The Success of Commercial Open Source Software

By Danny Windham
The Success of Commercial Open Source Software 2009-07-29
Yesterday, an article was published in Business Week by Peter Yared, who brought up some points that I’d like to address from the perspective of a successful Commercial Open Source company.   If you’ve not read his article, take a look here: The Failure of Commercial Open Source Software.  http://www.businessweek.com/technology/technology_at_work/archives/2009/07/the_failure_of.html   Hopefully I can give a different perspective on how we see almost the opposite results from many of the same points that he raises in his article.
Peter’s premise is that commercial open source software must be failing given the lack of liquidity events for companies delivering commercial open source offerings.  This is clearly the wrong way to measure success of commercial open source software – which should be based strictly upon the market penetration of such products and the level by which end users embrace and rely upon the offerings.  Using that as the metric our contention it that commercial open source software is a resounding success.
So, where’s the disconnect between these two seemingly interrelated principles?  After all, one should easily follow the other – right.  In reality, building a successful open source company in many ways is more difficult and time consuming than building a corresponding company delivering proprietary software solutions.  In the proprietary model, organizations are free to charge for their wares based upon determining a willingness to pay on the part of the buyer.  End users then often pay large acquisition fees for proprietary software choosing the most cost- efficient option among several proprietary alternatives.  Proprietary software vendors can then drive large revenues from a correspondingly small number of customers – and landing a series of these customers in a short period of time can translate into rapid revenue growth. Traditionally Wall Street has valued rapid growth above all else, which has provided a number of liquidity options for those companies fortunate enough to find themselves in this position – which has also produced several liquidity events that proved to be charged with irrational exuberance.
On the other side of the coin, commercial open source companies always have one additional competitor not found in the prior scenario – which is ‘do-it-yourself using in-house labor and pure open source software’.  This forces commercial open source vendors to routinely prove their value-add over pure open source alternatives (necessary if they are going to survive) and serves to limit what these companies can charge for their products and services to an amount equal to or less than the amount the end user determines as their internal cost to do-it-yourself.   Therefore, successful open source companies encompass a much more diversified customer base with much less reliance on any one customer than their proprietary brethren of similar size.
However, some companies may be well-suited to deploy OSS packages and some may be incapable of managing such a process – finding it more cost-effective to rely on vendors to perform services on their behalf.  We at Digium have embraced both models in our niche of the telecommunications software market by offering pre-built OSS platforms (SwitchVox) along with totally free versions of our software (Asterisk), as well as selling support and training for those “in-between” cases.  While the last year has been challenging, Asterisk and Switchvox both continue to gain market momentum, and typically at the expense of proprietary vendors.  Simply because there have been no big IPO events for OSS companies does not mean the industry is moribund – I think that you’d be hard-pressed to find anyone who has been cheering recently.  As far as COSS companies starting up: I’d say that market is quite vibrant, but the number of funding events that move small companies to notable companies has been small due to the market in general, and shouldn’t be taken as an indicator that is tied directly to COSS viability specifically.
I also don’t quite agree with the statement that the costs of open source and proprietary are the same.  For OSS, development costs are significantly lower – zero, in some cases – because of the community of developers who contribute to the software.   This is fairly obvious, and saying that the costs are identical is a bit disingenuous.  Support costs can be the same from a labor perspective, but it is possible for a skillfully-staffed IT department to benefit from the extensive support structures available via user-generated help, the wide market of consultants available for OSS, and from training in-house staff for self-maintenance.
Another great thing about OSS is that even if the company pulls support for the product (I’m talking to you, Microsoft Response Point) or goes out of business, there is an opportunity for your company (or consultants of your choosing) to continue making repairs and even advancing the tool despite lack of support from a vendor.
To look at Peter’s arguments from the business of selling software, there is a slightly different perspective.  Corporations based on sales of open-source software can suffer from exactly the same frailties as commercial software – I’ll certainly admit that, and we try to avoid those pitfalls, but I don’t think that similarity makes a significant difference to the end user.  One of the most obvious points of OSS is that it’s free – one does not need to spend money with an external company to realize the benefits of the package.  You can spend money with a vendor if you wish, if you find it makes more economic sense for your firm.  Buying brain-time to support software is expensive, yes – but OSS gives you the ability to opt-out of the huge initial costs of development pay-back.  OSS and proprietary vendors may charge exactly the same for their services, but since OSS platforms don’t have the same cost of development, they can make the same margins while offering a better product.
I’ll wholeheartedly agree with Peter that selling software is a miserable dance and a difficult hurdle for OSS to overcome.  Big proprietary vendors can spend huge sums on the up-front sales process, overwhelming most OSS vendors with the sheer volume of data, cut-sheets, and glossy noise of the pre-installation portion of any software sales  process.  Even through all that, OSS is on the rise in enterprise.  Why?  Because of the quality and flexibility of the solution, and the low cost (cost being the least significant issue, typically.)  OSS operates on merit, and is almost always championed by internal advocates who know what they need to get the job done in the fastest and most effective manner.  I will say that of Peter’s points, this one is the most true, but OSS is still making huge inroads even without the coordinated methods of giant salesforces.  I’ve always thought that OSS is really the friend of all developers, and the enemy of most salespeople, though it is good news for our salespeople when we compare the total cost of ownership of our OSS solution versus a fully proprietary alternative.
SaaS is on the rise in some sectors, but certainly not in all.  We find SaaS a great partner, as many SaaS companies that do VoIP, as an example, are using Asterisk in their service delivery platform, and they’re very interested in forming close bonds with us from a business and community standpoint.  SaaS is a great thing, but companies should understand completely the long-term costs as compared to in-house solutions.  Five dollars a seat per month sounds great, until you discover that over a five year plan that turns into something that looks like double or triple the cost of an OSS solution, even including labor and infrastructure, and now you have to contend with vendor lock-in and risks associated with SaaS companies having technical or business problems which may leave you in a very tight spot.    I think there are a number of good points Peter has made, but none of them lead to the conclusion with which the article is titled.  In fact, it seems to me that exactly the opposite is happening.  If success is measured by irrational buyout pricing and IPOs that make millionaires out of lunchroom attendants, then perhaps OSS has fallen short.  But by a more measured perspective of business growth, success in delivering solutions, and overall growth of OSS in businesses, I think we’re in great shape. I’d respectfully suggest he re-title your article to “The Success of Commercial Open Source Software” and we’d be on the same page.

