Open Source Commentary from Navica's CEO, Bernard Golden

October 2005 Newsletter

• Enterprise Software: Requiem for a Business Model

• Takeaways

Enterprise Software: Requiem for a Business Model

“We’re going into a phase of consolidation …. You have to figure out who gets bought and who just evaporates.” – Larry Ellison, June 2003

You have to hand it to Ellison, he lives up to his word. Having spent upwards of $15 billion on PeopleSoft, Siebel, and Retek, Oracle seems to be single-handedly trying to achieve his prediction. These purchases reflect Oracle's perspective that the software industry is entering a low-growth future due to slower and smaller buying patterns by corporate clients.

According to this this viewpoint, in a low-growth world the only way to achieve historical software margins is to spread fixed costs across a larger user base. If you can't get more users through selling your own products, you get them by acquiring other software companies. In pursuing this strategy, Oracle is following a well-trodden path behind many other industries: autos, steel, airlines, and so on.

Oracle's perception that the enterprise software industry has become low-growth and thereby ripe for consolidation is accurate; however, there is another force at work that will have even more impact: open source. Taken together, consolidation and open source will cause tremendous changes in the software industry and turn it into a shadow of its former self. One might even say that enterprise software is feeling poorly and the long-term prognosis is grim.

The State of Play in the Software Industry

There's no question that the phenomenon Oracle has observed is real. Growth rates in the software industry are way down. In the past, industry revenues typically grew at two to three times the underlying growth numbers of the economy. Of course, individual vendors could achieve results far higher than that. Many software companies saw 100%+ growth rates several years running.

The US economy has been growing pretty strongly over the past two or three years, on the order of 3% per year. However, industry revenues have not been growing at nearly the rate they would have in the past, given the general growth rate of the economy. Furthermore, what growth there has been in has not gone to the industry leaders -- the large enterprise software vendors have shown anemic growth over the past five years. Recently, Wall Street was cheered when one large software vendor managed to increase its quarterly license revenues by 1%!

The reasons advanced for poor license sales vary. One popular explanation is that users are still digesting the technology purchased in the late 90's: Internet, portal, ERP, and all the rest. After such a big meal, there's no appetite to swallow more software.

Another argument advanced is that, with the tech bust, IT organizations have been put on strict rations. They have little money to spend, so are only making tactical purchases to fill in small gaps in their IT infrastructure.

The boldest analysts feel that there is a reluctance to make the large investments in software typical of the past. In this view, IT organizations are making small initial purchases to prove technology rather than blindly buying large amounts before really knowing if the product will work in their environments. Having watched their predecessors get fired for multi-million dollar procurement mistakes, today's crop of CIOs make cautious, incremental software purchases.

Finally, some people say that software companies themselves have changed their strategy and are now more focused on harvesting maintenance dollars rather than making large license sales. Large sales are great, but they lead to lumpy quarterly results, so the subscription-like revenue streams of maintenance and support are more desirable.

All of these reasons are valid, although the last one is rewriting history to support a convenient rationale. Vendors that can't find top-line growth proclaim maintenance revenues their true aim. Anyone who has ever worked in the software industry knows that high-margin license sales are what make the industry magic and have always been the fervent aspiration of every self-respecting software company.

Taken together, these reasons account for a low-growth enterprise software industry; in a low-growth world like a consolidation strategy makes eminent sense.

However, an analysis of the enterprise software industry that is limited to considering low-growth scenarios fail to grasp the trend that will have the biggest impact of all – open source. We have only just begun to see the terrible effect open source will have on the commercial software industry.

Open Source Impact on Vendor Revenue Streams

Many discussions of open source treat it as if it is a unique beast, walled off and separate from the rest of the software industry. This perspective views open source and commercial software as two separate fiefdoms with little interaction and little impact upon one another. This ignores a central reality: every installation of an open source product represents the loss – or at least the potential loss – of a commercial product sale.

Far from having little impact upon the commercial software industry, open source is today affecting the industry significantly – but this is just a prelude to the dramatic impact it will exert in the future.

It's already happening – but it's anonymous

The first thing is to understand about open source use is that it's already widespread, but gets little publicity. In this sense, open source is a silent – but deadly – software revenue killer. Every open source installation is a foregone commercial installation, and indicates lost vendor revenues.

How much of this is going on? It's not easy to tell. If, in a purely proprietary software world, one product began displacing numerous competitive products, word would shortly get around. The company would brag about its success. The press would publicize the products in stories. Analyst firms would migrate the company toward the “magic quadrant.”

When the chosen product is open source, however, usually no one outside of the user organization is aware of the selection. The situation is, so to speak, akin to a tree falling in a deserted forest -- no one hears. This reflects the fact that almost all open source is downloaded and installed anonymously, which was commented upon in an earlier newsletter.

It's impacting commercial vendor margins

The second thing to understand about open source's impact on commercial software revenue is that it doesn't reflect a lack of software demand; rather, it reflects a growing preference for free software in place of commercial software. In other words, it's not really a low-growth world, it's just that much of the demand growth is being satisfied by open source use.

Today, company after company talks about how they're swapping in open source products in place of commercial ones; how they're architecting new systems to take advantage of free software. They haven't stopped deploying systems -- they're just building them a different way. And these aren't flaky companies, either. They're companies like ABB, Sabre Holdings, Charles Schwab.

