Supply-chain visibility for mid-market exporters: what's worth paying for

At Mahaya we sit across enough mid-market exporters (roughly the $5M–$80M revenue band, mixed FCL and LCL, two to six origin countries) that we've watched the visibility-tool category go through three waves of pitch in two years. The pitches change. The underlying buyer problem barely moves. So this note is a working answer to the question we get most: of the visibility tooling currently on offer, what's actually worth a line item in the operations budget?

The short version: the buyer should pay for two things. Vessel-level event data on the ocean leg, and a single normalized container ledger that survives across forwarders. Almost everything else marketed as "visibility" is either a side-effect of those two, or a dashboard layer that the operations team will stop opening by week six.

What the category actually sells

Strip the marketing and a visibility product is doing one of four things. It's ingesting AIS-derived vessel positions and carrier EDI to produce milestone events (gate-in, loaded, departed, arrived, discharged, gate-out). It's scraping or licensing terminal data to add dwell and rail-availability layers. It's wrapping a forwarder TMS to expose its events to the shipper directly. Or it's a portal layer on top of any of the above, with alerting, exceptions, and an ETA model.

The first two are the substance. The second two are packaging. A mid-market exporter shipping 200–2,000 containers a year does need the substance. Whether they need the packaging depends on the team. We've seen four-person ops groups run cleanly off a raw event feed piped into their own spreadsheet, and we've seen ten-person groups burn out without a proper portal.

What the data is and isn't

The honest framing on vessel events: the events you receive are the carrier's events, with the carrier's latency, with the carrier's definition of "loaded." A vessel reported as departed at 02:14 may have left the berth at 23:30 the prior evening. A container reported as discharged may be sitting unaccessed on a stack for 36 hours. The provider didn't lie. The carrier reported what it reported. Visibility tools mostly cannot improve on the source data; they can only present it cleanly and complain when it's late.

This matters because it shapes the buying criterion. The question is not "how accurate is your data," because almost every serious vendor pulls from the same handful of AIS/EDI sources. The question is: how do you behave when the source goes dark? A vessel skipping a port call, a terminal not transmitting for 18 hours, a discharge event missing entirely. The vendors that have been in this space longest have built fallback logic for these cases. The newer ones tend to silently propagate the gap. Ask for a worked example of the last time a major carrier feed went down for more than four hours. If the answer is hand-wavy, the product is not yet mature.

The container ledger problem

For exporters working with three or more forwarders, which is most exporters past about $15M, the higher-value problem is not vessel events, it's reconciliation. Forwarder A's TMS calls a container "in transit." Forwarder B's calls the same milestone "loaded on board." The exporter's ERP calls it neither. By the time month-end rolls around, the question "where is everything we shipped in March" requires a person.

A visibility platform that normalizes this, producing a single container ledger with consistent event names, consistent timestamps in UTC, and consistent unit of measure, is worth real money. We've seen this collapse a four-day month-end exercise into about three hours. We talk about how this connects to physical inventory placement in our note on inventory positioning; the short version is that you cannot position inventory you cannot count.

What we'd skip

Some categories that look like visibility but mostly aren't, for this buyer:

• Predictive ETA modules sold as standalone. The accuracy improvement over the carrier's own ETA is small (typically 6–18 hours on a 30-day voyage), and the carrier ETA is free. Pay for an ETA model only when it's bundled with everything else, never as an add-on line.
• "AI-driven" exception classification. Most exceptions on the ocean leg fall into about six categories. A rules engine and an experienced ops lead will outperform an opaque classifier, and the rules engine is auditable.
• Mobile-first executive dashboards. The CFO opens it twice, in the demo and at quarter-end. Build the data layer first; ship the dashboard layer when someone asks for it.

Where the spend actually pays back

In our experience the payback shows up in three places. First, in claims and demurrage: a clean event ledger lets the team file dispute documentation in days instead of weeks, which on a mid-sized exporter is a real four-to-six-figure line on a bad quarter. Second, in port-side decision-making: knowing whether to truck or rail out of a discharge port when the yard is congesting. We've written separately on how we monitor congestion; visibility tooling is the upstream input to that. Third, in commercial conversations with customers, where being able to answer "where is my order" in 90 seconds rather than 90 minutes is the kind of thing that shows up in retention.

The buying frame we'd use is the one the IMO publishes in its annual Review of Maritime Transport: measure the cost of a missed event, then ask what category of tool would have caught it. Spend against that. Everything else is optional.

Our working rule of thumb on the budget: for a mid-market exporter doing 500–1,500 containers a year, the right annual visibility spend lands somewhere between $18,000 and $60,000, with the higher end justified only if it includes the normalized ledger and if at least one ops headcount is being structurally avoided. Above that, you're paying for the brand, and the brand does not move the cargo.