. KANA sells customer service software and has over 700 customers, including 60% of the Fortune 100. They are also one of 16 strategic partners of IBM's WebSphere division.

The most exciting element of this deal is that KANA will ship the WaveMaker Studio as a built-in customization tool for their call center platform (similar to what the Force.com platform does for SalesForce.com). For WaveMaker, this will give us 700 licensed customers by the end of the year, putting WaveMaker in the top tier of Ajax tool providers in one fell swoop.

Cool technology and $2.50 will buy you a latte at Starbucks. Having a $70M company bet their future on your product puts you into the running for the Next Big Thing! Congratulations to the whole WaveMaker team for taking the step from demoware to solving hard customer problems.