Tech Industry Analysis

Open Source in the Mainstream - OSBC 2011 Notes

May 20, 2011                                                                               © 2011 MGI RESEARCH, LLC

Earlier this week we attended the OSBC 2011 (Open Source Business Conference) in San Francisco. Like the open source movement, this event has ebbed and flowed, and managed to stay relevant.  Here is our summary.

Open source software (OSS) is entering the mainstream – but to the chagrin of most VCs and early OSS companies, it has not played out as expected.  OSS is now widely adopted among the developer community, and if a company that develops software (either for internal consumption or as a business) has not incorporated OSS into the lifecycle of their development processes, they are behind.  OSS has emerged as a core enabler of cloud computing. Without it, much of the benefits of cloud computing would not be possible.  Successful software companies that are the commercial version of an open source project (Acquia/Drupal, Revolution Analytics/R, et al) have figured out that there is little advantage uniquely promoting their “open source identity. Rather, the successful ones (e.g., RedHat) push more traditional marketing messages of being better, faster, cheaper, more innovative than their competition.

The greatest barrier to more widespread adoption of commercial efforts associated with open source projects is skills – or the lack thereof. Corporate IT shops are reluctant to invest the time, money, and human resources necessary to ramp up in a new technology that may or may not be incrementally better than what they currently use. This is particularly true in core areas like database, where companies have major resource investments in Oracle, IBM (DB2) and Microsoft skills.  

The open source threat to established product franchises, - and their associated revenue and profit streams, centers around a couple of key areas. First, organizations (e.g., companies and government IT departments) that rely on open source rather than a “proprietary” solution from a “proprietary” vendor like IBM or Microsoft, often highlight the benefits of 1) better support, and 2) technical innovation received from the open source software vs the competition.  Beyond the communal nature of the support environment that is fostered within the open source community, there is little that is a truly unique differentiator. In fact, companies like SolarWinds have proven that a “proprietary” vendor can provide the same community-based benefits for customer support of an open source solution, without going the open source route, i.e., “customer-sourcing” of support and training.  Similarly, so-called proprietary vendors that continue to push the innovation envelope are less likely to encounter market-share competition or margin erosion from OSS vendors.

For a software vendor or user, the risks associated with open source now center on IP ownership.  It’s not an issue of licensing and the threat of legal action. Rather, it is an issue of being able to clearly identify ownership of all internally developed IP.  Users and vendors alike would be well-served to get their arms around this before it becomes a show-stopper to a merger or acquisition. Companies like Black Duck, Protecode, and Palamida are carving a niche market for source code identification and management. Two years from now these tools will likely be part of every development shop, and eventually be integrated into broader application lifecycle management (ALM) suites, like IBM Rational and Hewlett Packard’s BTO, as well as assorted agile-focused ALM tools from vendors like Rally, ThoughtWorks Studios, Collabnet, Atlassian, and Version One. This broader space is poised for consolidation – if and when IBM, HP, and even perhaps Compuware, recognize the growing market demand for such tools.

For the biggest success in open source – Red Hat and its Linux operating system, there appears to be a clear path to greater market share growth. The combination of growth in cloud computing and the clear shift from Unix to Linux as the server OS of choice give RHT a one-two punch that should boost expanded growth above the industry trend lines. 

Bottomline: We remain bullish on RHT and believe it will continue to out-perform given its market leading status and continued strong execution. Open source companies will have their share of market success and failure, and the gating factor will not be whether or not the company is open source, but more simply how well they execute a profit-driven business with a differentiated offering.

 

SugarCRM CEO Departs - Open Source SaaS Not Immune to Recession

May 11, 2009  (c) 2009 MGI RESEARCH, LLC

On May 7th 2009, SugarCRM (MGI-X:NR; pvt.), a highly visible open source CRM software company announced that its CEO John Roberts was leaving the business effective immediately. Board member and founder of VA Linux, Larry Augustin is assuming the role of CEO.  It is unclear if a search for a permanent replacement CEO is underway.  As MGI Research discussed in a note published on January 7, 2009, titled, "2009 Outlook for Open Source: Do or Die Time", the halcyon days of easy capital and low near-term profitability expectations for VC-backed open source companies are likely to come to an end this year.  MGI Research estimates that SugarCRM's revenues are close to the amount of total capital raised, i.e., $45 million, and it's probable that the company is encountering the same headwinds as others in the SaaS/enterprise software market - lower renewal rates, downward pressure on pricing, and drop in the deferred to realized revenue ratio, a key indicator of soft demand.  

