Explore — the lane map, the table, and the bubble chart
Three views of the same shipment rows, and how to tell a network's real shape from a previewed sample of it.
One origin, many lanes
Explore opens on the project's shipments: an origin, its destinations, and a card counting the lanes. Start with a consistency check — off a single origin, the lane count and the destination count should be the same number. When they are, the panel's counts hold together; when they are not, something is either double-counted or terminating where it started.
Four basemaps
Light, Streets, Dark and Satellite. Mostly cosmetic, with one real use: Satellite is how you confirm a destination sits at a distribution park rather than in the middle of a field. That catches geocoding errors which a stylised basemap hides completely.
Colour lanes by, and what greys out
Lanes can be coloured by volume, weight or pallets, each with its own legend. Lanes carrying no data for the chosen metric grey out rather than disappear — the honest choice, and it shows you at a glance how much of the network has no figures behind it at all. Cycle all three metrics: a lane greyed on one and coloured on another tells you which column is missing.
The table view
The same lanes as rows: origin and destination ZIP and city, shipments, weight, pallets, with a CSV export. This is where the arithmetic is checkable, so it is where the next three checks happen.
The check that catches ingestion errors
Compare shipments against pallets down every lane with data. If shipments exceed pallets by the same small constant on every single lane, and lanes with no freight still report exactly that many shipments, then non-freight rows — headers, subtotals, footers — are being counted as shipments on every lane. A constant difference repeated across every row is never a coincidence, and it is the most actionable thing this page will tell you.
Make the tail close
Add the shipments on the lanes that carry figures and subtract from the panel's chip. What remains has to be shared among the lanes without figures. If a plausible per-lane value overshoots or undershoots that remainder, either the tail is not uniform or the chip de-duplicates somewhere — worth settling before you quote the shipment count to anyone.
Check the legends contain the busiest lane
Legends can take their minimum from the real data and their maximum from a range that excludes the top lane. When that happens the lane the network actually runs on sits above the top of its own scale on every coloured view, and its bubble is drawn outside the plot area entirely. Test it directly: find the largest value in the table and check it against the legend maximum on each metric. If it is above, rescale before showing the map to anyone, because the most important lane is the one being rendered least accurately.
Sort order is not always a ranking
Scan a column you expect to be sorted before you read its order as a ranking. If the sequence breaks anywhere, the table is keyed on something else and row position carries no meaning.
The preview cap is the whole caveat
A banner states that large sheets are previewed to the first few hundred rows, and on a big file that is the single most important sentence on the page. Compare the lane count here against the lane count Insights reports off the full file: the ratio is the fraction of the network on the map. Export the CSV when you need the complete rows rather than the previewed ones.
Averages differing between pages is not always an error
An average weight computed here can differ from the equivalent card in Insights simply because the two are computed over different row counts. Both can be correct at once. The habit is to note the population beside any average you quote — and to look at the busiest lane separately, since a lane well above the network average distorts every figure it is averaged into.
How to use the page
Map for shape, table to check the shape has numbers behind it, and the preview banner before trusting any total. Run the shipments-versus-pallets check and the legend check on every new project — both take under a minute, and both find problems that no chart on the page will flag for you.
Ready to build your own?
Start free and ship your first app in minutes — or have us set it up with you.
More guides in Insights & Lane Analytics
More resources