Business intelligence
Show the work, and what it changed
Software that runs an operation is also the best record of it. We build the reporting layer on top: live dashboards, impact maps, delivery timelines and board-ready reports, reading from the same database the work happens in.
Deployed and running
Reach, by region and by site
A worked example on illustrative figures: 307 organizations across the nine US census divisions, reaching 240,500 people every working week. Shade the map by whichever measure matters, then take any region or site apart.
Interactive demonstration with illustrative figures. The regions, measures and sites show the shape of this reporting; they are not a client’s data.
Reporting
Every system we ship reports on itself
A worked example of the reporting layer we build in from the start, shown here on illustrative data for a fictional multi-site operator. Four views, four questions a monthly review has to answer. Every chart has a data table behind it, and every value in a tooltip is readable without one.
Who is using the system, month by month, across three teams.
Active users in December
1,178
up +6.9% vs November
Added since January
928
up +8.9% vs November
Field team share of users
58.4%
up +0.2 points vs November
Active users by team
Three teams on one axis. Color follows the team, never its rank, so the reading holds if the order changes.
- Field team
- Depot team
- Office team
We build these for clients who want their own delivery visible, and we run our engagements on them.
Delivery tooling
One program, a plan and a board
The scheduling and tracking surfaces we build for delivery teams, shown on an illustrative twenty week program. Every value on the chart is also readable without a pointer.
- Discovery
- Design
- Build
- Integration
- Launch
Bars carry the share complete. The percentage is written on a bar only where it fits without crowding; every value is in the table view and on focus.
Interactive demonstration. Move a card with the arrow keys when it has focus, or collapse a workstream on the timeline.
One dataset, different questions
A supervisor, a manager, a trustee and a funder are not asking the same thing. The same records answer all four, presented differently.
Operational
What is happening now. Live counts, queues, exceptions and alerts.
Management
What happened this period, against what was expected, and where it diverged.
Impact
What changed for people. Hours saved, reach, outcomes, and where geographically.
Board and funder
A short, defensible narrative with the figures that stand up to scrutiny.
A dashboard is only as good as its plumbing
Anyone can draw a chart. The hard part is that it stays true, stays current, and can be defended when someone asks where the number came from.
- It reads from the system of record
- Reporting sits on the same database the work happens in, so a figure on a board pack and a figure on a screen cannot disagree.
- Every figure is traceable
- A number on a chart can be opened down to the rows behind it. A report nobody can audit is a report nobody should act on.
- Scheduled, not chased
- Reports build themselves and arrive on a schedule, by email or into the tools your team already opens.
- Exports that survive the meeting
- CSV and XLSX for the analysts, PDF for the pack, and an API for whatever comes next.
Get in touch
Bring us the operation, we will show you what it is doing
If you already run software and cannot get a straight answer out of it, that is a reporting problem and it is fixable.