Open Sourcing Software in a Large Corporation?

code-libre asks: "I work for a small R&D group in a large corporation. We've spent the past few years developing a small but unique piece of software that was originally meant for internal use only. A VP recently approached us and asked if we could 'package and sell' the software so as to get a direct return on investment - at prices as much as $500k. Within our group, we worry about the costs associated with long term support and maintenance. We are also sure that a price over $10k is ludicrous, let alone $500k. I think it would be an excellent move to open source the software, but I need some advice..." "Even at a price of $10k, we don't expect to sell more than maybe 20-50 licenses. Costs associated with producing this software thus far are approaching $2mil, so we doubt our costs would be recouped. It is thus relatively easy to make the case that we _shouldn't_ sell the software.

On the other hand, it is software that will be vitally useful to those in the right markets. I would like to present the idea that releasing the software for free (and open source) will have two effects: one, branding the software turns it into a free piece of advertisement for our company in emerging markets. Two, open sourcing it will allow for others to help improve the software, which we in turn can use to our advantage - an indirect ROI.

Will points like this fly at a large corporation with little to no policy on giving stuff out for free? How can I convince an older generation of business leaders that FOSS is the way of the future? Ideally, I would like to help the company setup a internal group that could expedite small internal projects to the market place via FOSS routes. Any one have any experience with this?"

Comments

Re:While I support Open Source...

[alienw]

I think his VP is right: open sourcing it doesn't make any sense. In fact, for 90% of the software out there, open source is not the best model as far as ROI is concerned. Unless you have something you want to make into a standard or you have other ways to compete, you won't win anything by open sourcing stuff.

ROI - Thinking Backwards

[medcalf]

You guys are thinking backwards on ROI. The company has already spent the ~$2M - that's a sunk cost. Presumably, they have recouped, or will soon, the cost of this in internal business advantage. In other words, the costs to look at are the costs of selling and supporting it versus the income from selling licensing and/or support.

Without critical missing data, the ROI can only be guessed at. I'm supposing that the marketing types have already determined that number, from this point, is positive in a reasonable time period. Presumably, from the range of prices and size of the market as discussed, demand is price-inelastic, which makes sense if the business types are expecting a positive ROI from supporting the software. I cannot evaluate, either, the business impact on selling the product: if it provides a critical market advantage to the company (presumably it does, with that kind of R&D behind it), selling it may not make sense because of the loss of competitive advantage in a small market served by what (given the budget) must be large enterprises. I will simply assume that the business people have done their job, and that any loss would be smaller than the expected ROI gain.

Balanced against the net of ROI less the loss of competitive advantage, code_libre offers branding and getting back improvements. I suspect that the likelihood of getting the product improved by outside coders, given its niche character and the large amount of development effort already invested, is small. Thus the balance is net gain opposed to branding gains.

If the gain from branding can be quantified (or if a qualitative argument can be made that the company can be the single source of software, and thus related services and products, in a captive market), it is possible that the business would agree to open source the product.

But given just the information above, if I was asked to make the call, I would be disinclined to open source it. A captive niche market will often pay outrageous prices for software. Consider this: imagine if the software in question worked out when an airline should carry extra fuel, as opposed to when it should fly a leg with minimum safety margins, in order to take advantage of the cost differentials between fuel prices in different airports. Not a lot of customers, but each one could save millions of dollars a year (well, Delta did, and it can be extrapolated that others would, too). They would thus be likely willing to spend a million dollars to buy an enterprise license for the software: though the cost is huge, the benefit is even larger.

So from what I know, this would be a seriously uphill battle even if the company did have a robust history of open sourcing non-core software.

Re:stop thinking logically

[bradkittenbrink]

The poster is not thinking logically, for example:

"Even at a price of $10k, we don't expect to sell more than maybe 20-50 licenses. Costs associated with producing this software thus far are approaching $2mil, so we doubt our costs would be recouped. It is thus relatively easy to make the case that we _shouldn't_ sell the software.

The $2mil is a sunk cost. It is irrellevant to whether or not you should try to sell the software. The decision does not depend on how you can recoup all of your losses, it depends solely on how you can minimize your losses. That doesn't mean that selling closed source is the best way to minimize losses, but that's the way you have to think about the problem.