Why the last mile keeps letting you down

Cario provides a freight control tower enabling predictive visibility, proactive intervention and reliable last-mile delivery. Here’s how.

If there is one place freight plans tend to unravel, it is the last mile. Linehaul and distribution can run smoothly, inventory can be picked on time, and shipments can leave the warehouse exactly as planned, only for delivery expectations to fall apart at the final handover.

Customers are told one ETA, carriers deliver another, and operations teams are left explaining why something that looked on track yesterday is suddenly late today.

For many logistics leaders, this is not a failure of effort or intent. It is a failure of visibility.

“The last mile is where expectations are made or broken. If you cannot see what is actually happening as it happens, you are always reacting after the fact,” says Fiona Marshall, CEO of Cario.

The first and middle mile are typically more controlled environments. Freight moves between known locations, often on contracted lanes, using predictable assets and schedules. While disruptions do occur, they are usually easier to anticipate and plan for.

The last mile is shaped by far more variables

“It’s often where there’s complexity in what you’re delivering,” Fiona says. “Sending a pallet linehaul is very different from sending hundreds of parcels from a depot to a residential address, a hospital or a CBD location. There are a lot more factors that can make something go wrong.”

Traffic congestion, parking access, site readiness, staffing availability and customer-specific delivery rules all introduce uncertainty. Even small issues can quickly cascade, particularly in dense urban environments or regulated delivery settings.

In sectors such as healthcare logistics or critical supplies distribution, those disruptions carry consequences that extend beyond cost or service metrics.

Yet many organisations still attempt to manage the last mile using the same planning assumptions and tools they rely on upstream.

The limits of static ETAs

Static carrier-provided ETAs remain common across freight networks. These estimates are typically calculated at the time of booking and updated infrequently, regardless of what is happening across the network in real time.

“Carriers give you a single ETA, but they don’t take into account what’s happening at that point in time,” Fiona says. “They don’t account for traffic, weather, rail outages or network congestion.”

As a result, ETAs often appear compliant until it is too late to intervene. Issues surface only after delivery windows are missed and customers are already impacted.

“What we see again and again is organisations having plenty of data, but no way to connect it into something usable,” Fiona says. “They know what happened, but not early enough to change the outcome.”

From systems to control-tower thinking

Control-tower thinking shifts freight management from retrospective reporting to active, real-time decision-making.

Cario provides a cloud-based freight and transport management platform designed to give organisations control over complex, multi-carrier freight networks. At its core, the platform combines a freight management system for third-party carriers with a transport management system for organisations operating their own fleets or last-mile delivery services.

Rather than replacing ERP, WMS or finance platforms, Cario sits across them, connecting carrier, fleet and operational data into a single layer. This allows teams to see where freight is, what it is costing, and where delivery risk is emerging across first, middle and last mile movements.

That visibility is not confined to the control room. A control tower is only effective if it can communicate decisions back to the people executing the work.

Cario extends this capability to drivers through a mobile app that supports real-time route selection, optimisation and reporting, while also ensuring delivery processes and compliance requirements are met in the field.

“We position ourselves as a freight control tower for your whole freight,” Fiona says. “You want the visibility of everything, the physical movements and the financial movements, all in one place.”

For large enterprises moving freight at scale, this model is essential. Manually tracking thousands of consignments is impractical. Instead, teams need exceptions to surface automatically, early enough to act.

“The idea of a control tower resonates because executives want control, but they don’t want to be hands on with everything,” Fiona says. “They want the issues surfacing up that their teams can deal with.”

Predicting problems before they happen

The value of a control-tower approach lies in early intervention.

By analysing data across hundreds of carriers, customer systems, fleet operations and external sources such as traffic and weather, predictive models can identify when consignments are unlikely to meet delivery milestones.

“If you’re expecting a five-day journey and the parcel is still in the same place on day three, based on analysis of all other journeys we can predict that it’s going to be late,” Fiona says.

That early signal gives organisations options. In time-critical environments, it may trigger a decision to expedite a replacement shipment, reroute freight or escalate with carriers. In lower-risk scenarios, it allows proactive communication that protects customer trust.

“You want the power to know when things are going wrong, early enough to proactively manage them,” Fiona says. “If you’re operating at scale, you can’t do that manually. You need the system to tell you which consignments matter.”

Where AI adds value

Artificial intelligence underpins this capability, but its role is often misunderstood. AI is not there to replace operational judgement or flood teams with alerts. Its value lies in analysing large data sets, identifying patterns and highlighting only those exceptions that require action.

“We couldn’t do this without AI,” Fiona says. “But AI should be there to reduce effort, not create more. If your team is drowning in alerts, the technology is working against you.”

Transparency and governance remain critical. Predictive insights must be understandable and aligned with how teams actually work, or trust quickly erodes.

Confidence in the last mile

As customer expectations continue to rise, last-mile performance increasingly defines how organisations are judged, particularly when something goes wrong. Static planning tools and fragmented systems struggle in environments defined by variability. Control-tower thinking offers a way to connect data, surface risk earlier and support better decisions without rebuilding core platforms. Ultimately, in the last mile, confidence can be one of the most valuable assets a logistics operation can have.

“The goal is confidence,” Fiona says. “Confidence for your team, confidence for your customers, and confidence that when something changes, you’ll know in time to do something about it.