It’s no secret: your organization needs to get better at data.
That’s a sweeping generalization, but I think it’s true. If you work at a nonprofit, or work with nonprofits, I bet you want to do more with data. I know that I do! (If you’re thinking “Nah, we’ve analyzed all our data and we’re done now,” please leave a comment so I can give you a virtual high-five.) Last year, NTEN put out a report on The State of Nonprofit Data that identified a lack of on-staff expertise as one big barrier to organizations making use of data. If you don’t have data experts on staff, then chances are, the best person to do that extra data analysis you’ve been dreaming of is… you.
Anyone can learn to do good analysis. Working with data doesn’t have to be hard, but it can be intimidating when you’re starting out. Fortunately, you don’t need a degree in statistics or the resources of the Obama campaign to start analyzing your data. All you need is the curiosity to ask the right questions… plus Excel or another spreadsheet software. If you’ve always wanted to learn data analysis, or if you feel like you need to learn it to do your job better, here’s how to start.
Before you analyze, gather your data.
Figure out what’s most important to your organization. Understanding your donors? Driving more advocacy? Measuring the performance of your emails? Whatever your priorities are, pick a handful of important metrics and start measuring them regularly, on a timeline that makes sense to you. Make this as easy for yourself as possible, so you’ll actually fit it into your busy schedule: write instructions, schedule reports to run automatically, and keep your reporting organized in one place.
At Food & Water Watch, we do high-level reporting every month to see how we’re doing. Is our fundraising on track? Did our email list grow like we expected? We have sections for list growth, fundraising, and advocacy with a few key metrics in each section, and the report looks something like this:
Get comfy with numbers
When we were kids in school, most of us were taught that math is hard and numbers are confusing. It’s not true, and you need to un-learn that lesson if it’s scaring you away from your data.
Spend time working with your data. Calculate numbers like open rates and conversion rates (the percent of your audience that does whatever you’re asking them to do). Get familiar with what’s “normal” for your organization, and what’s unusual.
A good analysis project is to develop internal benchmarks for your program. “Benchmark” is a fancy way of saying “average.” What’s your benchmark open rate? Click rate? Action/donation rate? Do these rates vary for different types of messages? Why might that be?
Remember, if you see something strange in an analysis, do a gut-check. It’s easy to make mistakes, especially in Excel. Could your calculations be wrong? Or have you discovered something wonderful and unexpected?
Dive in by asking good questions
Analysis is all about questions. You’re a detective trying to discover what’s hidden in your data. Start out with big questions about what you’re trying to achieve:
- What are the best sources of new supporters for our email list? Can we capitalize on those sources to grow our list even more?
- Are our activists more likely to become donors than other folks? What happens if we target them with more donation asks?
- Why were our online donations really high/low last month? What should we do about it?
The best questions are both specific and useful – your data will give you information to answer those questions, and will suggest insights to help you improve your programs.
Then, gather whatever data seems relevant to answering your big question, and start digging in. Ask more specific questions of your data: Who donated, and when? What donation form did they use? Where did they come from?
Slice and dice the numbers by whatever characteristics will help you answer your big question. You can segment by donor or activist status, compare performance across issues or timeframes, and look for trends over time. Always keep an eye out for outliers – unusually high or low numbers that might skew your averages.
Use what you learn to inform new ideas for your program… or better yet, verify your findings by running some A/B tests on your next campaign.
Don’t bite off more than you can chew
One of the biggest barriers to analyzing our data is time. We’re all busy, and it’s hard to carve out the time for even the most essential analysis. Whatever question you’re investigating, make sure it’s one that will be useful to you… and that’s doubly true if it’s going to take up a lot of time.
The simpler the analysis, the more likely you’ll be able to produce meaningful results. If you’re frustrated by the analysis you’re tackling, try to simplify it: look at a smaller set of data, ask a slightly different question, or ask a friend or colleague for help. Don’t bang your head against the wall.
Remember, this is a learning process. The more you work with your data, the more skills you’ll have to take on bigger challenges down the road.
The Force is strong with this one…
Once you’ve mastered the basics and you’re ready for more, here are some ways to grow your skills:
- Learn Excel, and learn it well. In particular, pivot tables and formulas will change your life. Find a class, or look for tutorials online. Learning a new trick in Excel could save you hours, even days of work down the road.
- It’s not enough to analyze your data. You need to be able to present it to others in a way that won’t make their heads spin. It’s easy to do this badly. Study how others are presenting data, and learn from them. Especially, learn from the mistakes [http://wtfviz.net/ ] – if you can explain why these charts are hilarious, you’re on your way to being a good data analyst.
- Read up. My favorite blog on data analysis is Avinash Kaushik’s Occam’s Razor [http://www.kaushik.net/avinash/]. He focuses on the business sector, but he’s brilliant and funny, and a lot of his lessons apply to nonprofits, too.
Got other tips on learning to work with data? Share them in the comments!