By Ulrike Mueller, Demandware Engineering

I understand there are some concerns about the "hype" surrounding open source ecommerce software and how it could become a "trap" for some retail businesses, but I think it will be only a "trap" for those businesses that don't understand basic principles of IT—like total cost of ownership and return on investment. Companies who understand these principles and know how to use them are able to make an intelligent decision for their business. Any diligent analysis will factor out the pure software aspect—which encompasses the functionality and features—from the service aspects—which encompasses operating the software and providing capabilities like scalability and availability. It is important to remember that these are two separate aspects and just looking at open source software as a single "download-and-run" task does not tell the whole story. Looking at both aspects separately and analyzing them individually really shows the benefits of open source software for businesses as well as the software industry as a whole.

Let's look at the benefits of open source. But first to clarify something, when we look at the benefits, let's only consider open source software, which has an active community and is maintained by the community. We won't address open source software that has been uploaded to a code sharing website by a smart engineer because of altruistic motivation. This type of open source software is very important for the software development community, but not immediately for enterprises. However, open source software developed and maintained by a community has many benefits for enterprises.

Transparency is probably the most important benefit of an open source product. By transparency I am referring to the feeling of having access to every detail, the feeling to be able to control and adapt everything. In reality probably very few companies will ever peek into any detail or will touch a single line of code. But the option gives companies the feeling of freedom and not being strangled by a "black box." The transparency also nurtures a crowd of experts, who actually use this level of insight. They are part of the transparency value system: I don't have to look into details, if there are others who can do it for me. One type of experts are companies and individuals who will help enhance and evolve the software through direct contribution and contribution on all levels. It's probably the most efficient form of customer feedback in a software product. It can happen in a small scale with a knowledge base posting on a blog, with a bug fix or in large scale with a new feature. These mechanisms are how open source software evolves through crowd innovation. A special form is probably the concept of a plug-in or module ecosystem, found with open source products like Wordpress, Joomla or Drupal. These are software products built from the ground up as an ecosystem. Another type of expert in the community will help and commercialize operating the software or even completely provide this as a service.

Transparency also helps improve the quality of open source software. The best example for this is cryptography algorithms, reviewed by thousands of eyes in the community. Transparency and community together build some things that can be called crowd trust: it's not just me using this software, there others using it, there are experts around it, who can peek in every detail, these experts help others, these experts can help me. I can Google for expertise and find support from others. This also shows that success for an open source product does not come from a single benefit. All these benefits interact and nurture each other. Transparency brings trust, which brings community, which brings crowd support and crowd evolution, which leads to more adoption and larger community. The mechanism can be best called a viral vortex, once ignited it brings a growing community and benefits for all participants.

But keeping that "viral vortex" successfully running is not easy. Some people might also have the perception of open source as a chaotic process of creative people who write lines of code. Instead successful open source projects are very often controlled with a strong regime of values and vision by a few core individuals. The most prominent example is probably Linux and Linus Torvald. The strong regime and the core individuals are the guarantees for consistency and vision. It's also the reason, why very often successful open source products are associated with a company. Companies like Zend, MySQL, and Redhat are examples. They are the "keepers" of the product vision, the center for a community and also a guarantee that the community and the software product will not become fragmented. I would say it's even one of the success factors for an open source product. In today's world of growing adaption of SaaS and cloud computing, there is also a huge opportunity for a software company to specialize on delivering the operations, scalability and availability as a value-add around an open source product. Well executed and well managed, it really combines the benefits of a classical IT service business and the benefits of an innovative crowd-sourced open software product, and benefits both the software company and the user.

The benefits described above are actually all fundamental values, which companies should adopt regardless of whether they deliver open source software, deliver closed source software or deliver a pure service. They are fundamental values to achieve "openness" for any software company. The transparency, community and customer orientation can have many forms. It starts from the company having an open dialog with their customers, to a community writing knowledge base articles and documentation, driving localization, developing free SDKs and providing access to the source code. Java is probably the best example of the latter example, which is perceived as open, without being open source from a licensing perspective. But this openness has really helped with the adoption of the technology, with innovation happening around it and with community building. Open source is often not possible, because of complicated legal aspects, but even then a software company should incorporate the underlying fundamental values into their vision and management principals.

