The passenger journey, scored

BC Ferries surveys its passengers at every stage of the journey and publishes the results by route, one score per touchpoint per year, in an appendix to its annual report to the Commissioner. The operator runs that programme, reports on it and says in the same documents what it is doing about the weakest scores. What has not been done is the reading from outside: eleven years of it in one place, set along the journey it describes, held against the operating record for the same routes and the same years, and tested to see which touchpoints actually move the overall score. That is what this page is.

01 What the survey covers, and what it does not

Scores run from 1, very dissatisfied, to 5, very satisfied. Every figure on this page is read from the published appendix and carries the page it came from. A dagger, †, marks a touchpoint whose question the operator reworded at some point; hover it for what the report says. Those breaks are marked rather than smoothed over, because a reworded question is a new question.

02 The journey, stage by stage

Each stage lists its touchpoints worst first, because the worst score is the one that decides whether a passenger remembers the trip. The trend runs across the whole published history for that touchpoint, scaled to its own range, with that range printed beside it. Two stages have an operating measure that bears on them, drawn underneath on its own scale and named: those are two separate records of the same years, not one explaining the other.

03 Which touchpoints actually move the overall score

A satisfaction programme is only as good as its priorities, and the usual way of setting them is to correlate each touchpoint against the overall score and rank the result. That answer is often wrong, for a reason worth stating plainly.

So each touchpoint is tested twice. Across route years asks whether routes that score well here score well overall. Within a route asks the harder and more useful question: when this touchpoint moved on a given route, did that route's overall score move with it. Where the two disagree, the disagreement is the finding.

04 Where to spend, and what to leave alone

Impact against current score. A touchpoint that moves the overall score and is already scoring well is being done right and should be protected, not improved. One that moves the overall score and scores badly is where the money goes.

05 The weakest point in every stage

This is the first row of each table in section 02, gathered in one place. There is no operating figure beside it because there is not one to put there: that is the point of the section rather than an omission from it.

07 Does the felt experience match the measured record?

Every figure so far comes from the survey, which means the survey has been checked only against itself. The rest of this corpus is the operator's own operating and financial record for the same routes and the same years, read from the same filings by different extractors. Where a touchpoint has an operating figure that ought to move with it, the two can be set against each other. That is the only test on this page that uses something other than a survey to check a survey.

The survey waves run in August and November, so a wave labelled with a calendar year falls inside the fiscal year ending the following March, and every pairing here maps it that way. Getting that offset wrong by one year would quietly destroy any relationship there is. Money figures are per passenger and in fiscal 2026 dollars, deflated by the archived Bank of Canada series; nominal dollars over this span would describe inflation and call it spending.

08 What this rests on

Two limits worth stating rather than burying. The survey covers the busiest routes only, so nothing here speaks for the minor and northern routes, which is most of the network by route count and the part of it where service is thinnest. And a correlation between a score and an operating figure is not a claim that one caused the other; the operating series are drawn beside the scores so a reader can see the shape, not to assert a mechanism.