Open Source Business Research at OWF 2010

Update, 2010-11-05: If you like this blog post, you might also like my artikel on the single-vendor com­mer­cial open source busi­ness mod­el.

This after­noon, I’ll be pre­sent­ing my thoughts on the cur­rent state of open source busi­ness research and future direc­tions at the Open­World­Fo­rum 2010 in Paris. I have sum­ma­rized the­se thoughts in this blog entry, and they are aligned with the pre­sen­ta­tion I’ll be giv­ing. I should add that busi­ness research here means aca­d­e­mic busi­ness strat­e­gy and eco­nom­ics research, to the extent that a com­put­er sci­en­tist can relate to it, and that most of my research is actu­al­ly still tra­di­tion­al soft­ware engi­neer­ing research.

INTRODUCTION

To sur­vey the research space, I’m break­ing it up along the lines of involved actors or key roles the­se actors play. I see three main types of actors:

  • Pro­duc­ers
  • User/customers
  • Labor­ers

Soft­ware devel­op­ers are part of the pro­duc­ers cat­e­go­ry as long as it is vol­un­teer work; they are part of the labor­er cat­e­go­ry when it comes to non-self-determined work.

PRODUCERS

Let’s first turn to pro­duc­ers, that is, the soft­ware devel­op­ers, ven­dors, and dis­trib­u­tors who make much of open source hap­pen. A sim­ple frame­work to char­ac­ter­ize the­se pro­duc­ers asks the ques­tions of who, what, how, and why:

  • Who: vol­un­teer or cor­po­rate (employed) work
  • What: com­mer­cial or non-commercial soft­ware
  • How: com­mu­ni­ty owned or pro­pri­etary
  • Why: out of altru­ism or for-profit

The “what” dimen­sion tries to cap­ture whether there is sig­nif­i­cant rev­enue poten­tial in the soft­ware being pro­duced, which typ­i­cal­ly implies that com­mer­cial par­ties will join the table. In the “why” dimen­sion, “altru­is­tic” is a catch-all for any moti­va­tion that is not strict­ly for prof­it, for exam­ple, fun or altru­ism.

In the­o­ry, most com­bi­na­tions are pos­si­ble, but in real­i­ty we observe three main types of open source soft­ware projects and the respec­tive actors behind them:

  • tra­di­tion­al open source = {vol­un­teer, *, com­mu­ni­ty, altru­ism}
  • open source dis­trib­u­tors = {cor­po­rate, com­mer­cial, com­mu­ni­ty, for-profit}
  • single-vendor open source = {cor­po­rate, com­mer­cial, pro­pri­etary, for-profit}

Traditional Open Source

Tra­di­tion­al open source is the orig­i­nal open source. It is not a busi­ness mod­el, how­ev­er, it is a mod­el of open source soft­ware devel­op­ment. Research here most­ly address­es ques­tions like those for vol­un­teer moti­va­tion, e.g. fun or altru­ism or Lern­er and Tirole’s sig­nal­ing hypoth­e­sis [6] (which has recent­ly been called into ques­tion by Bitzer et al. [2]). Also, best prac­tices of open col­lab­o­ra­tion are active sub­ject of study, given that no sin­gle per­son can enforce par­tic­u­lar behav­ior and that many par­ties desired at the table are vol­un­teers [12].

Open Source Distributors

A tra­di­tion­al project with com­mer­cial poten­tial soon­er or lat­er will be picked up by a ser­vice com­pa­ny and/or dis­trib­u­tor to provide ser­vices that allow for effi­cient oper­a­tion of the project inside a cus­tomer enter­prise. Hard eco­nom­ic ques­tions faced by such com­pa­nies are how much to con­tribute to the project to keep it mov­ing for­ward, as well as how much inno­va­tion to con­tribute or poten­tial­ly keep pri­vate for com­pet­i­tive advan­tage. I’m not aware of any eco­nom­ic mod­els that can han­dle such multi-player sit­u­a­tions. The well-known tragedy of the com­mons prob­lem plays out here, as do ques­tions of pro­duct strat­e­gy, for exam­ple, forced ear­ly reveal­ing of inno­va­tion that would prefer­ably have been kept pri­vate.

