Which channels produced second-home buyers who closed?

Nobody needs a second home. The purchase happens when a family's circumstances, a market and a particular property coincide, and the browsing that precedes it can run for two or three years across dozens of visits and several agencies.

That makes last-click reporting close to worthless: it credits whichever branded search happened in the final fortnight and ignores the portal listing that first put the area in the buyer's mind.

Reconciling closed commission against the full enquiry history recovers the earlier touch, provided the buyer enquired at some point. Where they never did, the report says so rather than inventing a path.

Resort and coastal demand comes from identifiable feeder cities, and the mix changes with flight routes, remote work patterns and the tax treatment of second properties. Knowing which feeder markets produce closings rather than enquiries is the single most actionable output.

A buyer-origin column on the closing export makes that visible. Agencies routinely discover that one or two feeder markets produce most of the commission while the advertising budget is spread evenly across six.

Second-home buyers do not buy from a website, however long they browse one. They buy on a Wednesday afternoon, three days into a trip, standing on a terrace with a view they have been imagining for two years, having driven around the area with an agent who knows which villages have a doctor and where the road floods in February. The entire remote phase exists to produce that trip, and the trip is where the decision is either made or quietly abandoned.

Agencies understand this perfectly well and organise themselves around it, often paying for flights or a night or two of accommodation to get a serious prospect on the ground at all. What they cannot currently do is work out which channels produce people who actually come, as opposed to people who say they will and then do not. A source generating enquiries in volume but very few confirmed trips is expensive in a way no cost per lead will ever reveal, because the cost of the trip programme and of the agent's lost days is carried somewhere else in the accounts and never attributed back to the channel that caused it.

The inspection trip list is a short file with names, contact details and dates, and it uploads as a source file exactly as a campaign export does. Closings are matched back to it on the buyer's own contact details, so each channel can be read on the trips it produced and on the commission those trips eventually generated, months or years later. For an agency deciding whether to keep subsidising visits, and for which sources, that pair of numbers side by side is considerably more useful than the enquiry counts it currently argues from every budget season.

Agencies in resort markets very often manage rentals as well as sales, and the two sides sit in different systems, run by different people, and almost never speak to each other about anything except keys. Yet a guest who has taken the same villa for three consecutive summers has done more due diligence on the location than any buyer enquiry ever will. They know the drive from the airport, the noise in August, which beach is worth it and what the winter is like when the restaurants close.

That person is comfortably the most qualified prospect the business will handle all year, and they are sitting in the rental ledger already, with a name, an email address, a phone number, a payment history and a documented preference for exactly the sort of property the sales side is trying to sell. Nobody has ever thought of the guest list as a marketing channel, because it was compiled for invoicing and for sending out the key code, not for selling anything, and the two departments have never had a reason to exchange a file.

Uploading it as a source file changes that immediately and costs nothing beyond running an export. Closings are ranked against the guest list alongside portals, paid search and referrals, on commission earned rather than on enquiries generated, over whatever period the agency chooses. Agencies that run both sides of the business are routinely startled by how much of the year's sales revenue traces back to somebody who first arrived as a tenant with a suitcase. It also puts a defensible value on the rental book that has nothing whatever to do with the management fees it generates.

Resort transactions are shared more often than domestic ones. A buyer's agent in their home market takes a referral cut for the introduction, a co-operating agency in the resort splits the local side, a developer pays a different rate on new inventory than the resale market pays, and in some markets a finder who brought the buyer to the door takes a slice as well. The same headline percentage can leave the agency holding very different amounts on two apparently similar sales.

New-build and resale therefore need separating before any channel ranking means very much at all. Developer stock typically pays a higher rate and arrives with marketing support, cooperative advertising and a sales centre the agency did not have to build or staff. Resale carries the full cost of winning the instruction, photographing it, listing it across several portals and holding the owner's hand for nine months while nothing happens. Blending the two produces an average commission that describes neither half of the business, and it moves whenever the mix of available stock moves, which in a resort market is constantly.

Exporting retained commission with a stock-type column beside it settles both problems at once, using fields the agency's own accounts already distinguish. The report ranks channels on what the business actually kept, split between the two halves of the inventory, so a source producing referred new-build buyers is compared honestly against one producing resale buyers who arrived directly. Those are frequently different channels entirely, and funding them as though they were one is how an agency ends up over-invested in the cheaper half.