(ok with the pre-amble that I am VERY pro-OSS, and generally very bullish on the value for companies of every size to embrace an OSS strategy...)
I think a resource-constrained startup (pre-funded thru to series A) needs to be careful that every dollar spent is building direct value to the company that investors will want to invest in.
Creating good OSS does have an over-head (documentation, support, community collaboration, etc). Sure that can help you attract good developers to come work for you - but if you launch too many OSS project the overhead costs can become disproportionate to the funding of the company.
You also need to be careful that while you are building value early-on that you are not exposing yourself for your competitors (or your own employees) to fork what you are doing and build their own competitive service (it does happen but again, some of this comes down to license used).
I would not want to publish 'secret sauce' of a company (algorithms, development strategies, etc) before I got to a point where I could defend the business with network effect, traction, funding, etc.
There is no point of over-committing or over-exposing and then not having the resources to be around in 18 months time to keep the OSS project going.
I'm not going to name the company, but another extreme example comes to mind was of an early stage company in a specific vertical who completely funded a relatively high-profile (read: well hyped) open source project that was orthogonal but not completely related to the core competencies and interests of the parent company.
I would have been strongly against this if I was an advisor or investor in the company.
Suffice to say, things ended badly when the company ran out of money and now both their project and the open source project are dead in the water, + lots of hurt feelings all around.
I think a resource-constrained startup (pre-funded thru to series A) needs to be careful that every dollar spent is building direct value to the company that investors will want to invest in.
Creating good OSS does have an over-head (documentation, support, community collaboration, etc). Sure that can help you attract good developers to come work for you - but if you launch too many OSS project the overhead costs can become disproportionate to the funding of the company.
You also need to be careful that while you are building value early-on that you are not exposing yourself for your competitors (or your own employees) to fork what you are doing and build their own competitive service (it does happen but again, some of this comes down to license used).
I would not want to publish 'secret sauce' of a company (algorithms, development strategies, etc) before I got to a point where I could defend the business with network effect, traction, funding, etc.
There is no point of over-committing or over-exposing and then not having the resources to be around in 18 months time to keep the OSS project going.