Dirty Data Is Quietly Killing Your Nonprofit (And AI Isn't Coming to Save You)
- Michala Shanks

- Jul 22
- 4 min read
The first time I worked in a "real" nonprofit CRM, I thought my life was about to get so much easier. The sales pitch made it sound effortless. Everything was automated.

Everything was organized. Everything was going to save me hours. Lies.
Not because the software was bad. Most nonprofit CRMs are incredibly powerful. The problem was that nobody told me what came before the software could actually work.
Nobody explained mapping. Nobody explained field standards.
Nobody explained why searching for a donor before creating a new record would become muscle memory. Nobody explained that if your organization has been around for thirty or forty years, you're not inheriting a database. You're inheriting decades of habits from dozens of different people, each convinced their way was the right way.
And if nobody ever documented those habits? Good luck. I remember opening a database that had been passed from one staff member to another, then another, then another. Everyone had entered information a little differently. Some people abbreviated everything. Some people typed in all caps. Some linked businesses to people. Others didn't. Campaigns had different naming conventions depending on who created them. Notes lived in random places because everyone had their own system. The database wasn't broken. It was speaking six different languages. Then someone says, "We're migrating to a new CRM." Perfect. Except nobody tells you that moving dirty data into a brand-new platform doesn't magically make it clean. It just gives you a newer, shinier place to store the same problems. That's usually the moment someone casually asks, "Did you finish the field mapping?" The what? I still laugh thinking about the support representative asking if I'd watched the training videos. What training videos?
I genuinely thought I was just moving information from one place to another. I didn't realize every field had to be mapped intentionally, every relationship had to make sense, and every decision had the potential to affect years of donor history. That's not a software problem. That's an institutional knowledge problem. Every nonprofit has that one person who knows everything. They know Bobby Joe passed away three years ago.
They know the Smith Family Foundation should always be recognized under a different name. They know why a campaign code from 2016 still exists.
They know which spreadsheet everyone quietly ignores because the information was moved years ago. Until they retire. Or accept another job. Or burn out. Or simply leave.
If your organization depends on one person's memory to function, you don't actually have a database.
You have a really expensive memory.
And memories walk out the door every day. That's why documentation matters.
Not because policies are exciting. Not because SOPs look good in a binder.
Because documentation protects people. It protects your donors from receiving mail after they've passed away. It protects your volunteers from guessing how to enter information. It protects your staff from spending weeks trying to decode decisions made ten years ago. It protects your organization from losing relationships that took decades to build. I learned this long before I ever sat behind a desk.
I learned it as a volunteer. If you've ever worked registration at a gala, fundraiser, golf tournament, or community event, you already know what happens. There's a line out the door. Someone wants to buy another raffle ticket. Someone forgot to preregister. Someone insists they're already in the system. Someone hands you a business card. Someone else scribbles their email on a napkin. Everyone is doing their best.
And your volunteers are trying to keep the line moving while entering information as quickly as possible. Without a simple, documented process, everyone naturally develops their own way of doing things. Nobody's trying to create duplicates.
Nobody's trying to make your database harder to manage.
They're just trying to help. That's why good systems aren't about perfection. They're about consistency. Your volunteers shouldn't have to guess.
Your new development coordinator shouldn't have to guess. Your executive director shouldn't have to guess. If the process only exists inside someone's head, it doesn't really exist. And here's something I wish someone had told me years ago.
Don't underestimate simple tools.
Yes, there are incredible CRMs.
Yes, there are AI tools.
Yes, automation can save time.
But sometimes a well-organized Google Sheet with clear naming conventions, documented processes, and shared access is infinitely more valuable than expensive software nobody understands how to use.
Technology is only as good as the process behind it.
AI can identify possible duplicates.
It can suggest corrections.
It can automate repetitive tasks. But it cannot tell you that the donor who has faithfully supported your mission for twenty years prefers to be recognized by her maiden name, wants invitations sent to her home instead of her office, and would be deeply hurt if your organization forgot that.
That's still human work. And I hope it always will be.
Because these aren't just names on a spreadsheet.
These are people. They're the family that trusts your organization enough to give every month. They're the volunteer who shows up every Saturday because they believe in your mission. They're the donor whose planned gift may fund your programs long after they're gone. Clean data isn't about having prettier reports.
It's about stewardship. It's about honoring the people who make your mission possible.
If you're new to nonprofit leadership, fundraising, development, community engagement or whatever your organization calls the people responsible for keeping all of this together I made something I wish someone had handed me years ago.
The Nonprofit's Guide to Keeping Things Clean isn't a software manual.
It's the conversation I wish someone had sat me down for before I inherited my first messy database.
Inside you'll find practical SOPs, a database health check, cleanup strategies, migration guidance, and the kinds of lessons most of us learn the hard way.
Download it, share it with your team, and if you discover you're staring into thirty years of institutional memory disguised as a donor database...
I'd love to help you untangle it.
A Note from Michala
Thank you for spending a little time with me today.
If this guide helped, I'd love to hear from you. And if you ever need advice, support, or someone to help untangle a messy database, don't hesitate to reach out.
Happy cleaning!
— Michala



Comments