Date: Wednesday, September 25, 2019
Category: Apra News
On Tuesday, September 24, Apra members joined Mirabai Auer, University of Chicago (@mirabaieliza), Jay Dillon, Alumni Identity Fundraising Consultants (@jayldillon), and Jacob Tolbert, Millikin University (@crazybilly) as they led a discussion around data analysis and dreaming big. This Twitter Chat (#aprachats) reached over 5,000 accounts with almost 80 tweets.
Some highlights from the conversation include:
- Where do you begin if you’re new to data science?
- Develop a good research question that has the potential to impact your work
- Don’t get intimidated by the tools!
- Excel can take you a long way - learn vlookup and pivot table skills
- You can usually answer your research question with the tool(s) of your choice
- What are some common errors that may derail the analysis of existing data?
- Two filters separated by OR operator and forgetting to wrap them up in parentheses
- Gaps in the data
- Focus on new and creative ways to derive and capture more of this data
- Data quality
- Alumni and constituent relationships to our institutions -- and giving in general -- are shifting and we need our data footprint to better reflect interests and motivations
- We lean so heavily on donor history, which is extremely important because that's what we have. However, if we could get at donors underlying motivations and connections those would be so powerful in modeling.
- It seems that more and more, past behavior doesn't necessarily predict future behavior. Clearly, donor history has to be a part, but I don't think it can continue to be the biggest part.
- How can we quantify the level of connection constituents feel for our school/cause?
- We are just getting around to starting to bring in open and click data and I think there is actually a lot to that and I feel it is still not something a lot of people are using or data that many people can access.
- One thing we've noticed is that connections are changing, fracturing, so folks are less connected to the institution as a whole, and more connected to affinity groups.
- Engagements scoring - no matter how you do it - should have the primary goal of surfacing previously unknown prospects. Be suspect of any engagement metric that simply highlights the donors you already know!
Search #aprachats to read more about what it takes to be a leader in data science. Stay tuned to @Apra_HQ for future Twitter Chats.