Design the sidewalks last
You can't always predict how people will use your systems.
Do you remember your college campus, or the last big office campus you visited?
If it was like most such places, it had dozens of buildings connected by concrete sidewalks — where the planners wanted people to walk — criss-crossed with trails worn through the grass — where people actually walk.
This system has two problems:
Somebody spent a lot of money building sidewalks that nobody uses.
The dirt paths are hard to maintain. Muddy in the rain, neglected under heavy snow.
Campus personnel may try to force people onto the “correct” paths by erecting barriers. But the people just make slightly longer shortcuts by wearing a new trail around the barriers.
Probably not the result they were hoping for.
There's a better way:
In recent decades some landscape designers have caught on and decided to stop fighting the inevitable.
How?
They design the sidewalks last.
It's a simple concept:
You first build the buildings, and put in few or no paved pathways.
Then, in the second year when the lines of desire and convenience and efficiency have worn themselves into the grass, you pave them.
Sure, the campus is actually less convenient in the first year: it has no paved pathways at all.
But after the second year, the pathways are both convenient and maintainable. And you haven't wasted your resources building sidewalks that no one uses.
Here's the thing:
In any software system, including your open-source CRM, you have the same opportunity.
You can do it by implementing the human system first, with whatever ad-hoc data sharing methods are easiest. Shared spreadsheets, ad-hoc emails, whatever.
Work out the policies that fit your mission, and let this human system run its course for a few months.
By that time you'll see what shortcuts people have begun to take. You'll know where the big inefficiencies are, and where a well-designed CRM feature set can really help you.
No system will ever be “perfect.” You can save yourself some heartache by starting with what matters most — the human system — and investing in the software later.
All the best,
A.