We at Demandware have always been a customer oriented company and as a service company we live and breathe everyday with our customers and partners. Being open is a path we will and must follow.

Ulrike Mueller is Demandware's Chief Software Architect.

By Stephan Schambach, Demandware Founder & Executive Chairman

Thank you all for the responses to the post on open source. Great feedback and I have to say, while some of the comments are quite pointed, the dialogue and debate are excellent.

Based on some of the commentary, it would appear there's room for clarity and explanation in some of my original posting and there are a few general points that deserve response. The point-by-point responses are at the bottom of the posting.

Before that however, it would be useful to explain some of the background thinking that went into the post, based on my experience being a software entrepreneur—first in licensed software, now in on-demand—and a guy very close to the business (and technical) operations of a healthy growing company.

When we founded Demandware, one of our core philosophies was (and still is) to live what we preach—to focus on our core competencies and outsource the rest. In IT, for example, the rule is that if we need a system for something, then it must be SaaS or on-demand, and only if we don't find a SaaS vendor, we will look into alternatives, preferably Open Source, and commercial software only as a method of last resort.

As a result, we have a tiny 2-employee IT department for a sizable company, and there are absolutely no complaints about IT. We use an on-demand service for email and calendaring, Salesforce.com for CRM and marketing automation, Hubspot for web marketing, Intacct for financials, OpenAir for professional services automation, Concur for travel and expense management and the list goes on. And, believe it or not, we use the standard Demandware SaaS platform to run our website. I am very happy with the decisions we made. Business is not held up by IT.

Now picture a successful retailer with a great brand presence, a couple hundred stores, website etc. Their core focus is to merchandise, to understand consumers and to interact with them through multiple channels. Should this company focus on hiring software developers and IT staff, to build and operate a complex eCommerce platform consisting of server software (Open Source or commercial makes no difference here), databases, firewalls, redundancy, monitoring, backup, security ...? Or would it be smarter to enable the business side, the web-savvy merchandisers, to get their day-job done through a SaaS service that gives them all tools and flexibility without reliance on software developers and IT staff? I say, companies who know what their core competencies are the ones who will win.

OpenSource is a great innovation machine. It is incredibly successful where millions need infrastructure-type software that is free of licensing and modification restraints. But it is a rather technical building block, not a business solution. The higher up you move the software stack, the fewer potential end users and thus, a smaller community of developers. There are a finite number of companies in need of a high-end eCommerce platform, and building it "for free" is not an option for a viable business model. OpenSource vendors then inevitably offer a "pro" version that is not much different from a licensed software product, and it comes with the same problems: the vendor does not take responsibility for the actual business value generation. And this is, in my mind, no recipe for the future of enterprise software.

(BTW: we are not the first to have this debate. If you'd like to read a parallel discussion on the pros and cons of the open source business model, check out The Failure of Open Source Software on BusinessWeek.

As an end user, instead of just software download and customer/community support, I would like the vendor to deliver an internet-based system that is always up and running, that empowers the business users in my company and makes them more productive at what they do best. I want the vendor to accept responsibility for keeping this system modern and up-to-date with best-practice through automated software updates. I want the system to scale from where I start as a company, to wherever I can scale my business, without having to buy servers or re-architect my platform. I want to be able to configure and customize the system through Internet-based development tools. I want to interface it with my internal systems, or with other on-demand vendors easily.

Now to the specific comments:

We did not discuss OpenSource benefits enough. I believe I did discuss some benefits at the end of the article, but the point of the article was not to discuss the many benefits of OpenSource (those are known). Rather the point I was trying to make was that OpenSource in itself is not a recipe for success for a business application and the needs of retailers and brands.

Demandware holds retailers in a captive relationship, where OpenSource doesn't. Depends on how you look at it. In a heavily customized Open Source app, the business side—the eCommerce merchandisers—are dependent on either an agency for every change and pay through the nose, or they depend on an IT department and talent is hard to find, expensive, and even harder to keep. With Demandware, there is vendor dependency, of course. But what is delivered is well defined, and there is the same customization flexibility as with other development environments. Both paths create dependency: the question is, which path is goaled on your success?

