We’re often asked "so what can Filament actually do?" The short answer is SQL better than anyone you know.
If you’ve spent any time in Filament, you’ve probably seen Filly churn out a chart in seconds and wondered what’s actually happening behind the curtain. The answer is SQL (Structured Query Language). It’s the industry standard for talking to databases, and it’s what Filly uses to turn rows upon rows of data into the pretty dashboards you use for decision making.
When we’re speaking to potential customers, we’re often asked so what can Filament actually do? The short answer is: anything you can do with SQL, but we recognise that’s somewhat meaningless unless you know how to use SQL yourself. So we’re going to do a little statistics 101 here, because if you know the extent of Filly’s capabilities (just like anyone else in your team) you’ll be able to maximise their contributions.
There’s two main types of statistics: telling you what already happened (Descriptive Statistics) and helping you guess what might happen next (Inferential Statistics). One is based on facts, the other is an educated guess.
Most of the time you don’t need a crystal ball; you need to know what has happened, and be able to report on it in multiple ways. Your growth, your impact, your financial position. This is where SQL is the undisputed heavyweight champion. It takes thousands of messy rows; from Salesforce entries, Mailchimp clicks, and random spreadsheet cells, and performs "aggregations" (and can do it at a scale that would make Excel meltdown and give you the spinny wheel of death).
Just like Excel handles your basics (when it isn't crashing), and just like Apple Notes now tries to solve your math every time you type an equal sign, Filly has all your fundamentals covered off natively. SQL lets us fly through the standard stuff; means, medians, and modes, at lightning speed. While a single massive $50k grant can skew mean and make your donor base look wealthier than it is, Filly can instantly pivot to find the median, showing you what a typical supporter actually looks like. It also handles the measures of spread, like Standard Deviation, which tells you if your revenue is steady and predictable or if it’s wildly volatile. Instead of just seeing a total, you’re seeing the health and stability of your fundraising.
But the real magic of SQL is in joining, grouping, and filtering. It allows Filly to "slice" your data into geographic or demographic buckets instantly. Want to see which postcodes are driving the most volunteer sign-ups? Or which specific campaigns are attracting recurring donors versus one-off givers? SQL does this by sorting through the noise to find the "peak" frequency (the mode). This is the foundation of every Filament dashboard; it takes a giant pile of raw information and organizes it into a comprehensive scorecard for your mission. It enables you to see straight up exactly what you've achieved.
I’m hesitant to use the word predicting here. As a data purist it feels a little misleading. Inferential statistics takes a slice of your data and uses it to make an educated guess about a larger group, or a future outcome.
Using SQL, one of the ways Filly can help you do this is looking for correlations. For example: "Does spending more on Facebook ads result in more recurring donors?” (Or are we just lighting money on fire by sending it to the billionaires who should be promoting causes like ours for free).
One of the main ways we handle this in SQL is through linear regression. Don't let the name scare you, it’s just a way of drawing a straight line through a bunch of messy data points to see which way it’s trending. While simplistic, linear regression is an easy, effective way to identify what direction things are going in: are donations increasing? Is churn decreasing? More complex regression analysis is realistically beyond the norm for SQL, but we’re looking at how we can tackle this by bringing in a Python connection (soon, hopefully!).
If Filly shows you a trend line, it’s making an inference. It’s saying, "Based on the last six months of data (facts), we can infer that next month will look like X." It’s not a 100% guarantee (because, again, the world is chaotic), but it’s better than just vibing your way through a budget meeting.
When you’re using Filament, understanding the broad capabilities of SQL will help you know what to ask, and how to make the most out of the tool. The biggest hurdle (especially in a sector that’s often run on passion alone) is to break away from impact metrics that you’ve sort of made up, towards documented methodology so Filly can take over that work for you.