In our rapidly evolving fundraising landscape, organizations often find themselves juggling multiple specialized tools for different aspects of their operations. But what if there was a better way?
This post explores why choosing one versatile tool can be more effective than maintaining a complex ecosystem of specialized solutions.
The Problem with Too Many Tools
My father once told me that he thought the huge tractors that he saw on TV were silly. “They have to buy this one machine to do just the planting and a different one to do the harvesting,” he said. “On our farm,” he went on, referring to his father’s potato farm. “We used one tractor but attached tools to it.”
Sometimes when I think about us fundraisers trying to find that one magic bullet that wrap up our prospect pool into a tidy bundle, I reflect on Dad’s comment.
We buy all kinds of new shiny objects that promise a lot but that don’t talk to each other, like screening services, auto-writing toolkits, nudging reminder systems, and added database services.
Why Python is the Modern Fundraiser’s Tractor
All of these tools work within the context of their designs, but we sometimes forget what problem we went shopping to fix because each tool is sending out different information in different formats.
What if we returned to basics? Or better yet, what if we found a tractor that could attach and use all of these parts?
That is why data scientists love statistics programs like SAS, R, and Python.
With these open-source programs, I can not only bring data in from a variety of different locations in a wide range for formats, but I can also manipulate it (like cleaning, fixing missing values, building new fields that show more information), and apply modeling, machine learning, and artificial intelligence tools to it.
Python is getting the best press (and best developers) currently. Python programmers are creating new packages that do all kinds of data manipulation and examination. Like Apple apps or Facebook apps, Python’s packages are built by developers everywhere, and so the selection for solving a problem is huge. For instance, there are a number of packages that calculate RFM on your customer base.
What Python can do is take data from, say, your online community, database, and screening product and bring them all together into one flat file for examination to help you find out what methods, ask amounts, and time of year (and day) work with which audiences.
Python can also be used to clean up data for import back to your database, or to pre-process data for your dashboards. Power BI also handles all of these tasks, yes, but not the modeling, which Python offers – from linear modeling to time series to neural networks to random forests. And Python can talk to Power BI.
The hard part, like my father might interject right here, is that once you have a single tool that is this flexible, you have a steeper learning curve. I liken it to the finished dollhouse vs. the Lego set conundrum. For example, my father had to learn how to drive a tractor after his father bought one and retired the horses; a farmer today has computers and air conditioning in the tractor.
The Power of Open-Source Flexibility
To get a feel for what I mean by Lego set, take a look at this Python code example:
One of my small business clients told me that, to her, a Python script looks like, “What a toddler would type while playing on a keyboard.” Yet it’s an excellent tool because it is so flexible in what it can do with data.
The key is to be willing to learn it and then enjoy the wonder of learning each package that gets your business intelligence/donor modeling to where you want them to go.
Like managing the subway in a large city for the first time, navigating Python can be daunting to learn, especially if the general coursework out there focuses on projects that don’t matter to us (one Python course I took taught me how to draw a turtle on my screen).
That’s why we created a nonprofit-centric beginner’s course for aspiring Python users. You can check it out by clicking this link.
Making the Choice: Multiple Tools vs. One Flexible Solution
What do you do if you haven’t learned one of these flexible tools but you still need to put data together from your other tools for analysis and for dashboards?
You can do it, but you will end up programming bridges between each pair, and the bridges have to be written in some programming language. For instance, Power Automate talks nicely to Raiser’s Edge, but Power Automate feels like programming a robot from scratch. Power BI creates dashboards and manipulates data, but you still have to do the statistics in some other package.
What do you think?
Do you have more than two devices in your garage to cut grass, weeds, and bushes with? Or do you just hire that out (by the way, we at Staupell are available for data science consulting!)?
Do you instead prefer one good tool even if it takes a little more effort to use it on your large lawn? If you, like me, prefer to know what’s going on behind the scenes as well as on the screen, however, take a look at one of these programming languages.
If you are thinking you want something else, let me know at [email protected]. You may be finding yourself inventing a new software of service.