Enterprise OpenSouce has an easily upgradeable core. This may be true in an academic sense. The companies I interviewed for my original post though all ran their eCommerce sites on different versions of different OpenSource packages and said that they could not upgrade, because the effort to do so with all the customization and integration is not in their budget.

Many viable OSS businesses. Yes, there are many successful OpenSource businesses and we fully recognize this.

By Stephan Schambach, Demandware Founder & Executive Chairman

There is a lot of hype (again) around open source ecommerce platforms. They seem irresistibly sexy: no cost, just a download away, own the code and modify, customize and develop at your own pace. Many features are developed by the "Community," and you could argue that no commercial software company could employ as many engineers. Also, open source applications that run in the cloud seem like an alternative to Software-as-a-Service (SaaS). But the business benefits of an open source based ecommerce solution are not always what they seem. Here are some things to consider when deciding whether or not to build your site on an open source ecommerce platform:

First, if one looks at the history of open source innovation in nearly every sector—not just ecommerce—there’s a natural evolution of enthusiasm from a core set of pioneers, to mass adoption as the idea spreads and gains traction. At some point however, the masses fragment into spin-offs and subcultures, as the developments of one mass community becomes too unwieldy to address the specific needs of the sub-segments. So sub-segments secede, start their own project to address their needs, attract followers to the next new open source project. And history repeats itself. Such is the rise and fall of open source and it is—and will be—no different with ecommerce.

We’ve seen this before, with OS Commerce and XT-Commerce, whose users are now left with an old platform that receives very little community development. True to form, this has given rise to the new open source platform, Magento. Time will tell whether or not it will be the one platform that breaks the historical trend. Sure, there have been some examples of open source software that have remained successful in the long-term (Linux is a great example) but these are typically technical infrastructures that millions of people need, not business applications used by business users.

Second, most brands and retailers do not have the technical resources or desire to "develop" their own platform—for good reason. There is no enterprise solution that is more complex—and consists of more pieces—than ecommerce, and the mere server software is just a tiny piece of it. Experience has shown that the cost of an open source based "home grown" ecommerce site to be equal to or even greater than commercial offerings, and certainly more expensive than a SaaS platform. The reason is that the open source user owns the responsibility for all peripheral pieces belonging to a functioning ecommerce website such as hardware, databases, firewall, security, monitoring, content distribution, analytics, search engine, scaling, software upgrade, ongoing feature development and so forth.

Third, open source implementations will lock retailers into a captive relationship with a systems integrator or agency because once a site is implemented, they end up with a proprietary, totally customized solution that only this one company can maintain, leading to further economic disadvantage after the initial deployment.

Fourth, users of open source must upgrade their software themselves. Often this is simply impossible due to code-level customizations that are incompatible with the new version. With SaaS, the provider will upgrade frequently and keep all customers on the newest version automatically and without any service interruption. In other words, once an open source solution is deployed, it becomes a liability to upgrade and therefore cuts the customer off from affordable innovation.

And last but not least, in an open source environment the end user is responsible for quality assurance (QA). Most functions developed by the community are raw and have not gone through the same level of testing as with SaaS based systems. Being a beta tester can be expensive.

Now, that is not to say that Demandware is opposed to open source—we just use it differently. Much of our software stack—operating system, application servers, etc.—is open source. But we build a commercial SaaS platform on top of it and do all the heavy lifting for our customers. We take care of upgrading underlying components, should this be necessary, and we can afford it because we spread the effort among hundreds of customers.

All this is not to say that open source is not a viable option for your business, but rather to help you see beyond the flashing neon sign that says "FREE!" Before diving into any ecommerce platform, open source or otherwise, be sure to understand the total cost of ownership not just for the platform, but also for the ancillary hardware, software and support that is required.

As a result of the many comments we have received in response to this post, we have published a follow-up post—Open Source eCommerce, Part II. We thank you all for your comments and encourage you to read on and participate further.