Why it exists

Sales knew who billed the most, but not who left margin behind. It's not the same list: some high-volume clients operate near zero, and some small clients carry the result.

• Portfolio concentration. How much of revenue depends on the top five clients. It's a risk measure, not a success measure.
• Clients at a loss. Ones that bill and still leave negative margin, with the detail needed to go renegotiate rate with the case in hand.
• Average per shipment and mix by mode of transport and by office.
• Planned vs. actual transit, to know whether what's promised to the client resembles what actually happens.

It reads across four tabs: the consolidated view of both entities, each entity separately in its functional currency, and a holiday-close notice checklist — who's already been notified, who acknowledged it, and when.

The transit comparison comes with its coverage percentage attached to the average, because an average calculated over half the data isn't an average, and hiding that invites a decision made on thin air.

Where the data comes from

The main source is a different report from the rest of the suite, and it has its own gaps because of it: it carries no posting date and no actual transit dates, only estimated ones. The dashboard fills in what's missing by cross-referencing the other two webhooks — the billing date against receivables, and actual transit against the shipment feed.

That the data comes from two different sources isn't hidden: planned and actual are labeled differently in the interface, because comparing two things that weren't measured the same way is exactly where false conclusions creep in.

Build decisions

• The notice checklist groups by client name, not by code. The same client bills under a different code in each entity; grouping by code sends two notices for the same holiday close and makes the whole thing look disorganized.
• Only jobs with recognized revenue are included. Ones still open don't have margin yet, and putting them in a profitability average drags it down without anyone having done anything wrong.
• Margin is split across a shipment's invoices. If a shipment generates three cost entries and each one gets charged the full cost, margin comes out negative everywhere. It's split a third per invoice.
• The period is filtered by posting date, not departure date. That's the accounting basis, the same one the financial dashboard uses, so the two agree. The few shipments with no invoice fall back to their estimated departure date, and the cell says so instead of pretending it's the same thing.
• The exchange rate is looked up on its own, with two backups and a manual override. What value was used and where it came from is shown on screen: a figure consolidated in pesos without saying at what rate can't be audited.
• Operations born outside the ERP are flagged, not ignored. Since July 2026, part of domestic ground moves are created directly as transport bookings and never show up in the source report. They come in as their own row, with a label that flags them as not a normal shipment.
• No framework, no npm, like the rest of the suite: a self-contained HTML file deployed to the internal portal.

What's different in this demo

Same production file, data layer swapped out for a deterministic generator: same seed, same numbers for whoever opens the link. Dates are relative to the moment it's opened, because an operational dashboard frozen on an old date doesn't prove anything.

• Clients, amounts, and volumes are made up; so is the brand.
• The notice checklist actually works: whatever you check gets saved for the session and is lost on reload.
• Exporting to Excel is disabled: the viewer doesn't grant download permission.