Every logistics operator we talk to says visibility is solved. They say this with confidence — often mid-sentence, as if the claim doesn't require evidence. They have a TMS. They have carrier integrations. They get ETAs from their LSP portals. What more is there?
We disagree, and our investment in FourKites is the clearest expression of why we think the category is still in its first inning. Not because operators are wrong about what they have today. They're right about their current state. They're wrong about what "solved" means.
The visibility they have versus the visibility they need
What most European logistics operators have today is status visibility: milestone-based updates that tell them a shipment has departed, arrived at a transshipment point, cleared customs, or been picked up for last-mile delivery. This data comes from carriers, forwarders, and port systems — and it arrives in different formats, at different latencies, through integrations that break every time a carrier updates their API or an operator switches a lane.
What they actually need is predictive visibility: not where is the shipment now, but what is going to happen next and when, with enough lead time to act on the information rather than react to it. The difference sounds semantic. In practice it determines whether a supply chain team is managing exceptions or being managed by them.
Here is a concrete example of what that distinction looks like. A Swedish consumer goods company — mid-size, 40 to 50 inbound container movements per week from Asia — has standard carrier tracking integrated with their freight forwarder portal. On paper, they have visibility. What they experience in practice: vessels depart Shanghai, tracking updates arrive irregularly, an ETA shows in their system on day 3 of a 28-day voyage, and that ETA doesn't update when the vessel reroutes around weather off the Suez corridor. By the time their logistics team learns the ETA has slipped by 6 days, the production schedule at their Gothenburg DC has already been locked. The containers arrive, there's no labour allocation, the DC runs emergency overtime, and the margin hit for the quarter is partially traceable to an ETA that was wrong and didn't tell anyone it was wrong.
That is not a solved visibility problem. That is a status feed masquerading as predictive intelligence.
Why the "solved" perception persists
The persistence of the "visibility is solved" narrative comes from a specific dynamic: the operators who are most vocal about the category are the ones with the most sophisticated setups. A large 3PL running 500 truckloads per day across 12 European countries has teams dedicated to carrier integration, has negotiated real-time telemetry feeds with their top 20 carriers, and has built (or bought) proprietary exception management workflows. For them, the tactical visibility question is largely handled. They are not representative.
The mid-market — a regional distributor running 30 to 80 truckloads per day, a manufacturer with inbound raw materials visibility problems, a 3PL growing from 2 to 5 distribution centres — does not have that infrastructure. They have a TMS from a tier-2 vendor, carrier integrations that cover maybe 60% of their freight spend, and a logistics coordinator whose job is partly to call carriers and manually update ETAs in a spreadsheet. When we talk to operators in this segment across Germany, the Netherlands, Poland, and the Nordics, the picture is consistent: the visibility they have is just good enough that they don't classify it as a problem, but not good enough to actually prevent the exceptions that cost them the most.
This is the gap the category has to close. Not for the top 50 European 3PLs — they are partially covered. For the next 5,000 operators who have real freight complexity and nominal visibility tooling.
What FourKites is actually doing that matters
When we wrote our first check into FourKites in 2024, the aspect of their product that we weighted most heavily wasn't the breadth of their carrier network — though that matters. It was the predictive ETA layer and what they were doing to make it genuinely operational.
Most visibility platforms aggregate status data and display it. FourKites has invested heavily in building a model that takes the aggregated status data and produces probabilistic ETA predictions — accounting for lane-specific patterns, carrier performance history, weather, border crossing delays, and port congestion signals. The prediction is not a point estimate with no confidence information. It is a probability distribution that tells the operator: there is a 70% probability this arrives within the committed window, a 20% probability it arrives 4 to 8 hours late, and a 10% probability the delay exceeds 12 hours. Those are different operational responses. The point estimate treats them as identical.
What this means practically: an operator running a manufacturing line with a tight JIT inbound schedule can use probabilistic ETAs to pre-position contingency inventory for the scenarios where on-time probability falls below a threshold they define, rather than either carrying blanket buffer stock or scrambling when the late shipment hits. The savings from reducing reactive buffer decisions — even at a relatively small operation — tend to be larger than the cost of the visibility platform by a comfortable margin.