To reg­u­late issues of project gov­er­nance, col­lab­o­ra­tive inno­va­tion, and intel­lec­tu­al prop­er­ty, most of the­se projects will even­tu­al­ly move under the umbrel­la of an open source foun­da­tion [8] [10]. Capra and Wasser­man call this man­aged open source [3], I’ve called it sec­ond gen­er­a­tion com­mu­ni­ty open source [11]. Here, best prac­tices of foun­da­tion gov­er­nance are research top­ics. This includes prac­tices of project and com­mu­ni­ty man­age­ment, con­flict res­o­lu­tion, intel­lec­tu­al prop­er­ty han­dling, etc. Also of inter­est is path depen­dence, that is the (ini­tial) con­di­tions that let a tra­di­tion­al open source project grow into a suc­cess­ful man­aged com­mu­ni­ty open source project or that make it fail.

Single-Vendor Open Source

Single-vendor open source firms are soft­ware ven­dors who main­tain con­trol over the open source project and extract most of the val­ue using strate­gies like dual-license or open-core approach­es [13] [9]. Given that exert­ed con­trol typ­i­cal­ly hin­ders com­mu­ni­ty growth, one research ques­tion is how much con­trol to main­tain and how much to relin­quish for what form of return from a user com­mu­ni­ty. What exact­ly are the levers of con­trol?

The best prac­tices of single-vendor open source firms are not well under­stood and are an active sub­ject of research. Tony Wasser­man and I have been con­duct­ing inter­views to elu­ci­date exact­ly those best prac­tices that front-running com­pa­nies in this space are employ­ing. Our goal is to under­stand the best prac­tices by busi­ness func­tion as well as by cross-cutting busi­ness process­es and how they are supe­ri­or to tra­di­tion­al closed source ven­dors. Finan­cial bench­mark­ing is part of this research and our goal is to val­i­date whether long held claims, for exam­ple, of low­er sales and mar­ket­ing or low­er research and devel­op­ment costs are true.

It is also an open research ques­tions whether tra­di­tion­al soft­ware ven­dors will be able to remain com­pet­i­tive, whether they can adopt open source best prac­tices with­out becom­ing single-vendor open source firms, or, if they prefer, how they can tran­si­tion to an open source busi­ness mod­el with­out los­ing their shirt.

USER/CUSTOMERS

Exist­ing research work already inves­ti­gates how open source has made it into user com­pa­nies [5], though we can still improve our under­stand­ing of open source adop­tion and tech­nol­o­gy dif­fu­sion. For user com­pa­nies engaged with open source, the ques­tion not only becomes whether to use it, but also whether to con­tribute. Should it spend resources on the project? Should it gath­er influ­ence and for what ben­e­fit?

Open col­lab­o­ra­tive inno­va­tion is a major research top­ic of how users can inno­vate to their own and a community’s ben­e­fit [1] [4], but of course it also rais­es ques­tions of com­pet­i­tive gain and loss. From a producer’s per­spec­tive, the ques­tion becomes what intel­lec­tu­al prop­er­ty prac­tices to put in place and how to set up an open source soft­ware project, includ­ing its tech­ni­cal archi­tec­ture, to allow for best pos­si­ble user inno­va­tion.

In recent years, we have seen users not only adopt open source, but active­ly cre­ate and dri­ve projects for­ward. The tra­di­tion­al orga­ni­za­tion­al form of a con­sor­tium allows like-minded user firms to col­lab­o­rate on and facil­i­tate and finance the devel­op­ment of com­mu­ni­ty open source soft­ware. As a vari­ant of the open source foun­da­tions dis­cussed ear­lier, such user con­sor­tia need to answer ques­tions of project, com­mu­ni­ty, and intel­lec­tu­al prop­er­ty gov­er­nance. How open or closed should such a con­sor­tium be to its non-members? And also, again, how to make the eco­nom­ic case for par­tic­i­pa­tion given the threat of a tragedy of the com­mons sit­u­a­tion?

LABORERS

Open source is com­modi­tiz­ing soft­ware [14]. Sig­nif­i­cant bar­ri­ers exist to learn a closed source com­po­nent (license fees, edu­ca­tion fees), while lit­tle or no bar­ri­ers exist to learn­ing an open source com­po­nent. Thus, cheap labor can more effec­tive­ly com­pete with high-priced labor, and the usu­al mech­a­nisms will play out. This is per­haps where pub­lic pol­i­cy research is most like­ly to become impor­tant, as may be nation­al secu­ri­ty con­cerns.

