Open Source CRM: The Facts

The concept of open source software is simple. The source code is freely available, publicly modifiable and flexible. The model puts control in the hands of users in terms of development and user licensing is cheaper than the traditional proprietary software model. Having anchored itself within the infrastructure stack, open source is starting to break into the world of business applications and customer relationship management is an early target.

CRM is a mature market with high barriers to entry. Even Siebel is unable to maintain its independence, and market share is increasingly going to application giants SAP and prospective Siebel-owner Oracle. Despite these issues new entrants are jostling to enter the CRM sector, in the form of open source business applications providers. They are attracted to it because of its size: every sales, support, service and marketing person within an organisation is a potential user.

"CRM is a huge market and we do need a huge market. You need a large community," says John Roberts, CEO of commercial open software CRM provider SugarCRM. "We view CRM not as a nice application to have but as a mission-critical application. There are 50 million companies worldwide that need CRM to manage their customer interactions.

"The big players have a fraction of the market because they cannot reach companies worldwide: they do not have the Internet to [establish] the reach," he adds. By utilising the Internet as both a distributed development environment and a distribution channel, open source CRM companies can gain geographical reach without requiring costly company infrastructure.

The Internet continues to change the way IT works but the benefits of low cost and very broad reach are not restricted to open source vendors. Salesforce.com uses the Internet as both a distribution medium and an operating system. "There is an opening happening in the industry, in the technology made possible by the Internet and by the way we use technology," says Marc Benioff, Salesforce.com CEO.

The barriers to entry - and survival - in the CRM market are high. CRM is a matter of money, market share and integrated application suites, which is why dedicated CRM vendors are being acquired: from analytics specialist Epiphany to CRM market maker Siebel. "The barrier to entry to CRM is not software but who can spend the most on sales and marketing," says SugarCRM's Roberts. "Is this really the best manufacturing model for enterprise software? Is it in the best interests of the customers?"

Roberts believes the balance between R&D and sales and marketing is totally wrong. "Our approach was to develop in public and make the source code available from the beginning. There is an Orwellian dogma propagated among the software vendors that they can only give the .exe files to the software. They write in a closed environment because if they give the code away it will break and create fear and uncertainty. They have a dirty little secret: the reason why they are not willing to share their source code is because it is the base of their revenue stream," he says.

Open source uncertainty

Prospective buyers need to understand what they are getting into when they opt for open source over proprietary software and how to operate in an environment where it is as important to make requests known as it is to contribute to development.

Many of the services provided via the proprietary software vendor model are replicated and available at a lower cost under open source but there is uncertainty over the who, how and when. This is particularly significant to integration and vertical sector development, the most costly areas of application management.

In the current climate is it virtually impossible for new entrants to create a sustainable position in the CRM market using the traditional model. Open source represents the rise of engineering over marketing and has the potential to result in software that is developed faster, is more innovative and more 'verticalised' than proprietary software because it captures vision from developers worldwide and has a diverse user base.

The absence of licence fees is a large part of the open source CRM economic argument, but where CRM software is concerned licence fees are only 10% to 15% of the total cost of ownership. Implementation, hardware, the cost of internal or outsourced support, customisation and maintenance are where the real cost- of-business applications lie.

Phil Robinson, Salesforce.com senior VP for global marketing, says: "Open source reduces the cost of managing traditional software. We are not open source because we are not selling software. [With open source] the code is on the machine. That is different and you still get the traditional software headaches," he says. "On-demand delivers a service so companies can build an application, not maintain a development environment."

While open source CRM may win on economic grounds, it needs to offer something as good or better than conventional CRM software. Open source CRM is not as functionally rich as packaged CRM applications. It is still basic and lacking sophisticated functionality such as high-end or predictive analytics.

There are potential risk factors in terms of development, application maintenance and upgrades where custom integrations are concerned as well as potential security and stability issues. There is also the knowledge issue in that organisations that opt for open source need in-house skills to manage it from a technology and business perspective.

Roberts argues that with a self-interested, worldwide developer community engaged in peer review, open source software is robust, stable and secure, as well as enterprise class. He also refutes the idea that open source CRM is only for developers, pointing out that less than 5% of the SugarCRM customer base customises the source code. Where there is a requirement, there is a visual tool layer.

