Things are tough out there.
Nonprofits trying to use their data wisely are in a tough spot right now. Budgets are shrinking. Expectations are rising because so many people see AI as a way to magically solve problems. New tools crop up constantly that would be great to experiment with, but those experiments are challenging to fold into the existing tech stack. So many organizations feel like they’ve got increasing data problems, and even getting your arms around them can be a challenge nobody has time or budget for.
There is no way to magically solve these problems, but there are some things we can do to make the situation a little better. A big one is just to make sure that everyone in the organization is on the same page about data. This requires regular communication with people about the current state of things, what’s working, what isn’t, and what the plan is. For example:
Make sure everyone who generates data realizes that’s what they’re doing! If you asked these people at your organization where the data they generate goes, would you get a blank stare?
- An event organizer
- Someone who creates a widget on the website to gather information
- Someone who agrees to receive donations through a new processor – like DAF Pay or PayPal or Venmo
- Someone who creates a new newsletter
- Someone who agrees to collaborate with another organization for partner memberships
- Someone who orders merch
- Someone who sends direct mail
The easiest way to think about this is to restrict people to keep the tech stack consistent. That stifles innovation and encourages people to sneak around.
The harder way is to require people to communicate about anything that generates data. If you’re an organization that encourages experimentation, you might have a policy that people can try something new, but they have to communicate about the data they’re generating and have reasonable expectations about how that data will be integrated into the tech stack.
For instance, if the events team emails everyone to say, “We’re doing a brand new kind of event, the RSVP form goes live tomorrow. We’d like to use this data alongisde our member data in the future, but we get that’s going to take time. Let’s talk about the future of this data. Here’s a shared doc with info about what we’re doing.” That’s both communicating what’s happening and having reasonable expectations. If the expectation is that people will scramble and try to integrate that data overnight, you’re just asking for stuff that breaks and burned out staff. If the events team just doesn’t tell anyone what they’re doing and later people figure out that this data lives outside the usual tech stack, you didn’t get the chance to decide how to treat this data, and you didn’t get the chance to give the events team useful information like, all of our other systems want first name and last name separated, can you please do that on your events form instead of full name? Giving that info to them early can mean the difference between cleaning up 25 rows of data and 25,000 rows. Getting this data onto the list of known data sources helps everyone understand the complexity you’re dealing with and make plans for the next few months.
Write down the problems people have with data in one place and keep it updated
This sounds really simple, but most organizations don’t do it. If you don’t fully understand your problems, you won’t design the right solution. This can’t be just the problems the tech team has identified with the data, it needs to include everyone in the organization. Lots of people have problems they don’t realize are data problems, and so asking the question as broadly as you can really helps. You’ll also have to ask at least every six months, because things change!
Regularly review your problems to understand the cost and the source
I don’t know any organization that is fixing every problem they have around data, and that doesn’t need to be the goal of these reviews. When you’ve got reports from people of the problems they’re experiencing, you can figure out where the problem originates. If all your problems are originating from one place, that’s a sign it is time to invest time and money in that part of the tech stack. It is also really important to understand how much the problem is costing you. If most of your problems are in one part of the tech stack but the problems that are the most costly – the ones that do the most to prevent you from reaching your goals is in another, invest in fixing the piece that is most costly for the organization. Having the problems and their cost and source written down also means you can share with people internally why you’re not fixing something that bothers them, as well as help make the case for investment.
Do these things sound impossible?
These three things are not a small task. The bigger and more dynamic the org, the more time these things take. These things don’t have to be huge time sinks, but they do take real time and attention. If getting these three things done and keeping up with them sounds impossible for your organization , then it is probably time to recognize that you may not be staffed to use your data wisely right now. You may need to make a plan to keep your head above water until you can be staffed to handle things like this. These things aren’t a job you hire for, it is a way the people who do all the jobs collaborate, and collaboration takes time. Technology requires an investment of money, but without the investment of time, it won’t ever be as successful as it could be.
Artisanal Data 