Plan Compliance Is a Vanity Metric: The Drift Problem in Truck Route Planning

It is 6 am in a distribution centre outside Rotterdam. Your truck route planning software has done its job overnight. The plan is clean: 53 stops across the Randstad, eight trucks, 91% utilisation, every customer time window respected, every driver inside their hours. The dispatch screen is green.

By 11 am, two drivers are off-route, a customer in Utrecht has rescheduled, one delivery in central Amsterdam has been refused because the access window closed twelve minutes early, and your “optimised” cost per stop is already running 19% above the plan.

What just happened?

The honest answer is the one no one wants to write into a quarterly report: your route plan was never the answer. It was a hypothesis. And by mid-morning, the hypothesis had quietly broken.

Most European fleet operators are measuring the wrong thing. They are measuring plan compliance — did the route run as scheduled, did the truck hit the stops, did the system “work.” What they are not measuring is the gap between the plan they paid an optimisation engine to produce and the operation that actually ran on the road. That gap has a name. It is called drift. And it is where modern fleets are losing margin they don’t even know they had.

The Illusion of Optimisation

Every piece of truck route planning software on the market sells you the same beautiful artefact: a 06:00 plan. Clean lines on a map. Stops sequenced. ETAs locked. Constraints respected. It looks like an answer.

It isn’t. It’s a snapshot of a model, taken at the one moment in the day when reality is most cooperative — before a single vehicle has moved.

The plan assumes three things that European fleets know are never true:

• That the inputs are current. Traffic data is yesterday’s traffic. Demand is yesterday’s demand. The driver who called in sick at 06:15 isn’t in the model.
• That the constraints are stable. The Brussels low-emission zone boundary didn’t change, the access window at the Mercato in Milan hasn’t shifted, no new strike action is unfolding at a Calais crossing.
• That execution will match plan. Drivers will take exactly the modelled minutes per stop, customers will be home, gates will be open, pallets will be ready.

In practice, route plans start decaying the second the wheels start turning. By the time the second delivery is complete, you are no longer running the optimised plan. You are running a degraded version of it, held together by dispatcher judgment and driver workarounds. The plan compliance metric on your dashboard? It is measuring something that stopped existing at 07:14.

The Five Forms of Drift

Drift is not one problem. There are five problems, each costing money in different ways, and almost none of them visible in the reporting layer of a traditional route planning software.

Traffic drift. Real-time route adjustments are the industry’s stock answer, and they help — but only with traffic that has already happened. The HGV stuck behind a closure on the A2 outside Antwerp at 09:30 was routed at 06:00 based on a traffic profile that didn’t include the closure. Real-time re-routing kicks in after the cost is incurred. What is missing is anticipation, not reaction.

Exception drift. Failed delivery attempts are the most expensive form of drift in European urban operations. A single refused delivery in a city like Paris or Madrid doesn’t just cost the redelivery — it cascades. The driver loses 22 minutes, the next four ETAs slip, two customers drop into the next-day reattempt queue, and the planner spends 40 minutes manually re-sequencing tomorrow’s plan. Your plan compliance metric doesn’t see any of this. It only sees that the stop “wasn’t completed.”

Demand drift. Same-day order injection is now standard in European retail and 3PL operations. Your morning multi-stop optimisation was built around the orders that existed at 04:00. The 87 orders that came in between 04:00 and 11:00 are bolted on as exceptions, not optimised into the plan. Each one costs marginal miles you didn’t budget for.

Behaviour drift. The model thinks every stop takes 7.4 minutes. In reality, urban stops take 11. Rural ones take 5. The driver new to the Hamburg route takes 14. The plan was built on averages, and averages don’t drive trucks. Across an eight-truck operation over a week, the gap between modelled and actual stop times compounds into hours of lost capacity.

Capacity drift. Your LTL consolidation plan was elegant at 06:00 — three pickups grouped neatly into a single full-truckload equivalent. By 10:00, one pickup is delayed, one customer cancelled, and the “consolidated” route is now a half-empty truck running close to its planned cost without the planned revenue. Backhaul intelligence that worked yesterday breaks today.