Blurring the boundaries

As far as upgrade issues and user customisation is concerned, the distributed development methodology enforces separation of the various software layers - such as code from business process, where most customisation takes place - which should reduce the risk of upgrade errors. However, many of these developments are making their way into traditional software with the rise of distributed computing, service-oriented architecture and visual tools, making it a common factor rather than a unique open source advantage.

Despite being taken up by small to medium-businesses rather than large enterprises, open source software is not just an SMB solution, chosen because buyers feel they are priced out by the tier-one vendors. Open source is adaptable and can give SMBs what packaged vendors have failed to deliver for years and are only now starting to address. Historically, packaged vendors offered software that was too costly to customise to suit the specific needs of the SMB community. They are on the path of change now, but they have competition and open source is part of that.

Open source woos with the promise that organisations can get what they need and that the application can be moulded to the business. Users have a much greater chance of finding something that fulfils their needs with little modification and have the opportunity to influence development. It can also be a springboard to custom development. Even Siebel believes that the market for custom development is larger than packaged software and this is the basis of its planned Nexus project.

Where open source is concerned it is not so much a case of buyer beware, as buyer be aware. Those who take the open source route need to know what they are getting into and while the approach may suit some organisations it is not for everyone. There is risk involved. Although it can be mitigated by opting for commercially available open source software, there is a lack of sophisticated CRM functionality. Those with advanced requirements will need to be prepared to develop and/or or lobby for development

No technology lock-in

There are critical technology decisions to be made at every point as opposed to the comparatively easy single vendor application/platform approach. Before opting for a particular open source CRM project, organisations also need to ascertain how big and active the community is. On the positive side, there is no risk of being locked into a vendor-specific infrastructure stack, which is what Microsoft advocates with its Dynamics products - its re-branded Microsoft Business Solutions applications such as Great Plains and Navision - as does Oracle and SAP.

Although the open source versus traditional software arguments are valid there is a more fundamental issue. Open source CRM is benefiting from a change in market dynamics, the same change that has fuelled the success of the hosted CRM vendors. Hosted providers showed the buying community that there was an alternative to the traditional model. The exceptional growth rate each of the vendors engaged in web-based on-demand applications has experienced shows that the market was ready and waiting for something different.

Now that companies like Salesforce.com have proved the value of a new model, the market is more receptive to alternatives such as open source CRM. The most bizarre aspect is that although the concept of open source predates the hosted model, it has taken the hosted model to breathe life into open source applications.

Ironically, Salesforce.com is adding further credence to the open source approach by adopting open source principles. AppExchange, its forthcoming application exchange platform, combines the principles of the open source distribution model with the dot-com trading exchange concept. "This opens what we call the sixth level of on-demand," says Benioff.

"Where level one brought applications delivered as a service, level two was about integration, level three added customisation, level four concentrated on creating an ecosystem and level five was concerned with delivering a platform, level six takes things a stage further. It is about the ability to share, buy and sell applications and run them on the Salesforce.com platform."

Development for free

"The Internet democratised access to applications, now that has been amplified through AppExchange," says Benioff. "Developers can create applications with no software, hardware or IT complexity and distribute them without the same restraints."

In an effort to encourage usage the company will not charge for its own applications nor will it charge developers to use the service as a distribution channel for their applications. It is "development for free" says Benioff: free of licence fees as well as distribution and hardware costs. It all sounds remarkably similar to the open source mantra with the crucial exception that Salesforce.com still retains ownership of its code.

It is not so different to the open source environment of freely available source code and development, but by restricting it to the Salesforce.com platform it adds an element of control and formal certification.

According to Salesforce.com director of product marketing, Adam Gross, the company has made a major commitment to open source. As of April this year it has enabled open source developed components to interact with Salesforce.com via an API and has contributed large amounts of code to the open source environment.

Salesforce.com sees open source as a complementary operation. "The end game is that each subscriber makes other subscribers more successful because they contribute to each other. Open source is good, if you are technical. AppExchange takes that up a level so you do not have to be so technical," says Gross.

CBR Opinion

One of the most important aspects of open source CRM is that it represents another choice: traditional packaged, hosted or open source CRM is available. Salesforce.com was the first to show that there was an alternative to the traditional approach, and the open source CRM movement is benefiting. It is not case of one or the other, or even of Salesforce.com versus SugarCRM. As far as users are concerned choice and appropriateness are the watchwords and with the growing number of models, the chances of creating a better match between application and business requirement are improving.