We're not saying FourKites has no competitors or that the problem is technically simple. There are other well-funded visibility platforms doing serious work in this space. The category thesis doesn't rest on a single company winning. It rests on the observation that the underlying infrastructure the industry needs — predictive, probabilistic, multi-modal visibility that integrates into planning systems — is materially incomplete, and the companies building toward it are earlier in their growth curves than the market narrative suggests.
The integration layer is where value actually accumulates
One thing the visibility category got wrong in its first generation — the 2015 to 2020 wave of visibility startups — was treating the product as a dashboard. A better tracking map. Nicer ETAs in a UI. That's genuinely useful, and some of those companies built real businesses on it. But it also made the product feel like a reporting layer rather than an operational layer. Operators could ignore it between exceptions. It didn't change how they made decisions; it just gave them a slightly better picture of what was happening.
The shift that we think defines the current generation of visibility platforms is the integration into execution systems. When visibility data feeds into a TMS routing decision, into a WMS receiving schedule, into a procurement system that tracks purchase orders against expected inbound goods — that's when the product stops being a dashboard and starts being infrastructure. The switching costs change. The ROI calculation changes. The buyer profile changes: it's no longer just the logistics coordinator; it's the VP Supply Chain who signs because the integration touches procurement planning, and it's the CFO who sees the working capital impact in the numbers.
FourKites has been building the integration surface aggressively — their connectors into SAP TM, Oracle SCM, and the major European TMS platforms are a meaningful part of what we're betting on. The value of the visibility product compounds as it becomes embedded in the systems that actually govern supply chain decisions, not just the systems that report on them.
What founders building in adjacent categories should understand
For founders working in supply chain software — demand forecasting, procurement AI, warehouse intelligence, freight procurement — the visibility layer matters to you even if you're not building in it directly. Visibility is the data foundation that every other supply chain decision system depends on. A demand forecasting model that doesn't know the actual status of inbound purchase orders is operating on a materially degraded input. A warehouse WMS that doesn't receive accurate arrival windows can't optimise labour allocation for receiving. A freight procurement platform that can't account for actual carrier on-time performance on a lane is making rate recommendations with a significant blind spot.
This creates both a partnership opportunity and a competitive framing consideration. If you're building upstream or downstream of visibility — think carefully about how your product integrates with or depends on visibility data. The companies that are building end-to-end supply chain software stacks are increasingly using visibility infrastructure as the connective tissue. The standalone procurement AI or standalone demand forecasting tool will face pressure to either integrate with visibility platforms or build their own. Neither is cheap.
For founders specifically targeting the mid-market segment we described earlier — the operators with 30 to 80 daily movements and partial visibility coverage — there is a real opportunity in building visibility tooling that is lighter than the enterprise platforms and can land quickly. The mid-market doesn't need a six-month implementation timeline. They need something that covers their top 10 carriers, integrates with their existing TMS in two weeks, and shows them where their exceptions are going to come from. That product is not built yet at the scale and price point the market needs.
Why we keep investing in the category
Kvistlund has backed multiple companies in the visibility and freight-tech adjacent space across Fund I and Fund II. The consistency isn't momentum-chasing — visibility was part of our founding thesis in 2020 and our conviction has gotten stronger, not weaker, as we've watched the category develop.
The reason it gets stronger: the supply chain disruption events of 2021 to 2023 permanently reset operator expectations about resilience. Before that period, many European logistics operators were running supply chains optimised for average conditions. The events of that period — port congestion, carrier capacity shifts, geopolitical rerouting, pandemic-related DC closures — exposed exactly how brittle the average-case-optimised supply chain is when the variance increases. Operators came out of that period with a new mandate: they needed earlier warning, more accurate predictions, and shorter lead times between a disruption signal and a corrective action. That mandate is structural. It didn't go away when freight rates normalised in 2024.
The infrastructure to fulfill that mandate is being built right now. The category is not in its final chapter — it's in the chapter where the real technical differentiation starts to compound and the best platforms start to pull away from the rest. That's the inning we're in, and it's why we're still here.