Monday, December 31, 2007

Start with Relational Databases

I'll start with relational database management systems (RDBMS) as the first part of the software industry to analyze. Here are several reasons why I've started with RDBMS:

  1. RDBMS is a large market in its own right, producing $15-16 billion worldwide revenues in 2006 according to Gartner Group and IDC.

  2. Relational databases form the foundation layer of enterprise software applications. An understanding of the RDBMS market will help in understanding the enterprise software market, and in particular to evaluate the Oracle and Microsoft growth strategies in enterprise software. Furthermore, the continuing demand for enterprise software should support continued growth in the RDBMS market.

  3. Three players (Oracle, Microsoft, IBM) have demonstrated what appear to be "winning" strategies; together, the three have consistently dominated the market over the past decade. Their successes may offer insights for other parts of the software industry.

  4. The industry leaders now face a robust and increasing threat from open source, most notably MySQL. This struggle may offer a preview into the challenges that other parts of the software industry will face due to the incursions of open source.

  5. The underlying technology itself is maturing, but is by no means mature. RDBMS vendors continue to upgrade their products and release new versions, expanding the scope, scale and types of data their systems can manage.

Monday, December 10, 2007

Vivendi Acquires Activision

Last week, Vivendi announced the acquisition of a majority share in Activision. If all goes according to plan, Vivendi's online gaming and entertainment business (most notably Blizzard Entertainment) will be merged with Activision's business. Vivendi will end up with a majority stake in the combined company, which will continue to trade under the name Activision.

The press release spells out a strategic rationale for the transaction (emphasis added):

Said Robert Kotick, Activision's Chairman and Chief Executive Officer: “This is an outstanding transaction for Activision and our stockholders, as well as a pivotal event in the continuing transformation of the interactive entertainment industry. By combining leaders in mass-market entertainment and subscription-based online games, Activision Blizzard will be the only publisher with leading market positions across all categories of the rapidly growing interactive entertainment software industry and reach the broadest possible audiences. By joining forces with Vivendi Games, we will become the immediate leader in the highly profitable online games business and gain a large footprint in the rapidly growing Asian markets, including China and Korea, while maintaining our leading operating performance across North America and Europe. Activision stockholders will benefit from significantly increased earnings power and the recurring nature and predictability of subscription based revenues, while also having the opportunity, if they choose, to receive $27.50 per share for a portion of their shares in the post-closing tender offer.”

How compelling is this rationale?

Activision asserts that the merger will give the combined company "leading market positions across all categories of the rapidly growing interactive entertainment software industry." This rationale implicitly assumes that leadership in each category is a desirable position for the company. This assumption sounds plausible, given the high fixed cost and low variable cost structure of the industry, but would be worth verifying.

Assuming that category leadership is a desirable outcome, the economic value the merger creates, as well as the level of effort required for post-merger integration, will depend in part upon whether the leading market positions would arise from economies of scale or economies of scope. The merger could generate economies of scale if Activision and Vivendi had overlapping customer bases, redundant cost structures, or underutilized assets that could be rationalized as a result of the combination (a typical rationale for, say, merging two banks with overlapping branch networks). On the other hand, the merger could generate economies of scope if Activision and Vivendi had complementary businesses that could add value to each other (a typical rationale, for, say, merging a bank with an insurance company).

Scale-based mergers generally require far more effort than scope-based mergers, but have a much greater likelihood of creating value. For example, in the retail banking industry, scale-based mergers have gradually but painstakingly created value during the past decade of consolidation, whereas scope-based mergers in financial services have proven far less durable.

In the Vivendi-Activision deal, I'd suspect most of the anticipated synergies are due to economies of scope, as the press release already implies that each company is already a leader in its respective categories.

The press release only mentions one specific economy of scope, the Asian market access that Vivendi would provide to Activision. How much value does this synergy provide? If this were the only synergy, would a transaction with Vivendi be the cheapest and best way to get better access to China and Korea? Have Activision and Vivendi already developed a plan to capture the value of this opportunity?

Do the companies have any other synergies in mind? Does Vivendi intend to extend the Activision game franchises to more fully take advantage of the subscription-based business model? Does Activision have unique product development skills or distribution channels that could otherwise increase the value of Vivendi's Blizzard Entertainment franchises? Are there other cross-sales opportunities?

I'll be curious to see whether this merger was motivated by strategic considerations, and if so, to what extent the synergies are realized, and to what extent they cover the premium that Vivendi is paying for Activision. Or whether this merger is motivated more by financial considerations (such as finding a more optimal capital structure for Vivendi's game business).

Hat tip: WoW Insider

Wednesday, December 5, 2007

Market Map - First Draft

A market map shows an overall picture of the competitors in a market. It depicts the relative sizes of major submarkets, the extent to which each submarket is fragmented or consolidated, and the market shares of the key competitors within each segment.

The following market map shows the distribution of revenues of the 50 largest public software companies. This type of chart, called a marimekko, has the form of a rectangle whose total area represents the $150 billion of combined revenues of the top 50 public software companies. Each column of the chart represents a different submarket. The area of each smaller rectangle represents the revenue that a particular company earns from a particular submarket.

