If your business sells to a quantifiable and identifiable group of business customers who are in the tens of thousands, or less, there's really no reason not to individually identify every one of them and track all of your contacts with them. There are plenty of examples of easy cloud apps to enable you to do just that really easily.
And while that seems like an "of course" sort of thing, I'm constantly surprised at how few entrepreneurial companies in growth mode (which depends almost entirely on excellent execution) do a good job of it.
If you only sell to half of the customers in your physical market, how many non-buyers have you called on? If you are using a sales force making customer visits, how much behavior is driven by territory size and quotas as opposed to market coverage? (If my territory is 100 square miles but I exceed my quota by covering about half of it, what motivates me to go further and do more?)
What do the non-buyers buy, and from whom? How often have the people who buy changed or moved? Any former customers from other places now working there?
If your penetration is 50%, do you spend additional resources to call on the remaining or is this a competitive share issue? Do you attempt to sell more to your 50% or acquire the remainder? How would you do both? Is there some way you've inadvertently left the door open for competitors?
And, of course, there are plenty of near real-time distribution statistics available if your capture system works. Which salespeople/partners/territories are the best? Which need help?
It's tempting as a sales manager to manage sales results by manipulating the rewards systems and leaving the intermediate work - number of calls, on whom, for what, and so forth - to the sales people. It's a bad idea.
The element of measurement enforces sales discipline and helps the company own its customer information and at least reduces the need for detailed rebuilding after each sales personnel transition,
If you are going to survey customers, you can use their buying or non-buying history to pre-segment the samples to yield different views of the data.
Venture growth is largely dependent upon excellent execution. Market math is one important ingredient.