Thursday, July 27, 2006

OSCON - Business Models for Open Source Software Companies

Business Models for Open Source Software Companies by Tony Wasserman

I was hoping to get a little more out of this. I always have the expectation that when I hear an academic speak, they'd have all kinds of brilliant insights simply because they can claim they are an academic and get access to lots of great information that wouldn't be available to the press or to others in the industry. This presentation was nothing more than a general run down of the various open source business models out there.

Subscription Model
ala SugarCRM, Red Hat

Commercial and Open Source Products
ala Borland, CollabNet, SugarCRM

Support and Training Model
- Also, publications
Example: Many including O'Reilly
Works well for 1 / small man consulting shops and in develoing countries.

Dual License Model
Vendor offers same software under a commercial license who want or need commercial support.
Example: MySQL

Hosted Service
- Vendor uses open source software to create services that can be given away or sold to customers.
Examples: Yahoo, Google

Packaging Model
- Vendor integrates two or more open source products into a new product stack
- Vendor offers the "added value" stack along with support, training, and consulting
Examples: SpikeSource and OpenLogic

Commercial Enhancement- Vendor uses suitably licensed open source software and derives a new product that can be offered commercially, either as closed or open source.
Examples: EnterpriseDB and SRA OSS (PostgreSQL-based)

Consulting Strategy
- Vendor offers no open source software direcvtly, but helps clients make strategic decisions and investments related to open source
- Vendor obtains revenue by charging customers for their consulting time
Examples: IBM Global Services, Accenture, Gartner

Patronage Model
- Vendor offers open source software, money, equipment, or people's time to the community with no direct expectation of revenue.
- Vendor obtains revenue from other products and services, which may or may not be related to the open source software (and not even be open source).
Examples: Sun, IBM, Microsoft

Reseller Model
Offering free and open source software.

OSCON - Marketing to Dilbert: How to Invite Developers Into Your Project

Marketing to Dilbert: How to Invite Developers Into Your Project by Dave Rosenberg, Mulesource, Stephen O'Grady, Red Monk

Why Is OSS Marketing So Hard?
- Too much noise e.g., too many emails.
- Figure out how to engage audience.
- How to advertise without advertising?

These guys tried to get a discussion going around who does 1) a good job of marketing, 2) who does a decent job of branding but doesn't market well, and 3) who sucks?

1) Redhat, MySQL, SugarCRM, Firefox
2) JBoss, Novell, Sun

Stephen claims that "success of your application will be determined by how well you market to developers and IT staff, not CIOs." He claims this will transform the sales process.

Markets are conversations

- Market Research
- Customer Research
- Competitive Research
- Positioning
- Developer info
- Shared development

- Blogs
- Newsgroups
- Wikis
- Forums, IRC, etc.
- Events
- T-Shirts - 48% response rate when they ran a first 100 promotion.
- Contributions to projects
- Public Relations

Why to open source something?
- solve transparency issues
- solve product quality issues
- solve distribution quality issues (but other ways of solving this problem e.g., free software)

Best opportunities for open source is infrastructure software. Further you get out to the application space.

OSCON - Compiere - Building Successful Commercial Open Source Products

Building Successful Commercial Open Source Products by Jorg Janke

- Compiere is an Open Source ERP solution.
- Started in 1999 - apparently when they started, there were 15-20 open source ERP projects.. Profitable since 2003.
- ERP is a $50 - 70 billion market.

- 250+ customers.
- 100 partners worldwide in Eco-system - Compiere concentrates on enablement / 2nd level support / second level support / training. Partners concentrate on end user satisfaction, sales, tier 1 support.
- Funded - $6 million recently.

Business Models
- Claims business model is more about who your customer is than open / closed source.
- Open source (free) gets you in the door but that is about all you get for being open source. Market share.
- Repeatedly, it has been mentioned this week that lots will use open source product but very few will enhance it / contribute back to it. Lots of architects / advisors but few actual doers.
- Start of open source project - either create it yourself or solves a problem for a limited set of people, or you create it for other people's needs.

He distinguishes between 3 models and the "Commercial Open Source Contract"
- customer does own pre-sales
- pay for support / docs / services
- "if you need help, pay for it"

Where Open Source is Risky
"Open source business works best where software is a commodity and the revenue is in service, insurance, and support." - open source is risky where advantge is in the process / algorithm, where there is high risk, fast changing industry, e.g., payroll.

Compiere Process
- Compiere only supports the latest version.
- Offer easy, stable, proven automatic version migration Service.
- Accepted sponsored development if only on product roadmap
- Accepted development contributors if only very experienced.

Compiere Contribution Experiences
People contribute because:
- improve the product for their own needs
- reduce maintenance effort (core team maintains)
- gain experience. you coach me and I work for free.

People don't contribute because:
- gain competitive advantage
- sell enhancements to others
- keep changes confidential

Contribution LevelsUse and tell others about it
- Evangelists
- References
Use it and give feedback
- QA
Share exactly what they need
- Requriements - "most crucial contribution"
Develop it or help developing it
- Design... code
- code is only 10% of effort for business applications.

Pilot customer
Pilot customer was a complete pull. Released incomplete versions. Customers saw it. Attended trainings. Then sold support contracts.

Why raise funding?- Previous discussions with VCs were not good - product unproven, process unproven, partner model will not work.
- Went about validating it.
- It became a question of growing steadily and slowly or recognize window of opportunity. Appears they raised money for growth and to compete with big boys.
- Customer attraction. VC money validates company / model.