Marketing Expert's Corner

This article written in 2006


We in the IT industry habitually look at technology first, believing that a technical edge is more important than any other advantage a business could have. This isn't the way most industries work, but our industry has been hypnotized by the bright shiny objects of techno.  We make most of our decisions as Supply Siders, assuming that if you improve the product, the demand will come.  We believe the best product wins.

You don't need to read The Innovator's Dilemma to know this is short-sighted.  The hard reality is that over the long run, the best technology loses out to commercial advantages like market power, financial strength, and serious marketing.  In the end, it is almost axiomatic that the best technology loses, even when it's from a big company.  Think IBM.  Think DEC.

So when we in high tech look at the phenomenon of Open Source, we're looking at it the wrong way.  We're fascinated with its supply-side attributes...the things that aren't likely to matter much over the long term.  So, let's look at what open source does to sales and marketing -- the processes that will be the long-run determinant of success.

Open Source, inside-out

Listening to Open Source advocates (I count myself as a card-carrying member of that group), you hear about these kind of advantages:

  • faster, more open innovation,
  • lower cost of development
  • user-driven feature mix (avoiding the confusion of too many marginal features),
  • broader testing,
  • quicker time to market, and
  • dramatic reductions in cost of ownership.

While there are great examples of each of these advantages, it is rare for an open source project to achieve all of them.  And most of the advantages are stated from a supplier's point of view.  But what if Open Source's engineering aspects aren't the pivotal ones?  What if they are supporting arguments, and the real story is about business model? 

I have a confession to make:  I've been holding out on you.  This article was drafted over a year ago, but I didn't want to upset people by publishing it too early.  But John Roberts, CEO of SugarCRM, has been eloquently making the case about the open source business model, and it seems like the industry is ready to hear this -- look at this month's LinuxWorld keynote address:  

Death of the Enterprise Software Business Model

According to recent research by software analysts at Goldman-Sachs, Enterprise software companies spent 82% of new license revenues on sales & marketing in 2005, up from 66% in 2000. Effectively, traditional Enterprise software companies are charging their customers to sell to them.

The specific data in the paragraph above are very debatable, but the underlying issue is not.  The costs of the Enterprise Sales model can eat you alive unless you get hypergrowth or market dominance.

The Enterprise Software business model is characterized by large customers, high prices, direct channels, high commissions, and relatively low volumes.  The Enterprise sales cycle is high-touch, its salesman using a consultative style.  Thousands of software companies have followed this model, and most of the Software 500 continue to do so.  For software that is specialized and relatively low volume, there really hasn't been a great alternative. (Sure, consumer software can use channels, OEM bundling, and the web...but you need high volumes for that to work out).

But if you narrow the field a little bit,  looking only at consistently profitable software companies, the Enterprise model seems to be working only if you're an 800-pound gorilla.  IBM, SAP, CA, and Oracle do just fine with their software packages.  Cisco does fine with its network management software.   But there are hundreds of other Software 500 companies who have spotty profitability because of the cost of customer acquisition.  One delayed deal can blow their quarter, and that trashes their stock for months.

So if you have a new piece of software that maybe a thousand customers will be willing to pay for, how do you get it to market without losing your shirt?  

The Open Source Alternative

Instead of hiring an expensive sales force to do the market education, control all the conversations, and tap into the customer's buying process, you give away fully functional, somewhat-usable software to anyone who wants to try it.  You give complete access to your source files, so engineers in your target customers can see the inner workings and understand how they could add to it themselves.  The user convinces themselves they want to buy the "real version" before they even contact you. 

Essentially, you dispense with the entire cycle of lead generation, market education, prospect qualification, proof of concept, and chase-the-purchase-order.  You replace the Sales and Marketing engine with a dynamite web site, community development, and ecommerce system.  Early on, this means a marketing and "sales" staff of 3 world wide.  You add telesales reps only when the existing ones are too busy to take any more orders.  You don't hire a direct rep until you already have hundreds of (low-price) customers who can be up-sold.  

The cost structure scales with revenue.  More important, you cut in half the costs of the single most expensive part of running a software business -- Sales and Marketing can be done for 25% of revenues.  Because you aren't spending as much on Sales, your cash requirements will be much lower.  VCs will like the idea of a slow burn rate with the global reach of the internet (this is what they were banking on for all their investments during the bubble), but you will need to manage their expectations about revenue.

At the same time, you automatically expand the market.  Since you have no sales force anyway, there's no reason not to go after SMB or even micro-businesses via the web.  The goal at all times is to increase the size and dynamism of your community.  In fact, your community is more valuable than the code.  As Bernard Golden recently wrote, "an open source project without a community is just shareware."  Since your community is in cyberspace, countries you could never afford to do business in can still host your customers.  Large open source communities let you tap into and harvest long-tail markets.

Watch out though:  these open source community dynamics only really work well when you own your category, when there are no "competing" open source projects.  Avoid this kind of bifurcation at all costs.

Psychologically, open source feels really different.  You're giving away the crown jewels.  The vast majority of your users will never pay you a dime -- in fact, you'll have almost no idea who your users are.  But you'll get a hundred times more users than you'd ever have had with the Enterprise model.  The perceived free price can put you Inside the Tornado of hypergrowth much more readily than was possible before.  Once you have a bazillion users, you may just be able to harvest your community

Open source business models depend on the law of large numbers.  The numbers you really care about are numbers of downloads, numbers of community registrations, and numbers of posts per month in your online forum.  For the flywheel effect to kick in, you'll usually need more than 100,000 downloads.  Conversion rates in open source aren't necessarily any better than in closed source, so you may need 10,000 downloads (vice 10,000 leads) to do 10 deals.  (JBoss was the miracle project when it came to conversion rates -- 3% or more -- which was why it was such a tempting acquisition target).  Further, the deals must be smaller to drive the high volumes -- much closer to $1000 than $100,000 per purchase.

So watch out:  if your software really isn't interesting to more than a small group, an open source strategy will fizzle.  Your techno will simply be another anonymous project among the 130,000 ones hosted at SourceForge.  For highly specialized software, sticking with proprietary tactics is the only way to go. 

Phone Us +1 650 326 2626