The competitive impact of this specific event is minimal - SugarCRM has struggled to compete effectively against Netsuite, Oracle, RightNow, Salesforce.com and SAP.  But from an industry perspective, this event highlights the challenges even well capitalized open source software (OSS) vendors like SugarCRM face in building durable businesses, and how OSS companies by and large have failed to be universally disruptive.  The success of OSS (and indeed it has been successful in a few niche markts) should be measured company by company and market by market.  Ultimately, management execution and not the degree of open source conformance will determine success or failure in specific OSS companies.

Two core business assumptions regarding the OSS business will now come into focus and face a reality test:  1) that the recession will benefit open source companies, and 2) downloads are more important than bottomline profits.  CEOs of OSS companies should take note of this event - it could be the first of many replacements this year.  Investors should note that the success of Linux, and namely RedHat (MGI-X:840; NYSE:RHT) is not transitive - there are plenty of lackluster OSS results, and the OSS business model has not proven to be more efficient or even as profitable, as evidenced by RHT's and Novell's mediocre MGI-X scores.

SAP-Teradata Announcement Reflects Pressure on SAP vs. Oracle

April 28, 2009  (c) 2009 MGI RESEARCH, LLC

The recent SAP-Teradata announcement that SAP's BW (Business Warehouse) will be ported to Teradata (NYSE: TDC; MGI: NR) comes only one week after Oracle's public bid to acquire Sun Microsystems. The press release was light on details - e.g., shipment dates, channel issues, scant mention on either company's website, and in the short term was of greater benefit to Teradata, which extended its ability to re-sell the SAP/Business Objects BI products.  Given the tactical nature of the announcement, and the practical reality that it will be at least 6-9 months before customer beta testing of the SAP/Teradata combo, MGI Research sees minimal short term benefit of the announcement to SAP, and minor impact to IBM or Oracle.  In the bigger picture, the event highlights the tectonic shift that will occur if Oracle (Nasdaq: ORCL; MGI-X:2,314) closes the Sun (Nasdaq: JAVA; MGI-X:656) transaction. We re-iterate our view that HPQ or IBM are likely bidders for SAP within the next 24-36 months.   

Absent from the announcement was any reference to a deal around demand forecasting - a known weakness for SAP, particularly in retail and in spite of its Khimetrics acquisition, and a relative strength for Teradata.  The lack of any news means the expected Teradata/SAS (private; MGI-X: NR) partnership may still go forward. 

The deal underscores the emerging reality that Oracle will effectively distance itself from SAP, and now compete head to head with HP (NYSE: HPQ; MGI-X:1,822) and IBM (NYSE:IBM; MGI-X:1,769) if it finalizes the acquisition of Sun.  HP's relationship with Oracle will cool, and the HP/Oracle Teradata-killer product ("HP/Oracle Database Machine") is likely relegated to the junk heap of joint development projects. Oracle's strong MGI scores through 30+ acquisitions large and small points to its ability to innovate through acquisitive growth combined with operational efficiency.

For a company that historically possessed tremendous strategic vision and world-class execution, SAP increasingly looks tactical and lacking a coherent strategy.  SAP abandoned Cognos as the two companies were working joint go-to-market efforts when SAP launched its bid for Business Objects, pushing Cognos into the hands of IBM.  Global 2000 CIOs are eager to reduce their supplier base, and without a broader product set or a clear edge in innovation, SAP is at risk of losing its seat at the table of industry leaders inside the enterprise.

MGI Research believes IBM is an innocent bystander in this announcement - neither benefiting significantly from it, nor feeling any pain from SAP/Teradata announcement-ware.

Bottomline: The SAP-Teradata announcement does little to alter the rapidly changing industry dynamics in which Oracle is emerging stronger than ever, with impressive MGI-X scores, and SAP is playing a reactionary game in which it is unclear if it can or will compete via acquisition or internal development.  If Oracle is successful in integrating Sun (and it maintains ownership of the hardware assets), IBM and HP will be under pressure to make a run for SAP - a logical combination for both companies. Teradata may be a tuck-in acquisition for a larger player, or more likely, simply an industry orphan with a robust installed base and maintenance revenues.