A few additional comments:
  • This chart does not include the revenues from smaller software companies, nor from privately held companies, nor from other companies (such as HP, IBM, EDS, Accenture) who have significant software revenues but are classified in other categories by Google Finance. I expect to update the market map to include these revenues one submarket at a time after examining each submarket in more detail.

  • This chart shows the latest full fiscal year published for each company as of December 2007, and so the list of companies is already a bit obsolete. Several of the companies on this chart have already been acquired, or are in the process of being acquired, by other competitors.

  • Because this chart shows revenues, it will not display the unit market shares of open source / freeware competitors.

  • All the data in the above chart was sourced from Google Finance and/or annual filings with the SEC (Form 10-K for US companies; Form 20-F for foreign companies with ADRs). Within the 10-K and 20-F forms, the relevant information is usually found in a footnote with a title like "Business Segment Reporting".

Wednesday, November 21, 2007

Business Mix - Microsoft

In order to classify the larger software companies into the industry categories, it will help to quantify the mix of businesses that each participates in. Whereas the smaller companies will more likely be considered "pure plays" (falling into just one category), many of the larger ones will fall into more than one category.

As a rather obvious example, Microsoft has products and services that could be classified as Enterprise Applications (e.g., Microsoft Dynamics), as Infrastructure Software (e.g., Windows Server, SQL Server), and as Games (Xbox 360 and PC software). Microsoft also introduces another major software category, Desktop Applications, in which they are effectively the sole participant with their Office suite. And Microsoft has several non-software businesses (e.g., Xbox consoles, advertising, MSN) -- although it is far from clear that these non-software businesses make any positive contribution to Microsoft's market cap!

Pages 23-28 of Microsoft's FY 2007 10-K provide much of the data needed to infer Microsoft's revenues and earnings at a fairly detailed level. I won't show all the analysis here, but I should note that I did make four assumptions: (1) that Xbox consoles wholesaled at $350 each during FY07; (2) that the Server and Tools business split 45%-45%-10% SQL Server-Windows Server-Other; (3) that Microsoft Office comprised 92% of the Business Division's revenues; (4) that each division's operating margin could be applied uniformly to all the businesses within that division.

It's interesting to note that Microsoft's "core businesses" -- Windows and Office -- are two businesses in which it makes virtually all its profit, and (unsurprisingly) the two businesses in which it has nearly dominant market share. Both of these businesses face limited growth and potential eroding share, which is no doubt part of the reason that Microsoft has attempted to diversify beyond this core.

But other than the relational database business, where Microsoft appears to earn a decent profit and where it has a respectable market share, none of the other businesses appear to be making any significant contribution to Microsoft's profitability. Microsoft has a disadvantaged position against Google in the advertising business, and falls well behind Oracle and SAP in the enterprise applications arena.

Already this picture raises some questions about Microsoft's business portfolio. Is Microsoft taking the best steps possible to defend its core business and maximize its profit potential? Does Microsoft ever have a chance of achieving significant profitability in its non-core businesses, and what have to be true in order to believe that they could? Are there more sensible business adjacencies that Microsoft should pursue instead of these businesses?

Monday, November 19, 2007

Adjusting the Categories

A few additional thoughts on how the categories are organized:
  • It may make sense to analyze the enterprise software vendors based on the business processes they cover. These processes could include product design, inbound logistics, manufacturing, outbound logistics, sales and marketing, procurement, HR, planning, finance and accounting. Smaller enterprise software vendors might offer a product line that covers only one or a few processes, whereas the large vendors (SAP and Oracle) would offer a far more complete suite.

  • If we do analyze the enterprise software vendors based on business processes, then it might make sense to classify the engineering and design software companies within this framework, perhaps as Product Lifecycle Management applications supporting the product design process.

  • The two largest "Other" companies in the previous post are VMware and Adobe. VMware would almost certainly belong in the Infrastructure Software category. Adobe is a bit harder to classify; perhaps it will make more sense to think of Adobe as a portfolio of businesses that spans several categories.

Market Cap of Each Category

The following chart compares the market capitalizations of the companies in each category. (I've renamed several of the categories but for now have kept the same classifications).

Microsoft aside, the enterprise software vendors (led by Oracle and SAP) form the largest category. The second largest category consists of the "infrastructure software" companies. Video games, the only consumer segment that appears here, form the third largest category. The engineering and design category is even smaller.

Also interesting to note, aside from video games and some portion of Microsoft's product line, there is almost nothing here that would count as a consumer software segment. Most of the big companies who write software that consumers use (Google and Apple and AOL, for example) essentially give the software away for free and make their money elsewhere (advertising, content, services).

Sunday, November 18, 2007

Preliminary Classification of the Top 50 Software Companies

Business Definition, Customer Sharing, Competitor Sharing

This post outlines a preliminary classification of the 50 largest software companies. This time I'm using Google Finance's Software & Programming category as the data source. I'm again using market capitalization as the measure of size. I've excluded six companies (VMW, GA, SAI, STV, OMTR, SLH) which do not have any Related Companies listed by Google.

