Think about this:
Let's say that to develop a good suite of flash memory file management software might require a man-year of effort, at a cost of $200,000. I might be high or low on that, so scale these number depending on your own assessment of the job.
Now, let's say you are Lucent, and you have a new chip for use in portable audio recording applications that will work with flash memory, and you need to get 300 development kits into the hands of potential customers as soon as possible. The sooner they have them, the sooner they will be able to design your chips into their products.
You have two choices:
1. Develop your own flash memory software, and incur a 6-month delay.
2. Pay a licensing fee of $10/unit to a third party for software that is already written.
The first choice will cost you $200,000 and a 6-month delay. The second will cost you $3000 and give you the software immediately. Which one would you choose?
Now, let's say you are a company just entering the personal audio player field. You already have competitors that you are playing catch-up with. You expect to ship 100,000 units in the first year, 500,000 in the second. You have 3 choices:
1. Develop your own flash memory software, and incur a 6-month delay. Cost: $200,000 for 600,000 units, or .33/unit. Your cost/unit in the first year is much higher, $2/unit, because of low volume of sales.
2. Pay a licensing fee of $5/unit to a third party for software that is already written. Cost: $3M.
3. Develop your own flash memory software, but use third-party software for the first 6 months of production (50,000 units), at a cost of $8/unit. Total cost = $200,000 + $400,000 = $600,000, or $1/unit.
Which of the above choices would you choose?
Now, let's say you project sales of 5M units total over several years. Let's say the third-party will license you the software for $2/units with this kind of commitment. Your costs are now (1) $200,000, or $0.04/unit (2) $10M or $2/unit (3) $600,000, or $0.12/unit.
Again, which one would you choose?
If I were Lucent, I'd choose (2). I need to get those 300 development kits out the door as soon as possible. As a bonus, it's cheaper to pay $10/each than to develop my own software.
If I were the personal audio player company, I'd choose (3). It's important to get to market as quickly as possible, so I am willing to pay a premium price for software that is available immediately. But I also must lower costs for the long haul, and so I will also immediately start a parallel development of my own software. The cost savings in doing so are huge - I can save 90% of the cost of the software by doing this.
There is a bit of a toss-up here, and some manufacturers may choose(1). There is other custom software that will have to be written anyway, and it will not spring forth over night. So, I could just add a couple of developers to the team, and have the flash memory software available at the time that everything else is done. However, there is also the risk that something might go wrong in the development process, and there could be a delay. If I want to insure maximum success, then, I would still go with (3), if for no other reason than as a hedge against problems in my own development effort. (Indeed, I would do this with as many components of the system as possible, not just the flash memory software.) I would make sure, though, that I have the ability to cancel the contract, perhaps paying some penalty, if my own development effort is sucessfully completed in time for product launch. |