OSBC Open Source Conference Highlights

April 1, 2009 (c) 2009 MGI RESEARCH, LLC

We recently attended InfoWorld’s 6^th annual OSBC event in San Francisco – where the open source community convenes in a business setting.  The major theme was the attraction of open source products' low cost during a period of IT budget contractions.  Our major interest was to measure the relative success of open source companies in reaching the enterprise market and their ability to threaten the established vendors.  The short conclusion is that open source projects continue to make inroads into Silicon Valley development efforts, but as a business, open source companies have yet to prove their long-term viability with a business model capable of delivering attractive returns to investors. 

Conference Highlights

Open Source = Cheaper Alternative to Established Vendor Offerings

A key theme from most open source companies was “our product is the low-cost alternative to the expensive ‘proprietary’ solution from major vendor X”.  Interestingly, only Ingres was able to produce numbers remotely close to a true TCO analysis comparing the cost of adopting Ingres versus the competition.  Most open source vendors have not made the marketing transition from targeting their developer communities to delivering durable marketing messages to the enterprise buyer. 

Show Me The Money!

Linux has clearly established itself in the datacenter, and is increasingly gaining traction on a certain percentage of netbook shipments.  Multiple data points, examples and anecdotes were provided, and this is corroborated by what we are hearing elsewhere in the industry.  Nonetheless, as a venture capitalist heavily invested in open source highlighted during a keynote panel, even RedHat, which is viewed as the poster child of open source success with its Linux products, has only managed to win less than 5% market share.  In the database sector, we estimate mySQL’s revenues to be in the $80 million range, and when combined with the revenues of Ingres (FY’08 revs of $68 million, profitable), they are insignificant to a $15 billion+ database market.  Our takeaway from this is that open source companies are impacting the margins of established companies, but have not gained sufficient credibility within the enterprise to cause long-term damage to the core business model of companies like Microsoft, Oracle, SAP, Salesforce.com, and others. As previous MGI Research has detailed (see Research Note - "2009 Outlook for Open Source - Do or Die Time?"), the total revenues of companies like Alfresco, Jaspersoft, SugarCRM, Talend, and others are not nearly as impressive as the venture dollars these companies have attracted, and most of the leading venture-backed open source companies are on their Series C, D, or even E rounds of capital. Surely the pressure is on them to start delivering profits in advance of a desired investor exit.

Will Recession Push Corporate Buyers to Consider Open Source Products?

It’s unclear if the pressure on IT budgets will drive CIOs to replace installed products with open source alternatives.  One of the major hurdles to enterprise adoption of open source was the lack of internal skills necessary to implement and manage an open source product.  Headcount is typically the largest cost line-item in an IT budget, and an obvious target for savings.  It’s unlikely that CIOs can manage a headcount reduction (and the loss of skills), and then expect to adopt an open source product that demands a new set of skills the remaining IT department employees are unlikely to have. 

Database Market Update

In the database world, Ingres may do better than mySQL, given its legacy as a proven enterprise-class product and the $30 million of R&D investment pumped into improving the product over the last three years.  In addition, even Sun folks mentioned the impact of losing mySQL founder Martin Mickos, who leaves Sun at the end of this week.  Mickos was treated like a rock star at this event, and there are rumors that he is planning on buying mySQL back from IBM, in the event that IBM is able to acquire Sun.  It’s clear that mySQL market momentum suffered just after it was bought by Sun, and Oracle’s strategic move of buying InnoDB, the leading transaction storage engine of mySQL, also slowed the growth of mySQL.  Microsoft appears to be a beneficiary of the cost pressures on database spending, as SQLServer is perceived as a lower-cost, low risk alternative to Oracle and IBM.  The core Sybase database business appears to be under pressure, as mySQL and Ingres both make inroads into the installed base that is growing tired of paying maintenance and perceives Sybase to be an absentee owner of the product.

Overall Conference Mood

Not surprisingly, conference attendance was off – organizers claimed 600 registrations, but the kick-off keynote session had about 200 in a ballroom that was reduced in size by one-third.  Our non-empirical estimate is 60-70% of attendees were vendors, 20-30% were industry service providers (lawyers, VCs, press), and 10-20% were user organizations.  A noticeable difference from other recent events was the energy emanating from the open source community.  It was ever so faintly reminiscent of the buzz of years past in the tech industry.