Already today can we observe that a new class of open source soft­ware devel­op­ers, com­mit­ters to high-profile projects, are appro­pri­at­ing some of the val­ue that the soft­ware they are devel­op­ing is cre­at­ing [7]. The­se com­mit­ters are gain­ing more inde­pen­dence from employ­ers as their eco­nom­ic val­ue lies increas­ing­ly in them­selves as an agent of their own inter­ests. Whether this leads to a two-class soci­ety of com­mit­ters and less self-determined labor­ers remains to be seen and I view it as an inter­est­ing labor eco­nom­ics research top­ic.

INDUSTRY

One of the enjoy­able things about open source busi­ness research is that the­o­ry and prac­tice, that aca­d­e­mics and prac­ti­tion­ers, can be close­ly relat­ed. I fol­low the writ­ings of well-known ana­lysts, con­sul­tants, jour­nal­ists, and indus­try thought lead­ers like Car­lo Daf­fara, Rober­to Gal­lop­ini, Matt Aslett of 451 Group as well as the Red Monk ana­lysts and Olliance Group, Glyn Moody, Matt Asay and many more (less fre­quent) writ­ers. To the aspir­ing Ph.D. stu­dent in search of a top­ic I rec­om­mend to fol­low them as well.

CONCLUSIONS

Excit­ing times! I see a lot of busi­ness and eco­nom­ics research ahead. As a sum­ma­ry of this blog entry, you can down­load the OWF 2010 Slide Pre­sen­ta­tion as a PDF. As men­tioned, it is nei­ther com­plete nor ful­ly con­sis­tent but I believe it should serve well as an idea-generating sur­vey for future research.

REFERENCES

[1] Bald­win, C., von Hip­pel, E. Mod­el­ing a Par­a­digm Shift: From Pro­duc­er Inno­va­tion to User and Open Col­lab­o­ra­tive Inno­va­tion. HBS Work­ing Paper 10–138/MIT Sloan Work­ing Paper 4764–09, 2009.

[2] Bitzer, J., Geisheck­er, I., Schröder, P. Returns to Open Source Soft­ware Engage­ment: An Empir­i­cal Test of the Sig­nal­ing Hypoth­e­sis. Wis­senschaftliche Diskus­sion­spa­piere V-321–10, Uni­ver­si­ty of Old­en­burg, 2010.

[3] Capra, E., Wasser­man, A. A Frame­work for Eval­u­at­ing Man­age­ri­al Styles in Open Source Projects. In Pro­ceed­ings of OSS 2008. Springer Ver­lag, 2008.

[4] Ches­brough, H. Open Inno­va­tion. Har­vard Busi­ness School Press, 2003.

[5] Dedrick J., West, J. Why Firms Adopt Open Source Plat­forms: A Ground­ed The­o­ry of Inno­va­tion and Stan­dards Adop­tion. MISQ Work­shop on Stan­dards Mak­ing, 2003.

[6] Lern­er, J, Tirole, J. Some Sim­ple Eco­nom­ics of Open Source. Jour­nal of Indus­tri­al Eco­nom­ics, June 2002.

[7] Hann, I­. et al. “Why Do Devel­op­ers Con­tribute to Open Source Projects?” In Pro­ceed­ings of Sec­ond Open Source Soft­ware Work­shop, 2002.

[8] O’Mahony, S. Non-Profit Foun­da­tions and their Role in Community-Firm Soft­ware Col­lab­o­ra­tion, 2003.

[9] Olson, M. “Dual Licens­ing.” In Open Sources 2.0, Chap­ter 5. O’Reilly, 2005.

[10] Riehle, D. “The Eco­nom­ic Case for Open Source Foun­da­tions.” IEEE Com­put­er vol. 43, no. 1 (Jan­u­ary 2010). Page 86–90.

[11] Riehle, D. “The Eco­nom­ic Moti­va­tion of Open Source: Stake­hold­er Per­spec­tives.” IEEE Com­put­er vol. 40, no. 4 (April 2007). Page 25–32.

[12] Riehle, D., Ellen­berg­er, J., Mena­hem, T., Mikhailovski, B., Natch­etoi, Y., Naveh, B., Oden­wald, T. “Open Col­lab­o­ra­tion with­in Cor­po­ra­tions Using Soft­ware Forges.” IEEE Soft­ware vol. 26, no. 2 (March/April 2009). Page 52–58.

[13] Riehle, D. “The Com­mer­cial Open Source Busi­ness Mod­el.” Infor­ma­tion Sys­tems and e-Business Man­age­ment vol. 8, no. 4. Springer Ver­lag, 2010. Forth­com­ing.

[14]  Riehle, D. “Open Source Devel­op­er Careers.” Linux-Tag, 2010.