Articles & Resources

I was reminded of something very important recently: sometimes it is OK for data processes to be messy.

As a data practitioner, it is tempting to try to make every process/tool/system perfectly tidy. I want the things I build to work well and align with best practices. I also believe fiercely that the organizations I work with (no matter their size) deserve the best of the best.

But I also believe that data tools should support the work of people, not make it harder. This means we need to be flexible enough to meet people where they are. This can be messy -- and that's completely OK.

Data processes and systems can get tangled for a lot of different reasons. Here are a few of the most common ones I come across in my work:

1.     Some team members struggle with technology.  Let’s be honest, not everyone is comfortable with technical systems. Sometimes it is easier to adapt to the skill level of existing employees than to try to upskill the team. Adapting data processes this way can be very frustrating, but it is also very common. If this is your situation, you are not alone!

2.     You don’t have the bandwidth to learn a new tool. Change takes a lot of mental energy. Nonprofit teams work very hard, and don’t always have the capacity to tackle something new. It is OK if you decide to stick with an imperfect tool or process because your team just doesn’t have the energy to learn something new right now.

3.     You don’t own the data source. Small organizations usually do not have the resources to build their own systems, or even customize a complex tool (e.g. Salesforce). This means they may not control the types or quality of data they have access to. I experienced this SO many times as a program manager, and now I see with my clients almost every day.

4.     Goals and priorities are evolving. When goals are in flux, updating systems and processes can take a backseat. This can make data processes feel out of sync with your needs and cause frustration for team members. Even so, it usually makes sense to wait until your goals are clear to make significant updates to your data systems and processes. You will get there!

These are just a few of the causes of messy data processes that I have seen both as an employee and a consultant. These issues can be annoying, but they are often harmless in the short-term. That said, if your processes are broken enough that it is causing real problems for your organization, there are things you can do. Here are a few strategies that I have seen work.

1.     Get clear about what's (really) not working. Some issues are annoying but ultimately harmless. Make a list of all the issues you’re struggling with and sort them by impact to your organization. Be as specific as possible in identifying the most serious issues and their impact on the organization.

2.     Start with a specific goal and a clear win. Choose a clear, manageable first step to untangle your data systems. Keep this first step small and make sure it will have a meaningful, positive impact on your team. This will create momentum and reduce resistance when you decide to make bigger changes.

3.     Create a judgement-free zone.  Technology doesn’t come easily to everyone. Before implementing a new process, put a variety of supports in place to help team members learn in ways that work for them. Make sure that staff know someone is available to answer questions. The more confident and empowered you can make people feel, the less resistance you will face.

4.     Automate, automate, automate. If your messy processes are creating repetitive, manual work, automation can be a game-changer. The software platforms you use may already have built-in automation functionality. If not, consider using integrations or third-party connectors like Zapier or Make to automate repetitive tasks.

5.     When all else fails, declutter. If you aren’t sure where to start, look closely at your data systems. Are you collecting any data that you no longer need? Do your reports include fields that aren’t useful or don’t get updated? Once you get rid of the things that DON’T matter, it will be much easier to see how you can make what DOES matter work better. I have seen this approach work magic when systems were so tangled that finding a starting point felt impossible.

Last but not least, remember this: it is OK if your data processes aren’t perfect. As much as we want data processes to be neat and tidy, working with humans is messy. At the same time, if you get clear about what’s not working and taking thoughtful steps, you will make can make meaningful progress.