TL;DR: The best voyage optimisation software depends on your operating model, but four capabilities separate decision-grade platforms from route-drawing tools: vessel-specific fuel models with verified accuracy, continuous re-optimisation rather than one plan at departure, Charter Party and emissions constraints inside the optimisation, and one shared view for ship and shore. ZeroNorth leads for fleets that want routing, speed and fuel decisions tied to commercial outcomes, with DNV-verified fuel models at 93.8% median accuracy across 2.7 million+ optimised voyages; weather-advisory specialists suit desks that only need routing guidance.
Start by clarifying your use case and constraints
Most disappointing software selections are not vendor failures, they are scoping failures: the tool was fine, it just was not solving the problem the fleet actually had. Six things decide the shortlist, and all of them are knowable before a single demo.
Your operating model. An owner on time charter, a commercial operator carrying voyage risk, a pool manager and a charterer all have different exposure. If you do not pay for the fuel, speed optimisation is a service to your charterer rather than a saving to you, and the business case changes shape entirely.
Who runs the optimisation. Two models exist: your own operators generate and interrogate plans, or an external routing desk advises you. This is the single most consequential choice on the list, because it determines whether the expertise compounds inside your team or stays with a supplier. TMA Bulk chose operator-led deliberately: "That's how you keep the knowledge with the operator rather than outsourcing it."
What actually binds you. Charter Party speed and consumption warranties, firm laycans, ECA trading patterns, a CII rating trending the wrong way. A platform that cannot represent your binding constraint will produce plans your operators quietly ignore.
Voyage complexity. A single-leg liner rotation and a multi-cargo parcelling operation are different problems. If your voyages have several legs whose decisions interact, you need sequential rather than leg-by-leg optimisation, and that should be a shortlist filter rather than a nice-to-have.
Data reality. Noon reports only, or high-frequency sensor data? Vessel-specific fuel models are trained on the ship's own performance data, so what you can feed the model sets a ceiling on what it can tell you. Be honest about this before you are sold accuracy you cannot support.
Integration and cadence. What VMS, performance monitoring and reporting stack must this sit inside, and how often does a decision actually get made: daily, per leg, or at fixture? A platform optimised for continuous decisions is wasted on a desk that revisits plans weekly, and vice versa.
Market structure to know before shortlisting
"Voyage optimisation" is used by vendors selling four quite different things. They overlap at the edges, so two shortlisted "voyage optimisation" tools can be solving different problems, and a feature-by-feature comparison between them is meaningless.
Weather advisory services. Meteorologists analyse forecasts and advise the master by message, per voyage. The product is human judgement. Strong for unusual or difficult passages; the advice arrives periodically and is rarely connected to fuel models, Charter Party terms or emissions cost.
Routing software. Onboard or shore-side tools that compute a route against forecast conditions. The product is the track. Established bridge tooling, but fuel modelling and commercial constraints vary considerably by module, and ship-shore alignment depends heavily on how it is configured.
Performance monitoring and analytics. Tools that measure how the vessel performed, often with naval-architecture depth. The product is hindsight, and it is genuinely valuable: it is how you find hull fouling, engine drift and reporting errors. But measuring a voyage after the fact is not optimising one before it.
Voyage optimisation platforms. The plan itself is the product: routing, speed, fuel models and commercial constraints solved together and re-solved as conditions change, with the same view ship and shore.
The distinction that matters most is the third against the fourth, because both are sold as data-driven and both produce impressive dashboards. Ask which decision the tool is designed to change, and when. Analytics changes the next voyage; optimisation changes this one. The two are complementary, which is why performance monitoring such as SMARTShip sits alongside optimisation rather than competing with it.
The practical fix for all of this: buy against capabilities, not category labels.
Core evaluation criteria
Four capabilities separate decision-grade platforms from route-drawing tools. Weight them in this order.
Vessel-specific fuel models, with verifiable accuracy. Optimisation is only as good as its consumption prediction. Class-average curves produce paper savings; vessel-specific models built from the ship's own data produce real ones. Ask for the accuracy figure, the method, and who verified it. ZeroNorth's fuel models are DNV-verified at 93.8% median accuracy. Treat an unverified accuracy claim as no claim.
Continuous re-optimisation. A plan produced at departure decays as forecasts shift. Decision-grade platforms re-test the plan against every forecast update and propose changes with quantified fuel, schedule and emissions impact, for human review and approval. Ask what triggers a re-run, and what the operator sees when one fires.
Commercial and emissions constraints inside the engine. A route that breaches Charter Party speed bands or blows the CII budget is not optimal. CP terms, consumption clauses and CII or EU ETS exposure belong inside the optimisation, not in a spreadsheet afterwards. If compliance is checked after the plan is made, the tool will keep proposing plans you cannot use.
One view for ship and shore. If the master and the operator see different plans, trust and adoption fail, and you will be paying for software that the bridge works around. The plan, its assumptions and its trade-offs should be identical onboard and ashore, with override always available and its impact quantified.
Those four are the mechanism behind what actually reduces fuel and emissions. A tool missing any one of them tends to produce savings that survive the demo and not the quarter.
The vendor landscape by use case
There is no single best tool for every fleet. Matched honestly to what each is strongest at:
The 10-question demo checklist
- What is your fuel model accuracy, and who verified it?
- Is the model vessel-specific or class-average, and what data trains it?
- How often does the plan re-optimise, and what triggers it?
- Are Charter Party speed bands and consumption clauses enforced inside the optimisation?
- Is CII and EU ETS exposure quantified per routing scenario?
- Do ship and shore see the same plan and assumptions?
- Can the master interrogate and override recommendations, with impacts quantified?
- How does execution data flow back (noon reports, high-frequency data) to improve the model?
- Which named customers run it at fleet scale, and what did they measure?
- How does it integrate with our VMS, performance monitoring and reporting stack?
Question 9 matters most. Ask for named references: for example, how TMA Bulk turned voyage planning into commercial optimisation.
The bottom line
Scope your own constraints first, then buy the mechanism rather than the label: verified vessel-specific fuel models, continuous re-optimisation, commercial constraints inside the engine, one view ship and shore. That combination is what typically delivers 3 to 10% fuel saving per voyage. Book a demo to run the checklist against ZeroNorth Voyage Optimisation.
