A shipper puts a tracking platform in place. GPS pings update every 15–30 minutes. The ops team can see exactly where every load is at any given moment. And then, six weeks later, WISMO calls are still coming in at roughly the same rate as before. The team is frustrated. The platform is working — the data is flowing — but the problem it was supposed to solve isn't going away.
This is one of the more common mismatches in logistics technology adoption, and it stems from a category confusion that the freight-tech industry has not done a great job of correcting: visibility and prediction are solving different problems. Treating a visibility platform as an ETA tool is like treating a GPS map as a navigation system — you can see where you are, but the map doesn't tell you whether you'll get there before rush hour.
What Visibility Actually Is
Freight visibility, in the technical sense, is the continuous or near-continuous reporting of a shipment's current location and status. A tracking platform with ELD integration pings the truck's position. A mobile app shows the driver's current location. A carrier API surfaces a status event — Picked Up, In Transit, Out for Delivery, Delivered — as those events occur.
All of this is reactive information. It tells you what has happened. The location ping at 11:03 AM tells you where the truck was at 11:03 AM. The status event "Departed relay terminal" tells you it left the relay at a certain time. None of this, by itself, tells you when it will arrive. It gives you raw inputs that a human — or a model — could use to estimate arrival, but the estimate requires more than the location point.
Visibility platforms are genuinely valuable for what they are designed to do: reducing the number of calls a 3PL has to make to a carrier to ask "where is this load." If a dispatcher can open a screen and see the truck is 180 miles out, they do not need to call the carrier. That is real operational value. But knowing the truck is 180 miles out does not tell you whether it will arrive in 3 hours or 4.5 hours — because that depends on traffic, driver hours of service remaining, whether there is a mandatory rest stop needed, and the carrier's typical performance on that final segment of the corridor.
What Prediction Requires That Visibility Does Not Provide
A reliable ETA prediction needs to combine the current location data that visibility provides with at least three additional layers:
Historical transit time context. How long does it typically take a truck at this exact segment of the route to complete the remaining distance? Miles remaining divided by average speed is not good enough — it ignores real-world corridor behavior, congestion patterns, and relay-point handling time. A model trained on historical BOL data for this carrier on this lane can provide a distribution of how long "180 miles out on I-20 at 11 AM" typically resolves to — not a single number, but a range with a confidence band.
Driver hours of service (HOS) status. A truck 180 miles from its destination with a driver who has 45 minutes of driving time remaining before a mandatory rest period is in a fundamentally different situation than the same truck with 7 hours remaining. Visibility platforms show location. HOS status, accessible through ELD integrations, is what converts location data into arrival probability.
Corridor-specific real-time factors. Weather in the path, known construction delays on the specific route segment, time-of-day congestion patterns near the destination metro. These need to be applied against both the remaining distance and the historical transit distribution to produce an adjusted probability range.
When all of these inputs are combined, what you get is not "the truck is 180 miles out" but "based on current position, HOS status, this carrier's historical performance on this lane segment, and current weather on the corridor, the P80 arrival window is 2:15 PM to 5:00 PM with the median at 3:30 PM." That is what a consignee or receiving DC can actually use.
Why WISMO Calls Persist After Visibility Deployment
A consignee calls to ask "where is my shipment" for one of two reasons. First, they genuinely do not know its location or status — the visibility problem. Second, they know it is somewhere in transit but they do not know when it will arrive — the prediction problem.
Visibility platforms, when deployed with customer-facing portals, largely solve the first trigger. A consignee with portal access can see the truck is 180 miles out and in transit. But that does not prevent the second call: "I know it's on the way, but will it be here by 3 PM? We have a receiving appointment and I need to know if I should reschedule." That call requires a predicted arrival time, not a current location.
This is why WISMO call volumes often plateau after a visibility deployment rather than continuing to fall. The first wave of calls — pure status uncertainty — gets eliminated. The second wave — arrival time uncertainty — persists because visibility data alone is insufficient to answer it.
When Visibility Is Enough
We are not saying visibility tools are insufficient or that every shipper needs a full prediction layer on top of them. For many use cases, current-location visibility is exactly the right tool. Internal dispatchers who just need to know whether to reassign dock doors, or 3PLs whose customers have flexible receiving windows where ±3 hours does not matter, are getting most of their value from visibility alone.
Prediction becomes critical when the cost of arrival-time error is high. That is typically any combination of: tight receiving appointment windows at destination DCs, time-sensitive freight (perishables, production-line just-in-time deliveries, retail distribution center cut-off times), or high customer service sensitivity where a missed delivery window triggers a claim or SLA penalty.
For these use cases, knowing where the truck is right now is necessary but not sufficient. The question is not "where is it" but "will it be there in time" — and answering that with confidence requires the prediction layer that visibility platforms, by design, do not provide.
A Practical Framework for What You Actually Need
A useful diagnostic for a 3PL ops team evaluating their current tooling: log your last 30 WISMO calls and categorize them. How many were asking for basic status (where is the truck, has it been picked up, has it cleared the relay)? How many were asking specifically about arrival time or requesting a tighter ETA window?
If the distribution is roughly 60/40 or worse toward arrival-time questions, a visibility platform alone is likely not closing your WISMO gap. The information your customers are calling for is not on any screen they have access to, because it hasn't been calculated yet — it requires prediction, not tracking.
The two capabilities are complementary. Real-time location data is an input to a good prediction model. A prediction model without visibility data loses a significant signal. The mistake is treating them as interchangeable, or assuming that the first one you deploy handles the full problem. For most active 3PL environments handling volume where arrival timing matters, both layers are necessary — they're just answering different questions.
