Support and Services Matter

Yesterday, a blog post I read from rPath CEO Billy Marshall piqued my interest. Marshall dismissed the idea that open source software firms can be successful on the strength of their service and support offerings–a point of view that runs counter to conventional wisdom that when you’re selling something that’s free, you’d better surround it with some attractive complementary pieces. Citing his previous experience leading sales at Red Hat, Marshall suggests an alternative to great support and service: selling your product really, really, really hard. In support of his statements, rPath’s chief points to Red Hat and Oracle financials, and how where Oracle spends 25% of its revenue on sales, general and administrative costs, Red Hat spends a whopping 47%–that’s compared to 15% on R&D and 18% on service and support. That’s enough for the numbers–Marshall doesn’t provide a citation for them, Red Hat and Oracle are very differently sized, and the support, R&D and SG&A figures he offers for each firm add up to 61% for Oracle and 80% for Red Hat, so there may be differences in the way the two companies sort their numbers. Now, Red Hat saw and continues to see an opportunity undercutting higher-priced Unix hardware/software combinations with a Linux and commodity hardware duo. They’ve thrown a lot of resources toward winning big Unix migration deals, and they’ve done well for themselves. This doesn’t mean, however, that Red Hat or other companies can’t succeed by selling support and services. As I see it, the world is grossly underserved with regards to IT, and major opportunities exist for firms that set out to gather freely available software components and shape them to serve the needs of individual companies. Marshall writes that support is a bad model because customers want software that doesn’t need support. However, simply directing your sales execs to excise mention of support from their spiels doesn’t make support unnecessary. Marshall also writes that “customers will pay a premium for great software, if they cannot get the same great software cheaper from somewhere else,” and then goes on to describe how at Red Hat, he boosted sales by convincing customers that they couldn’t get Red Hat Enterprise Linux anywhere else, and that support is irrelevant. Of course this isn’t true, as you can acquire RHEL sans support without paying Red Hat anything by downloading CentOS Linux, and you can now acquire RHEL with support from Oracle for less than what Red Hat charges. The difference for Red Hat is, and–if Red Hat is to remain on top–must continue to be… Support and Services. Hey, if you don’t need any help from your Linux vendor, why pay them a dime? If it doesn’t matter that your bug or feature request take precedence over open source’s upstream masses, again, why pay?