Map boards
Filter it down, down, down
Most field software has a map, and most of those maps are the dispatch board with pins on it: today’s jobs, four dropdowns, one answer. Useful for an hour on a Tuesday morning and useless for planning anything.
This is a different animal. A map board is a saved set of filters over any dataset in the platform — sites, jobs, equipment, recurring work, route stops, who pays where. You narrow it until it says something, then you save that as a board and it is there tomorrow.
A map board is a saved set of filters
It is the same idea as a named dispatch board, pointed at geography. A board is a question you ask often enough to be worth naming — Dollar General PMs, Tuesday’s bank run, every rooftop unit we have not touched this year — and once it is named, asking it again costs one click.
- Unlimited boards. Making another one costs nothing, so people make them. That is the difference between a tool somebody uses and a feature somebody demoed.
- A board survives. Save it, close the browser, come back tomorrow, and it opens exactly where you left it — same layers, same filters, same colouring.
- It is yours, per company. Your boards do not follow you into somebody else’s books, and a shop with two divisions keeps two separate sets.
The same saved-board idea the dispatch board runs on. A board is a question with a name on it.
Any dataset, not just today’s jobs
This is where a map stops being a dispatch view. The layer chooser offers the real objects in the business, each with its own filters and its own facets:
| Layer | The question it answers |
|---|---|
| Sites | Where work happens. Every building and suite you serve, with the zone resolved from its postal geography. |
| Jobs | Who is doing what, where, today — the board’s own answer, on a map instead of in columns. |
| Work orders | The job rather than the visit, reaching back past the board’s window into history. |
| Recurring work | What you have promised and have not done yet, with the room left on each promise. |
| Equipment | The units under service, and when each one is next due. |
| Route stops | One run, in sequence. The route plan → |
| Routes | Which runs exist, for whom, on what day. |
| Who pays where | The payer behind each site — which matters enormously when a national account changes hands. |
Each layer is powered by the same published capability the screens and the API use, so a layer is never a special report somebody has to maintain separately. How that works →
The number narrows, which is the whole point
Open the sites layer and it says how many rows you are looking at. Tick one zone and that number drops. The state list collapses to the states that survive, the city list collapses under it, the legend follows, and the table beside the map follows too. Tick another thing and it narrows again.
That sounds obvious and almost nothing does it, because doing it properly means every filter has to re-count every other filter after your choice — which is the difference between a facet list you can trust and one that offers you options returning nothing.
A facet ignores its own selection
Pick one zone and the zone list still shows the others with their counts, so you can swap sideways. Every other facet re-counts against your choice. That asymmetry is deliberate and it is what makes the thing usable.
Colour by anything
Colour the pins by zone, by status, by assignee, by priority, by which truck, by whether it is still open. The legend is the answer to “what am I actually looking at.”
Numbers and text, not just lists
Facets know what kind of thing they are. A count of suites is a number, a site name is text, active is a flag — and each filters the way that kind of value should.
Then save it
The whole arrangement — layers, filters, colouring — saves under a name. That is the board.
The sites behind the sites layer. Same data, same capability — the map is another way of asking.
Two layers at once, filtered separately
The reason this matters is comparison. One dataset on a map is a picture. Two datasets on a map, each filtered on its own terms, is a planning tool.
- Overdue recurring work against the sites you already visit weekly. If the PM you are behind on is a mile from a stop you make every Tuesday, that is a decision, not a report.
- Tomorrow’s jobs against the equipment coming due. One truck is already going there. Somebody should know that before the truck leaves.
- Work against who is paying for it. A national account changing property managers is invisible on a job list and obvious on a map.
- Narrowing one layer leaves the other alone. Each layer keeps its own filters, its own counts and its own colouring, which is what makes it a comparison rather than a merge.
Equipment is one of the layers. Units under service, when each is next due — and where they are.
It tells you what it is not showing you
A map is a confident-looking thing, and that is exactly what makes a dishonest one dangerous. Nine pins on a screen look like nine jobs. If the other thirty could not be placed, you have just planned a day against a picture that quietly left them out.
- Every layer says how many of its rows are not on the plot. Not buried in a tooltip — on the face of it, next to the count. When you first switch a map on, that number is the most useful one on the screen, because it tells you how much of your address book has real coordinates yet.
- A partial pull is announced. If the server holds more rows than it sent, the header says so and every count on that layer renders as a floor — 482+, not 482. A number that might be wrong is labelled as a number that might be wrong.
- Sites at the same address stack into one counted marker rather than hiding each other, so a strip mall with eleven suites does not read as one job.
- If it cannot draw a basemap, it says so in words and still plots the real coordinates in true relative position, rather than showing a grey rectangle and letting you wonder.
And this is how territory actually works
Underneath the map is the thing that makes it more than a picture: a zone is a profit-and-loss centre, not a colour. It owns a team — a dispatcher, a billing clerk, a fleet of techs — and a timezone, and it is the unit you measure the business in.
Its geography is assembled from real postal boundaries — carrier route, ZIP, county, state, each one more specific than the last, glued together into a territory and resolved most-specific-first. So when a job arrives — typed in, or over the API from a broker at two in the morning — it lands in the right team’s queue without a human deciding whose it is.
Zones, each with its own timezone. Departments cut across them — everybody has plumbing — so neither one nests inside the other.
The zone is stamped onto the job when it is created, as a value. A territory realignment next year changes where new work lands; it does not quietly rewrite which team did last year’s.