Customer: Should we do a POC?

Asked by a customer recently whether they should do a POC or not.

Here is my response:
  • Do in a POC the things that are most important for your business and your project
  • Do POCs only with the vendor who is in a position to deliver what you need. 
  • Only do a POC if you need to. Many major businesses find much better ways to evaluate their alternatives than doing a POC in their environment. 
  • A good POC should not be a trial run or test install or pilot. 
First of all, a POC is to show that there is business value in doing a project. Running a long evaluation of a product to check that every feature does what it is meant to won't get you to your business goal (unless you are in the business of running evaluations).
It is easy to fall into the trap of trying to create a test implementation or product QA lab and think you are doing some beneficial POC for your project.  However usually there is some key reason you are looking at a product.  If the product is able to show that it fulfills this key requirement then usually it is a successful POC.