How we compare transfer options

We compare the whole journey, not just what is bookable

A comparison that quietly omits the public bus because no one pays a commission on it is not a comparison. Every transfer method we can verify for a route appears in the table: private transfer, taxi, shared shuttle and scheduled public transport, whether or not a commercial partner sells it.

On the one route we have published so far, the cheapest option by a wide margin is the scheduled coach, which earns us nothing at all. It is labelled as the cheapest option anyway, because it is. How we make money explains why our commercial arrangements are structurally unable to change that.

Facts come from named sources, and carry the date they were checked

Each factual value on a route page — a distance, a fare, a journey time, a set-down stop — is stored with the source it came from, what kind of source that is, the date it was checked and how long it can be trusted before it needs rechecking. That is why route pages end with a list of sources rather than a generic "information may change" line.

Not all sources carry the same weight. In rough order of authority:

• Official and government sources, such as a transport authority's published taxi tariff, or an airport operator's own description of where a service departs from.
• Transport operators, for their own fares, timetables, stops and journey times.
• Commercial providers, for their own products and availability. A provider showing one price for one product tells us about that product, not about the market.
• Values we calculate ourselves from a documented formula and published inputs, which we label as calculated rather than as an observed fare.

Secondary sources — aggregators, travel articles, search-result snippets — are not promoted to primary evidence. They can point us at something worth verifying; they cannot be the reason a fare or a departure time appears on a page.

Not all journey times mean the same thing

Two numbers that both read "45 minutes" can be entirely different claims, so we keep the distinction visible rather than flattening it into one column of minutes.

Scheduled

An operator's own timetabled duration. It is a commitment to a schedule rather than a measurement of traffic, and it already includes the stops the service makes.

Estimated

A driving time we derived, typically free-flowing, without traffic. It is a floor for a normal journey rather than a typical door-to-door time.

Typical

An observed normal range rather than a single figure or a guaranteed duration.

Fixed

A duration the source states as fixed for the stated conditions.

This matters more than it sounds. A coach scheduled at 45 minutes and a car estimated at 40 to 45 minutes are not evidence that the coach is as fast as driving: the coach figure is a commitment that already accounts for its route, and the driving figure excludes the traffic you will actually meet. Printing both as bare numbers would make a false equivalence look like a finding.

Recommendations, and why some routes have no overall winner

We can label an option for a specific context — the cheapest fare, the option that suits families, the one that works for a late arrival — when the route data supports that specific claim. Those labels come from the route's own recorded facts, not from an editorial impression.

We do not require every route to have a "best overall". On many airport routes there genuinely is not one: the cheapest option is not door-to-door, the door-to-door option costs several times more, and which of those is right depends entirely on who you are travelling with and what you are carrying. Where that is the case, the recommendation is deliberately left unset and the page explains the trade-off instead of manufacturing a winner.

A badge without a route-specific reason behind it is decoration, and decoration on a decision tool is worse than nothing.

Unknown stays unknown

When we have not verified something, the page says we have not verified it. We do not fill a gap with a plausible-looking number so that a comparison table looks complete.

A live example, at the time of writing: for our launch route we have verified the coach operator, the fare, the journey time and the stops it sets down at, but not its first and last departures of the day. So the route page tells you the coach may not be running if you land late in the evening, rather than implying a service we have not confirmed. The dedicated bus guide stays unpublished until that gap is closed.

The same applies to luggage allowances, accessibility provision and anything else we have not checked against a source we would stand behind.

The right answer depends on the traveller

Most airport transfer questions do not have a single correct answer, because the factors that decide it are yours rather than the route's:

• Party size. A per-person fare multiplies with your group; a per-vehicle fare is shared. The cheaper option can swap as the party grows, which is why we never compare the two headline numbers directly. Seeprice methodology.
• Luggage. A cheap fare that leaves you at a bus station is not cheap if you then need a taxi and your cases across town.
• Arrival time. Scheduled transport stops running; taxi ranks do not. Night tariffs can be materially higher than daytime ones.
• Door-to-door need. Some options end at your address, others at a fixed stop. That difference is often worth more than the fare gap.
• Certainty. Booking in advance fixes a price and a vehicle size before you fly. Turning up at a rank costs less and commits to nothing.

So the comparison shows you the trade-offs and tells you which option wins on each one, rather than pretending the decision is simpler than it is.