The Related Companies lists on Google Finance help to describe a sort of clustering of the companies. In the chart below, a solid line between two companies indicates that each is on the other's Related Companies list. There are also a few Top 50 companies which did not appear on the lists of any other companies; in these cases if there was a one-directional relationship then that relationship is indicated on this chart with a dotted line.

I have tentatively identified four clusters that seem to emerge out of these relationships. Each cluster is highlighted in a different color in the chart, and can be identified with a different type of customer group.

The first group consists of companies who sell primarily to video gamers. These are indicated in light purple.

The second group consists of companies who sell primarily to general business managers. These are dominated by the ERP, CRM, and Business Intelligence vendors. These are indicated in green

The third group consists of companies who sell primarily to engineering and design professionals. These are indicated in yellow

The fourth group consists of companies who sell primarily to IT and E-Commerce professionals. These are indicated in blue. This is a rather diverse group, including such companies as security software makers, middleware vendors, and Indian IT development shops, and will almost certainly need to be subdivided further

The categorization is not comprehensive. Several companies did not clearly fall into any of these categories; these companies are indicated in white on the chart. Additionally, MSFT seems to participate in at least three of the four categories

My next steps will be (1) to compare the sizes of these categories and (2) to select one of the categories for more detailed industry analysis. After analyzing the first industry, I expect to revisit the overall classification scheme

Thursday, November 15, 2007

Organizing the Project

To help organize the work, I have drafted a high-level framework that shows the steps I expect to take while analyzing each part of the overall software industry:

These steps may be repeated for each industry, company, and investment hypothesis under consideration.

After repeating these steps enough times, some patterns may emerge that could then help test and/or revise the original hypothesis.

Wednesday, November 14, 2007

Identifying the Major Software Companies

Who are the largest publicly traded software companies?

I searched the Technology section of Yahoo! Finance and selected the twenty largest companies, by market capitalization, across four categories: Application Software, Multimedia & Graphics Software, Security Software & Services, and Technical & System Software.

The results are as follows:

(The actual market caps for MSFT, ORCL and SAP are $312, $100, and $60 billion respectively. This is my first try at using Google Docs for charting, and it is not yet clear to me how to create the visual cues on the chart itself to indicate that these three companies have values that exceed $40 billion.)

This approach does miss a few companies that I would probably consider in scope for this blog. These companies are buried in other categories. For example, Symantec is classified in the Internet Software & Services category, which mostly contains companies whose primary source of revenue is from services rather than software sales. As another example, NAVTEQ appears in the Business Software & Services category, which is dominated by companies such as ADP whose primary source of revenue is from services rather than software sales.

Tuesday, November 13, 2007

Launching the Software Strategy Blog

This blog will explore the following hypothesis:
As the software industry matures, software companies will have to employ more "old economy" strategies and fewer "new economy" strategies in order to compete successfully.
Let me explain how I am using several of these terms today. Please bear in mind that one of this blog's goals will be to clarify these definitions!

  • Software industry: At least initially, I would like to concentrate on companies whose primary business is developing and selling software. This would include such names as Adobe, Autodesk, Intuit, Microsoft, Oracle, SAP and Symantec. This definition excludes a number of well-known companies which are renowned for their software development capabilities, but who make their money elsewhere. For example, this definition excludes Google (most of whose revenue is from advertising), D. E. Shaw (most of whose profits are from trading using proprietary software), EDS (most of whose revenues are from services), or indeed most of the "Web 2.0" companies. It remains to be seen whether this is a meaningful distinction.

  • "Old economy" strategies: Here, I refer to the business strategies that companies in the manufacturing and service sectors have used to compete successfully over the past four decades. These would include such concepts and disciplines as economies of scale, barriers to entry, process re-engineering, six-sigma quality, etc. These strategies are the bread-and-butter of most traditional strategy consulting firms and most business school curriculums. You tend to read about these sorts of strategies in the Harvard Business Review.

  • "New economy" strategies: Here, I refer to the business strategies that have helped many of the relative newcomers in the technology industry succeed. These strategies would include such concepts as first-to-market, innovation, knowledge management, multifunctional teams, etc. You tend (tended) to read about these sorts of strategies in Business 2.0 and Red Herring long before the "old economy" press heard about them.

  • Compete successfully: A slightly briefer way to say "create and retain sustainable competitive advantage" (using the language of Michael Porter's Competitive Strategy canon) or a slightly more verbose way to say "win" (using the language of the Boston Consulting Group's book Hardball).

Of course, for a long time software companies have successfully employed "old economy" strategies (Microsoft comes to mind), and "new economy" concepts such as product innovation will always be an important aspect of any successful software company's strategy. But I suspect that the relative importance of these two kinds of strategies will continue to shift.

I hope to learn and share some useful insights while writing this blog. Who knows where this may lead, but perhaps there will be something useful learned which may have applications for strategic planning and/or investment decisions.

I expect to draw upon a range of sources in addition to conducting some original analysis; these sources could include published studies and articles on the industry, news coverage of current events in the industry, books and publications which more broadly consider competitive strategy topics, SEC filings, and equity analyst reports.

Monday, November 12, 2007