It is the second week of November. Three rooms are occupied, the summer staff left in September, and the owner is at the desk with a laptop doing the things that piled up since May: last season’s reconciliation, forty reviews nobody replied to, a rate sheet for next year built mostly on a feeling, and an enquiry that came in at eleven last night from someone comparing four properties.
Destination BC’s regional numbers put Vancouver Island hotel occupancy at 88% in August and 41% in January. Annual occupancy across the region was 68% in 2024 with an average daily rate of $263.
Behind that ratio is the actual shape of the work. In August the property is full, the staff are in, and the problem is throughput. In January the property is two fifths full, the staff are gone, and the same owner is still answering enquiries, chasing deposits, reconciling last season, replying to reviews, planning next year’s rates, and doing the paperwork that has accumulated since May. The volume drops and the job list does not.
That asymmetry is the reason this piece exists. Most technology aimed at accommodation is aimed at the busy month, at filling rooms and managing channels. The work that actually wears an operator down happens in the other eight, and it is administrative, repetitive, and made of information that already exists somewhere in the business. That is precisely the shape of work that has become solvable.
The labour arithmetic underneath
Two sets of numbers explain why the shoulder season is where this bites.
On staffing, go2HR’s provincial labour market research, surveyed in 2021 and published in March 2022, found 97% of BC tourism employers anticipating difficulty filling positions, with an average difficulty rating of 4.3 out of 5. Seasonal worker return rates ran 49% in a typical year and fell to 38% in 2021. In accommodation specifically the typical-year rate was 39%, and in 2021 it was 32%. That survey was fielded in 2021, so read the direction as durable and the exact figures as a snapshot. Around three in five of the people trained last summer are new people this summer, which means the training cost recurs annually and the institutional knowledge does not accumulate.
On the owner’s own time, CFIB reported in 2023 that small business owners work 54 hours a week on average and 59 where labour shortages bite, with 84% of affected hospitality owners reporting increased hours. Separately, CFIB’s 2024 red tape research found Canadian small businesses spending 735 hours a year on regulatory compliance, of which 256 hours is unnecessary, and the burden falls hardest on the smallest: businesses under five employees spend 198 hours per employee against 8 hours for firms over one hundred.
Set those together and the picture is clear. The person doing the administrative work at a small property is usually the owner, usually outside normal hours, and usually in the months when there is no revenue coming in to justify hiring help. The constraint is not headcount in July. It is one person’s attention in November.
Where the return actually is
Four areas hold up under examination. They share a property worth naming, which is that in each case the raw material is already being produced by systems you already pay for, so nothing new has to be collected and every answer can be checked against a source.
The enquiry that arrives at eleven at night
An enquiry from someone comparing four properties is a perishable thing. There is honest industry data on this from outside hospitality: a study of over 15,000 leads and 100,000 call attempts found the odds of making contact dropping roughly a hundredfold between a five-minute and a thirty-minute response, with a steep decline across the first hour. That study is from business-to-business sales and from 2007, so treat it as an analogy rather than a hospitality statistic. The mechanism it describes is familiar to anyone who has watched a booking go elsewhere overnight.
What is worth building here is narrower than a chatbot on the website. An enquiry-handling step that knows your actual availability, your actual rates, your actual policies on dogs and late arrivals and the road in winter, that answers accurately at eleven at night, and that hands anything unusual to a person in the morning with the context attached. The value is in the accuracy: a generic assistant that guesses your cancellation policy creates a problem rather than solving one, which is why this only works when the system is wired to your real booking data rather than to a description of your property.
The reviews nobody has time to answer
This is the best-evidenced item in the article, and it is the one most often left undone.
Michael Luca’s Harvard research found that a one-star increase in Yelp rating produced a 5% to 9% increase in revenue for independent restaurants, with an effect close to zero for chains. Chains have brands. Independents have reviews. Cornell’s hotel research found that a one-point rise on a hundred-point review index was associated with up to 0.89% higher average daily rate, 0.54% higher occupancy, and 1.42% higher revenue per available room. Both studies date from 2011 and 2012 and remain the standard citations in the field, so read them as established mechanism rather than current measurement.
On responses specifically, Tripadvisor’s research with Ipsos MORI across 23,292 users in twelve markets found 77% more likely to book when owners give personalised responses, 63% more likely when owners respond to most reviews, and 89% saying a thoughtful reply to a negative review improved their impression.