Yesterday, an article was published in Business Week by Peter Yared, who brought up some points that I’d like to address from the perspective of a successful Commercial Open Source company.   If you’ve not read his article, take a look here: The Failure of Commercial Open Source Software. I don’t often make direct replies to articles like this, but it tied in well with some of the thinking I’ve been doing recently about why enterprise adopts Commercial OSS and how well Digium’s model fits into that space in conjunction with our traditional role of SMB products and software.  Hopefully I can give a different perspective on how we see almost the opposite results from many of the same points that Peter raises in his article.

One of Peter’s premises is that commercial open source software must be failing given the lack of liquidity events for companies delivering commercial open source offerings.  This is clearly the wrong way to measure success of commercial open source software – which should be based strictly upon the market penetration of such products and the level by which end users embrace and rely upon the offerings.  Using that as the metric our contention is that commercial open source software is a resounding success.

So, where’s the disconnect between these two seemingly interrelated principles?  After all, one should easily follow the other – right?  Widespread use of a certain software package should quickly lead to financial success, or so it would seem.  In reality, building a successful open source company in many ways is more difficult and time consuming than building a corresponding company delivering proprietary software solutions.  In the proprietary model, organizations are free to charge for their wares based upon determining a willingness to pay on the part of the buyer.  End users then often pay large acquisition fees for proprietary software choosing the most cost- efficient option among several proprietary alternatives.  Proprietary software vendors can then drive large revenues from a correspondingly small number of customers – and landing a series of these customers in a short period of time can translate into rapid revenue growth. Traditionally Wall Street has valued rapid growth above all else, which has provided a number of liquidity options for those companies fortunate enough to find themselves in this position – which has also produced several liquidity events that proved to be charged with irrational exuberance.

On the other side of the coin, commercial open source companies always have one additional competitor not found in the prior scenario – which is ‘do-it-yourself using in-house labor and pure open source software’.  This forces commercial open source vendors to routinely prove their value-add over pure open source alternatives (necessary if they are going to survive) and serves to limit what these companies can charge for their products and services to an amount equal to or less than the amount the end user determines as their internal cost to do-it-yourself.   Therefore, successful open source companies encompass a much more diversified customer base with much less reliance on any one customer than their proprietary brethren of similar size.

However, some companies may be well-suited to deploy OSS packages and some may be incapable of managing such a process – finding it more cost-effective to rely on vendors to perform services on their behalf.  We at Digium have embraced both models in our niche of the telecommunications software market by offering pre-built OSS platforms (SwitchVox) along with totally free versions of our software (Asterisk), as well as selling support and training for those “in-between” cases.  While the last year has been challenging, Asterisk and Switchvox both continue to gain market momentum, and typically at the expense of proprietary vendors.  Simply because there have been no big IPO events for OSS companies does not mean the industry is moribund – I think that you’d be hard-pressed to find anyone who has been cheering recently.  As far as COSS companies starting up: I’d say that market is quite vibrant, but the number of funding events that move small companies to medium-sized notable companies has been small due to the reluctant market in general, and shouldn’t be taken as an indicator that is tied directly to COSS viability specifically.

I also don’t quite agree with the statement that the costs of open source and proprietary are the same.  For OSS, development costs are significantly lower – zero, in some cases – because of the community of developers who contribute to the software.   This is fairly obvious, and saying that the costs are identical is a bit disingenuous.  Support costs can be the same from a labor perspective, but it is possible for a skillfully-staffed IT department to benefit from the extensive support structures available via user-generated help, the wide market of consultants available for OSS, and from training in-house staff for self-maintenance.