Five forms of drift. Five places margin leaks. None of them measured by the metric most operations report on.

Across Europe, unsuccessful drop-offs represent a massive drain on margin, tacking on €1–€3 in additional costs per package beyond the initial logistics fee. These failures frequently spiral to over €14 per occurrence once warehousing, re-routing, and support overhead are tallied. For the 2024–2025 period, first-attempt reliability has slipped, with 18-22% of B2C deliveries failing to hit the mark on the first try.

Why The Drift is Invisible

Walk into a European fleet operations room and look at the dashboards. You’ll see route completion percentage, on-time delivery percentage, kilometres driven, fuel consumed, driver hours used. Useful numbers. Reassuring numbers.

None of them measure drift.

Drift lives in the gap — the difference between what the route planning software predicted and what actually happened. To see it, you have to be measuring things most fleets don’t:

• Cost-per-stop variance between plan and actual, by route, by region, by day
• Replanning frequency — how many times after 06:00 did dispatch manually re-sequence a route?
• Dispatcher override rate — how often is the system’s recommendation overruled by a human?
• Plan-versus-actual utilisation — not “did we hit 91%,” but “did we hit the 91% the plan promised, or did we drift to 76%?”
• Cascading exception rate — when one delivery fails, how many downstream stops slip?

If you are not measuring these, you cannot see drift. And if you cannot see drift, you cannot manage it. You will keep buying better route planning software, watching the 06:00 plans get prettier, and quietly wondering why the savings the business case promised never quite land.

This is the silent crisis in European fleet operations. Plan compliance is reported. Drift is absorbed laterally — by dispatchers working late, by drivers improvising, by customer service teams handling the WISMO calls. It doesn’t appear on any P&L line because no one is named in charge of it.

The Category is Incomplete, Not Broken

None of this means truck route planning software has failed. The category solved a real problem, moving fleets off whiteboards and spreadsheets into structured, constraint-aware planning. That was necessary work. For European operations dealing with multi-country regulations, ULEZ-style access restrictions, EU driver hour rules, and increasingly tight urban delivery windows, the underlying engine still matters.

But solving the morning planning problem is not the same as solving the operation. A planning tool, however sophisticated, optimises a moment. It does not orchestrate a day. The gap between those two things is where drift lives, and it is the gap the next generation of route planning has to close.

The shift is not “more real-time.” Reactive re-routing is still reaction. The shift is from plan-time optimisation to execution-time decisioning: a system that treats the 06:00 plan as a starting hypothesis, not an answer, and re-decides continuously as conditions change. When same-day orders arrive, the system absorbs them into the plan instead of bolting them on as exceptions. When a driver runs 14 minutes long at a stop, the downstream sequence is rebalanced before the next ETA slips. When a customer in Lyon cancels, the consolidated load is replanned before the truck departs the depot.

This is the work agentic systems are now doing in production at European fleets — running continuous re-optimisation against live constraints, escalating only the genuine exceptions, and absorbing the routine drift that previously lived in the dispatcher’s head.

It is also where Locus operates, but the larger point is not about any one platform. It is about a category that needs to grow up. Optimisation alone is not enough when reality refuses to hold still.

The Next Five Years

The next five years of European fleet optimisation will not be won on better 06:00 plans. The 06:00 plan has been a solved problem for some time. The next five years will be won on what happens between 06:00 and 18:00 — on whether the operation can decide as fast as reality changes, or whether dispatchers, drivers, and customers absorb the gap by working harder.

Plan compliance is a vanity metric. Useful as a hygiene check, dangerous as a north star. The fleets that win the next cycle will be the ones that stop reporting on whether the plan ran, and start measuring whether the operation actually delivered the economics the plan promised.

Drift is the tax most European fleets are paying without seeing the invoice. The first step is reading the invoice. The second is deciding you are done paying it.