I have a habit that my friends find strange. I read nonprofit RFPs. Not for work, or at least it didn't start that way. Organizations publish them when they go looking for a new CRM or a case-management system, and buried in the boilerplate there is almost always a paragraph where someone finally says, in writing, what their week actually looks like.
One of them, from a food bank, mentions that client applications are collected in a survey tool and then "manually imported" into the system of record. Another, from a youth organization, says the whole point of the project is to get rid of "large spreadsheets being used by multiple departments." A third says staff have "adapted" to a system that can't be customized "by relying on external resources, such as Excel."
Adapted. That's the word I keep coming back to. Nobody chose this. The system couldn't do the thing, so a person started doing it, and then it became the person's job.
Here's what that looks like in a real week, and I'm describing it from the inside because I've watched it happen at a foundation we now work with. Somebody fills in an intake form on the website. A coordinator copies that into the case-management system. The same coordinator adds the person to a spreadsheet, because the case-management system has no way to show which cohort someone belongs to. When the person is enrolled, development finds out in a Tuesday meeting and adds a note to the donor database, because a funder wants a count. Finance asks for a headcount at month end and gets it from the spreadsheet, which is now a fortnight behind.
Every tool in that chain works. The organization has just never connected any of them, so the connection is a human being.
Shane, who runs the American Paragons Foundation, put it plainly when we first sat down with him. The work he actually cared about, getting athletes the support they needed, was maybe a fourth of his week. The rest was chasing athlete information across forms and spreadsheets and re-entering the same data into three tools. He wasn't exaggerating. We mapped it.
What makes this expensive is that it never appears anywhere. There's no line in a budget that says "moving information between systems by hand." It's smeared across the day of a program coordinator, a development associate, an office manager and the executive director, in pieces of ten or fifteen minutes. Each piece is too small to fix. Add them up over a year and you're looking at a salary. Sometimes more than one.
And time is only half of it. Every time a person carries a fact from one screen to another, the fact can change on the way. A date typed as the wrong month. A surname spelled two ways, so the same person exists twice. A status updated in one place and forgotten in the other, so the case system says active and the roster says gone since March. None of this causes a crisis. It causes something slower, which is that people stop trusting the numbers. When two systems disagree, someone has to decide which one is telling the truth, and the way they decide is usually to open the spreadsheet they maintain themselves. So the spreadsheet wins, and the fragmentation gets a little worse.
I used to think the fix was obvious: buy the bigger platform, the one that does everything. I've watched enough of those projects now to have changed my mind. The bigger platform has its own gaps. Within a year there's a spreadsheet growing around the edge of it, and it cost four times as much.
What actually works is duller than that. You keep the tools that do their job. You decide, for each kind of information, which system is the home for it: participants live in the case system, donors in the donor database, money in accounting. Then you build the connections so that when something is created or changed in its home, the other systems that need it get it without anyone touching it. And you give the team one screen that shows across all of it, so the cohort view that lived in Excel lives somewhere that's always current.
At American Paragons, that meant athlete intake, roster management, program enrollment and status tracking ended up in one place, talking to each other. Information gets entered once. The coordinators who used to be the thread between systems are still there. They just do the work the copying was eating.
If you want a place to start, take an hour with the people who handle information every day and draw, on paper, the path of one record from the moment it arrives to the moment it shows up in a report. Most organizations have never done this. Almost all of them are surprised. The drawing shows you where the same fact gets typed twice, where it sits waiting for someone to move it, and where two copies of it have quietly stopped agreeing.
That drawing is where every conversation we have starts. It's also worth doing even if you never call us.