build with your customers and shoot to be client funded.
@Casey has the simplest, clearest suggestion, and I agree. ("build with your customers and shoot to be client funded.") There needs to be a highly compelling reason to use a non-standard CRM product and become 100% dependent on a system that doesn't have popular development support. The amount of difference from available CRMs needs to be greater than 50% different.
The most important steps for your enterprise are going to be your marketing strategy, testing your assumptions, validating your product decisions (repeatedly), and building to suit based on tested FACTS, not ideas. Your product has to have clear personal benefits for users. Developing in tandem with actual future users is going to be essential to remain distinct. But also remember that your sample size needs to be big enough that you are getting a wide variety of input, not just a handful of similar-sounding thoughts.
An ERP and CRM are very time consuming, if you take 5 years to build it will take another 5 to rebuild when nobody is using the language you think today is the best choice.
So, if I could give you a suggestion, build a framework to build the CRM/ERP
At least is what we did here (after remaking the software 3 times we have learned the lesson :) ).