How It Works
The chat interface is your home base. Think of it as having a conversation with a data analyst who already knows every CRM, payment platform, and spreadsheet you’ve connected — and never goes on leave.
Open Dashboards in the sidebar and ask a question in plain English. Filly works out which of your tables matter, writes the SQL, runs it against your warehouse, and returns an explanation alongside a chart. A new dashboard is saved automatically, so you can come back to it, share it with your team, or keep building from there.
Ask Better Questions
Filly is a world-class SQL engineer, but a terrible mind-reader. “How are our donations looking?” returns a generic total. The fastest way to get something you can actually use is to be specific. Most prompts that work follow this shape:
“Give me the [metric] grouped by [dimension] filtered by [attribute] for [timeframe] shown as a [chart type].”
- Metric: the number you care about — total revenue, average gift, donor retention rate, donation page conversion rate.
- Dimension: how you want it sliced — by campaign, state, channel, month.
- Attribute: the filter — only active members, only first-time donors, only campaigns tagged “EOFY”.
- Timeframe: be explicit — “last 90 days”, “FY25”, “Q2 this year vs Q2 last year”. “Recent” means nothing to Filly.
- Chart type: line, bar, pie, KPI card — or just let Filly choose if you don’t know yet.
“Show me average gift size by state for active recurring donors in FY25 as a bar chart” gives you something you can put in front of your board. The more specific the brief, the less guessing Filly has to do.
Example Questions
Fundraising health
- “Total revenue this financial year, broken down by campaign.”
- “Average gift size by acquisition channel for new donors in the last 12 months.”
- “Compare one-off gifts vs recurring gifts as a share of total revenue, monthly.”
- “Top 10 donors by lifetime value, with their most recent gift date.”
Donor retention & lifetime value
- “What’s our donor retention rate year-over-year?”
- “Show me lifetime value by acquisition source for donors acquired in 2024.”
- “Which donor segments lapsed after a single gift in the last 18 months?”
- “Recurring donor churn rate by month, last two years.”
Acquisition & conversion
- “Donation page conversion rate by traffic source, last 30 days.”
- “Compare Meta Ads spend to new recurring donors acquired, monthly.”
- “Which email campaigns drove the most first-time gifts in the last quarter?”
- “Donor acquisition cost by channel for FY25.”
Programs, events & volunteers
- “New volunteer sign-ups by source channel for the environment program, Q4.”
- “Event attendance by postcode for our 2025 gala series.”
- “Grant pipeline value by stage and funder.”
- “Program participants by month, split by funded vs unfunded streams.”
Build a Dashboard That Tells a Story
Most dashboards are digital junk drawers — six charts, thirty numbers, no decision at the end of it. A good dashboard guides the viewer’s eye from the big picture to the deep dive without making them work for it. Eyes scan in an F-pattern: top-left, across, then down. Use that.
1. KPI cards at the top
Lead with the “so what?”. Total revenue this year. Active monthly donors. Donor retention rate. If a board member only looks for ten seconds before their phone pings, these are the numbers they should walk away with. Ask Filly: “Show total revenue YTD as a KPI card with comparison to the same period last year.”
2. Trends underneath
A single number is a statistic; a number over time is a story. Drop line charts under your KPIs to show momentum — where the spikes were, where the dips happened, where you’re heading. If one $50k bequest is dragging the mean up, ask Filly for the median instead to see what a typical supporter actually looks like.
3. Breakdowns in the middle
Once the board sees the trend, they’ll ask “where is it coming from?” Bar and pie charts answer that — fundraising channel, region, campaign, donor cohort. If June spiked, the bar chart should explain whether it was the EOFY appeal or the Meta Ads campaign.
4. A trend line at the bottom
The bottom of the dashboard isn’t where leftover charts go to die. It’s where the strategy lives. Ask Filly for a linear regression on recurring revenue or active donor count — a projection turns a report into a roadmap. “If we stay on this path, here’s where we’ll be in six months” is a much more useful sentence than “here’s what happened in March.”
Lead With Leading Metrics
Lagging metrics tell you what already happened — total revenue at year-end, donor retention rate after the appeal closed. Accurate, but by the time you see them, the window to change anything has closed. Leading metrics track the behaviour that produces those results: donation page starts that didn’t finish, recurring sign-ups this week, lapsed-donor reactivation rate, average ask amount being completed in your form. A healthy dashboard carries both — lagging metrics to report, leading metrics to act on.
Tips for Better Results
- Be specific about time periods. “Last 30 days,” “FY25,” “Q2 vs Q2 last year.” Vague timeframes get vague answers.
- Iterate. Your first prompt is the start of a conversation, not the whole conversation. “Filter that to NSW only,” “now break it down by month,” “switch to a line chart.”
- Ask for the right average. One $50k bequest will inflate your mean. If you want to see a typical supporter, ask for the median.
- Compare. “How does this compare to the same period last year?” turns a number into a verdict.
- Kill your darlings. Before sharing a dashboard, ask: “if I deleted this chart, would the decision change?” If no, delete it. Design for the eye, not the ego.
Garbage In, Garbage Out
Filly’s outputs are only as good as the data they’re built on. If the same supporter appears under three different emails across Salesforce, Mailchimp and Raisely, you’ll double-count them. If a spreadsheet has “to follow up” typed into a column that should hold numbers, you’ll get a confidently-wrong chart back. The fix is upstream — connect tools directly where possible, keep identifiers consistent across systems, and bring in historical data when you migrate CRMs. The preparation you put in once saves you arguing with the dashboard for the next five years.
Save and Share Your Work
Everything saves automatically. Dashboards persist as soon as Filly creates them — there’s no “save” button to forget.
Private by default. A new dashboard is visible only to the person who created it. When you’re ready to bring the team in, hit Share. From there you can grant view or edit access to:
- All members of your organisation in one click,
- Admins only, if a dashboard is just for leadership or the board, or
- Specific people, added by name or email — handy for giving your bookkeeper access to one finance dashboard without opening the whole account.
Need to send a dashboard to a funder, journalist or board member who isn’t in your Filament account? Generate a public link from the same dialog — read-only, revocable any time, no login required at the other end.
Navigating Your Dashboards
Dashboards open as tabs at the top of the screen, the way a browser works. Switch between them with one click, close the ones you’re done with, and Filament remembers which tabs were open next time you log in.
- Browse view: the pinned chart icon at the top-left of the tab bar — or Dashboards in the sidebar — opens the full list. From there, filter by All dashboards, My favourites, or My dashboards, and use the search bar to find one by title.
- New tab: the + button on the right of the tab bar starts a fresh dashboard from a prompt.
- Favourites: hit the heart icon on any dashboard — either in the Browse list or in the dashboard header — to favourite it. Favourites appear in the Favourites section of the sidebar for one-click access from anywhere in Filament.