You’ve been reading our past posts and believe in the power of data. You know where your organization’s data is housed. You’ve been brushing up on your Excel skills and have cleared your calendar in heady anticipation of this moment—You are going to analyze some data.
You open the spreadsheet and a new world unfolds before your eyes. You drink in the intoxicating mix of names, dates, columns, codes, open field text responses…everything you should need to answer that pressing business question for your team.
So umm…now what?
There are many ways you can slice subsets and compare factors across even a simple dataset, and this sheer magnitude of combinations and directions can overwhelm even a savvy data analyst. In our office, we refer to this frozen-by-possibilities status as “Analysis Paralysis.” Here are some tips to keep you from getting lost in the data and to ensure your analysis continues moving in the right direction.
Perform a Gut Check: It’s tempting to cannonball into the data and immediately try to answer your highest-order question:
- “Was our Group Attendance Package successful last year?”
- “What is our member retention rate?”
However, this approach to analysis can result in overlooking serious deficiencies or errors in your data and can leave you feeling technically overwhelmed.
Instead of starting with the most complicated question, get to know your data by calculating some gut-check metrics:
- How many total registrants did we have last year?
- How many members do we have currently?
Working through these base metrics builds your confidence in your analysis skills, allows you to spot omissions in your dataset, and can even reshape your higher-order questions.
For example, you might notice that only 10% of Group registrants provide an email address at point-of-sale, versus 100% for other types of registrants. Your business question could evolve from “Was the package successful” to “Was the package’s success worth missing out on 9 of 10 marketing leads?”
Keep Your Eye on the Prize: It’s easy to become a slave to permutations. With all the ways you can dice data, how do you know when enough is enough? Before you spend your lunch hour calculating the ROI of event attendees with last names that start with K who registered for the event on Tuesdays, take a step back.
Referring back to the business questions at the heart of your analysis can help you remember the bigger picture and avoid data analysis rabbit holes. Picture yourself presenting the analysis in a conference room: It’s highly unlikely that your association’s senior team will be won over after you subject them to a 15-minute explanation on how you sliced and diced the data on each slide. They want clear answers based on reasonable assumptions, and your analysis should follow the same principles.
Anticipate Follow-on Questions: This might seem to contradict the last point, where we told you to analyze less, but hear me out. You can reasonably assume that some pieces of your analysis will lead directly into other questions and analyses.
If you find that your organization is losing 100 members per year, the second-order question is probably “What type of members are we losing?” Losing 100 student-level members is very different from losing 100 of your largest-tier corporate members.
If you anticipate that someone listening to your presentation will probably ask a follow-on question or want a piece of information before making a business decision, then provide the answer in your analysis. Building your analysis along logical lines helps you stay on track and provides structure to your final product. And who doesn’t like responding to a mid-presentation question with the omniscient:
“Great question! I’ve addressed it right here on the next slide.”