I don't really disagree, but a startup trying to do business with a small number of very large businesses can also cause problems. When your customer knows that the startup is dependent on their business (to some extent), they have a lot of leverage, and many big companies will naturally exploit that.
I've heard lots of stories (and experienced first hand) how a big company that is doing business with the startup will demand lots of outrageous things and cause the startup to expend lots of resources just to keep them happy. The trick is structuring the licensing agreement/contract so that if this happens, you're compensated for the additional time you have to spend.
Yes. This is actually something that we have run into with at least one prospect already.
I have a pretty easy response to unreasonable customer demands: "Some customers aren't customers that you want." My partner agrees and so far it's kept our head in the right place.
We also are targeting some of the smaller clients so that we can a) polish our offering b) get some "street cred" and c) just make sure that we can make money in this space :)
Absolutely: choosing the right customers is important. Both in terms of finding customers who are good from a revenue-per-effort ratio, and also customers who will help you develop your product in interesting ways (e.g. in ways that might be applicable to a broader market). Often, finding a great customer is an essential first step toward building a great product, because they can work with you to help you understand the sort of product they need (and by extension, the sort of product a broader class of attractive customers requires).
I have seen it both ways unfortunately, so that's one of the reasons that I am somewhat cautious (I guess that's what happens with 13+ years of experience in the field).
One that sticks in my mind is one of the first "real" jobs I had. The company was a solid, well established company that was about 20 years old. Best place I ever worked. They had a piece of supporting software that was written in VB on Access. The software was basic, but it worked as a great support to the methodology and theory of the company.
One big (REALLY BIG) client came in and needed the software to be housed on central servers, with the ability to roll the data up for reporting purposes.
So the company went off in the weeds, building a huge (2.5meg!) java applet, basically tailored to this one client's specifications. The company had no java experience. Had to hire java guys. No Solaris experience (as required by the client). Had to buy hardware and hire people. They had no oracle experience. Again, more $$ and more people.
Within 3 years the company went through two acquisitions, one of which was a reverse buy out of a crappy company that had IPOed. A year later the company folded.
A 20 year company was destroyed by this choice to build a bespoke system for a huge client.
Another memory is making said 2.5meg craplet run under the IBM JVM for OS/2 Warp 3. On token ring. Banks got sold some horrible systems!
I've heard lots of stories (and experienced first hand) how a big company that is doing business with the startup will demand lots of outrageous things and cause the startup to expend lots of resources just to keep them happy. The trick is structuring the licensing agreement/contract so that if this happens, you're compensated for the additional time you have to spend.