Certainly open source does not account for all “lost” software revenue, but at the margin it is clearly having an impact. And margin in software is like money in politics: all-important.

Software is the highest-margin product ever invented. Typically, software products achieve 95% gross margins. No wonder software companies love license sales! Less often remarked upon is the flip side of low variable costs: software company cost structures almost seem like they're made up of completely fixed costs, primarily headcount. License sale revenues during any given quarter first have to cover the fixed costs of the company; after those costs get covered, nearly every dollar of license revenue turns into profit. Every sale beyond the tipping point drops straight to the bottom line, which means that sales at the margin are precious, since they provide the luscious net profitability that investors love so much.

So it's easy to see that open source use can have a significant impact on software vendor profitability. Every open source implementation shaves nearly the entire license fee of the commercial product from the net profitability of the vendor.

It's Worse than Margins: It's Prices, Too

Even more pernicious than the displaced profit that open source causes is the pricing pressure it presents. Everyone who has watched Wal-Mart's relentless "Always Low Prices. Always" tramp through the retail industry understands this. When a competitor appears with consistently lower prices, you can either reduce your prices, or focus on presenting a better value proposition.

What do you do when your competition is free?

Many commercial vendors are still attempting to maintain a price umbrella, proclaiming that open source software is amateur-developed, unsupported, and insecure. By contrast, they say, their commercial product is employee-developed, has a commercial institution associated with it, and is highly secure.

Many of them claim that open source poses no threat to their business – that no real customers are using open source. Oracle's recent purchase of InnoDB, a company that provides a key piece of technology to MySQL, gives the lie to that claim. It's hard to read the motive for the purchase as anything other than an attempt to cripple MySQL.

Leaving aside the validity of all these claims, the fact remains that increasing numbers of organizations appear to be willing to adopt open source software. Clayton Christensen, in his several books on innovation, shares many examples of industries that were crushed by competitors offering good-enough products at prices 20% below the incumbents. Software incumbents appear to be gambling that customers will assess switching costs as outweighing license savings, and therefore will stick with the incumbents. They're probably right – in the short term.

In the long term, organizations will move to open source due to its price advantage. How will commercial vendors react then? They will have only one option – drop prices.

During the antitrust hearings relating to the Oracle takeover of PeopleSoft, there was some very juicy testimony relating to pricing deals offered by sales representatives of the vendors. Discounts up to 90% were proffered to customers. So, it's easy to see that software pricing can be very flexible, given the right circumstances. After all, with such high margins, there's lots of room to reduce prices.

It's inescapable that the long-term impact of open source will be to reduce prices and thereby reduce software company profit margins. Evidence of this can be seen in the recent quote by SAP's sales head, Leo Apotheker: “There is significant price pressure out there in the market ... I don't expect pricing pressure to go away in the near future. It might actually never go away because of the market dynamics.”

How will software vendors respond to this relentless price pressure?

The Software Company of the Future

Going forward, software companies will have to be much more cost-conscious than ever before; the lavish cost structures, enormous sales rep packages, and the seemingly endless “initiatives” (read: aimless new product introductions) will all be consigned to history.

In short, the software industry is not going to be nearly as much fun in the future as it has been in the past.

In the coming years we will see the software sector behave like other sectors that have experienced fearsome competition. Sectors like the steel, auto, and airline industries. It is going to be a wrenching time, with constant cost-cutting via layoffs, benefit trimming, and perk reduction. Make no mistake about it, there will be a large human toll.

This relentless financial pressure will also play itself out in the way software companies do business.

Today, software companies provide a bundle of benefits to customers. The most obvious benefit is the software itself, of course. Maintenance and upgrades are another benefit.

Less often noted are the other services software companies include with their product. For example, software companies provide technical services like architecture reviews, roadmap presentations, seminars, and whitepapers at no charge. One might even include the investments software vendors make in briefing technical analysts as a bundled service, as customers can take advantage of analyst knowledge that, at least in part, is created by software vendors.

It's easy to underestimate the value these “soft” product elements provide. As they're thrown in with the license purchase, a typical customer will take them as a given and fail to understand how useful they are.

It's going to be much clearer in the future, however, because margin pressure will force companies to stop offering these things as giveaways. They will become a separate offering, available at a set fee. A customer request for a roadmap presentation will be met with a request for a PO number.

In other words, software will move from being a bundled product to an unbundled product, with each product element available and priced separately. This will enable software companies to set prices for their base products much nearer the free competition presented by open source.

Takeaways

It won't happen next year, or the year after. Slowly but surely, however, open source will impact the software industry dramatically. The first casualties will be the weaker second- and third-tier vendors. Ultimately, every vendor will experience the menace of open source. It's inevitable. It happens to every industry: high-margin early years followed by lower-margin maturity as the initiative loses its novelty and gets assimilated into the quotidian business fabric of society.

Software companies will begin to resemble other companies in the economy and face similar challenges. This doesn't mean that there is no opportunity for software-based businesses – far from it. It does mean that the license-based software industry will look far different in the future than it has in the past.

The phrase that is useful is “opportunity – and margin – migrate to the edge.” Future software opportunities lie in new vertical applications, new devices, new frontiers – all located at the edge of the software stack, the organization, and the culture. If, as a vendor -- or a vendor employee -- you don't like your prospects, move to the edge. Don't bother staying where you are – you won't like the view from there.

Next month: Free software is great, right? Then why has my job gotten more complicated? What an open source world means to IT users.