Another great thing about OSS is that even if the company pulls support for the product (I’m talking to you, Microsoft Response Point) or goes out of business, there is an opportunity for your company (or consultants of your choosing) to continue making repairs and even advancing the tool despite lack of support from a vendor.

To look at Peter’s arguments from the business of selling software, there is a slightly different perspective.  Corporations based on sales of open-source software can suffer from exactly the same frailties as commercial software – I’ll certainly admit that, and we try to avoid those pitfalls, but I don’t think that similarity makes a significant difference to the end user.  One of the most obvious points of OSS is that it’s free – one does not need to spend money with an external company to realize the benefits of the package.  You can spend money with a vendor if you wish, if you find it makes more economic sense for your firm.  Buying brain-time to support software is expensive, yes – but OSS gives you the ability to opt-out of the huge initial costs of development pay-back.  OSS and proprietary vendors may charge exactly the same for their services, but since OSS platforms don’t have the same cost of development, they can make the same margins while offering a better product.

I’ll wholeheartedly agree with Peter that selling software is a miserable dance and a difficult hurdle for OSS to overcome.  Big proprietary vendors can spend huge sums on the up-front sales process, overwhelming most OSS vendors with the sheer volume of data, cut-sheets, and glossy noise of the pre-installation portion of any software sales  process.  Even through all that, OSS is on the rise in enterprise.  Why?  Because of the quality and flexibility of the solution, and the low cost (cost being the least significant issue, typically.)  OSS operates on merit, and is almost always championed by internal advocates who know what they need to get the job done in the fastest and most effective manner.  I will say that of Peter’s points, this one is the most true, but OSS is still making huge inroads even without the coordinated methods of giant salesforces.  I’ve always thought that OSS is really the friend of all developers, and the enemy of most salespeople, though it is good news for our salespeople when we compare the total cost of ownership of our OSS solution versus a fully proprietary alternative.

SaaS is on the rise in some sectors, but certainly not in all.  We find SaaS a great partner, as many SaaS companies that do VoIP, as an example, are using Asterisk in their service delivery platform, and they’re very interested in forming close bonds with us from a business and community standpoint.  SaaS is a great thing, but companies should understand completely the long-term costs as compared to in-house solutions.  Five dollars a seat per month sounds great, until you discover that over a five year plan that turns into something that looks like double or triple the cost of an OSS solution, even including labor and infrastructure, and now you have to contend with vendor lock-in and risks associated with SaaS companies having technical or business problems which may leave you in a very tight spot.    I think there are a number of good points Peter has made, but none of them lead to the conclusion with which the article is titled.  In fact, it seems to me that exactly the opposite is happening.  If success is measured by irrational buyout pricing and IPOs that make millionaires out of lunchroom attendants, then perhaps OSS has fallen short.  But by a more measured perspective of business growth, success in delivering solutions, and overall growth of OSS in businesses, I think we’re in great shape. I’d respectfully suggest he re-title your article to “The Success of Commercial Open Source Software” and we’d be on the same page.

– Danny Windham, CEO

Related Posts

There Are 2 Comments

  • Peter Yared says:

    Danny, thanks for the balanced, rational response to the article I wrote. For whatever reason you are the only one who actually addressed the points that I brought up about the costs of running a software business, the enterprise sales model, and SaaS. Great discussion. Peter

  • Bandaru Lokamanya says:

    We have to develop Digium swithvox phone system in c#.net and asp.net.We have the following requirement like
    Connect to the phone system API and create an extension for the user.
    i. You can obtain the last extension used from a local text file that we will call “LastExtension.txt”.
    ii. The phone system requires: Firstname, lastname, email, extension, extension password, phone password.
    1. Extension Default Password: This is just extension repeated twice (eg: extension 101 password is 101101)
    2. Phone password: Same as default extension password
    e. After creating the extension, update the new user’s AD info with the new extension in the main telephone line
    g. Display confirmation to the user that everything is setup.

    These all should develop in c#.net and asp.net.Please provide the guidance for all this one to do in web based application using c# and asp.net.

Add to the Discussion

Your email address will not be published. Required fields are marked *

About the Author

Danny Windham

Danny Windham joined Digium in February, 2007 as CEO. In this role, Windham was responsible for setting corporate strategy and executing day to day business operations. Prior to joining Digium, Windham served as president and chief operating officer of ADTRAN, a global provider of networking and communications equipment. Windham joined ADTRAN in 1989 following ADTRAN's successful acquisition of Processing Telecom Technologies, a company Windham co-founded in 1986. Prior to becoming president/COO in 2005, Windham served as the senior vice president and general manager of the Enterprise Networks Division. Windham holds a Bachelor of Science degree in Electrical Engineering from Mississippi State University where he was named a Distinguished Engineering Fellow in 2001 and also holds an MBA from Florida Tech.

See All of Danny's Articles