The obstacle is never that owners disagree with this. It is that writing a specific, non-generic response to every review, including the unfair ones, in August, is not going to happen. A drafting step that reads the review, pulls the relevant booking so the response can reference what actually occurred, and produces a draft in your voice for you to edit and post turns a job nobody does into a few minutes of editing. The draft stays a draft. A published response written entirely by a machine reads exactly like one, and the research above is about personalised responses, so the person has to stay in it.
The cancellation and deposit question
The cancellation numbers vary enormously by how a booking arrives. D-EDGE’s distribution analysis found that in Europe in 2023, about 18% of direct-booking revenue was cancelled against 42% of Booking.com revenue, and separately that prepaid bookings cancel roughly half as often as average. SiteMinder’s analysis of over 130 million bookings put the global cancellation rate at 19.15% for 2025, alongside an average booking lead time of about 32 days.
The honest position is that reliable no-show data for small independent operators does not appear to exist publicly, and the figures circulating on vendor blogs cite each other rather than any primary source. What the distribution data does establish is that your cancellation rate is partly a consequence of how you sell rather than a fixed feature of the business, and that is worth knowing before anyone proposes software for it.
Where a system helps is in the sequence around a booking rather than the booking itself: the deposit that gets requested on a schedule, the confirmation that goes out at the right interval, the pre-arrival message that catches the guest who has changed plans while it can still be resold, the flag on a booking pattern that historically cancelled. That is a multi-step job with a person on the consequential end, which is a sequence rather than a single task, and it is the kind of thing worth building once and running for years.
Three seasons of your own data
Every property has a booking history nobody has ever properly read. When did the shoulder actually start last year against the year before. Which rate changes moved occupancy and which moved nothing. What does a Tuesday in May actually earn. Which channel brings guests who cancel. What is the real cost of opening in the second week of April rather than the first.
Destination BC’s regional profile notes that while most overnight travellers visit between July and September, the April to June shoulder accounts for a considerable share of visitor nights. SiteMinder’s global analysis found peaks flattening, with 65% of markets seeing their busiest month take a smaller share of annual arrivals in 2025. Both suggest the shoulder is where the decisions have moved, and the shoulder is exactly where a small operator’s instincts are least tested, because it is the part of the year with the fewest data points in anyone’s memory.
This is the data you already own and are not using. The work is unglamorous: getting three or four seasons of bookings, rates and cancellations into one place where the questions can actually be asked. It is also the item on this list with the longest-lived benefit, because it compounds every year you keep collecting.
The thing to check before anyone quotes you
One practical warning, because it decides what is buildable.
Booking and property management systems for small operators vary widely in what they let anything else connect to. Checkfront, founded and headquartered in Victoria and now part of a larger group, publishes API documentation while describing the current version as in maintenance mode and providing access through a sales conversation rather than to trial accounts. Cloudbeds documents its API publicly and gates production access behind an enabled partner developer account. Little Hotelier, owned by SiteMinder, published no developer API under its own name as of August 2026, and integrations run through SiteMinder’s platform instead. Lodgify publishes a public interface for its own users. Newbook, common in RV parks and campgrounds, is now owned by Storable. All of that is drawn from each vendor’s own public documentation as read in August 2026, and vendors change these terms, so confirm directly before planning around any of it.
The pattern across the category is documented API, gated access. That is rarely a technical obstacle and frequently a commercial one, and it holds well beyond booking software. It is the first thing worth establishing, because a plan that assumes your booking system will hand over its data depends on a commercial decision you do not control. Anyone scoping this work who has not asked that question yet has not started.
The part worth carrying out the door
The gap between 88% in August and 41% in January is the defining fact of an Island accommodation business, and the work that gets left undone lives in the trough rather than the peak. The enquiry that goes unanswered overnight in November, the reviews from August that never got a reply, the reconciliation that takes a week in February, the rate decision made on a feeling because nobody has time to read three seasons of history.
Each of those is made of information the business already produces. None of them requires new data collection, new staff, or a change to how guests are treated. What they require is someone deciding which one costs the most and building for that one specifically, with a person still on the end that matters, because a guest can tell the difference between a considered reply and a generated one.
Start with the review responses if you want the fastest visible return, and with your own booking history if you want the one that keeps paying.