Which channels produced installed jobs, not quote requests?

A typical enquiry is a daughter in one city calling about her father's stairs in another. She researches, she calls, and she often pays. The survey is booked in the parent's name, the invoice may be in either name, and the delivery address is the parent's.

Any join on name or address fails on a large share of these, and the usual conclusion is that paid search produced nothing. Matching on the phone number and email that appear on both the enquiry and the job record recovers them, and flags the rest for a person rather than guessing.

Home accessibility generates a lot of information-seeking enquiries: people comparing a stairlift against moving house, or pricing a walk-in tub a year before they need one. Some convert months later, many never do.

Ranking channels on enquiry volume therefore rewards whichever channel attracts the most browsing. Ranking them on installed revenue — two thousand to ten thousand dollars a job — is a different list, and it is the one that decides next quarter's budget.

Most dealers in this trade sit inside somebody else's network. A manufacturer runs national advertising, takes the enquiry, and passes it to whichever dealer holds that territory. The enquiry arrives with a name, a telephone number and a postcode, and sometimes a note about what the caller asked for. What it never arrives with is any account of what it is worth. The dealer is invoiced for the lead, or gives up margin on the unit, and has no way of comparing that cost against the enquiries their own advertising produced last month.

The comparison is not difficult once installed revenue is attached to both. The manufacturer's lead list uploads as a source file, the dealer's own campaigns upload as another, and both are ranked on what the jobs actually invoiced. Because the match runs on the customer's telephone number or email, an enquiry passed across in March and an installation completed in June are one row, regardless of which system either half was sitting in. Neither system has to be touched, and neither party has to agree to anything for the dealer to run the comparison.

Dealers who run this tend to find the answer varies by product rather than being uniform. Network enquiries may dominate one line and be worthless in another, and the dealer's own local advertising may quietly carry the bathing conversions while getting no credit anywhere. None of that is visible while the two sources are counted in different units, one in leads supplied and the other in clicks bought. The network will not produce that breakdown, and it has no particular reason to. Only the dealer holds both halves of the arithmetic.

A straight stairlift goes in, and eighteen months later the same household orders a second one for the back stairs, or a grab-rail package, or a ramp when the wheelchair arrives. Rental units come back, get reconditioned and go out again. Service contracts renew annually, batteries get replaced, and a curved rail is eventually bought outright by the family who started on a rental. All of it belongs to whoever produced the original call. None of those later orders required a second advertisement, and most of them required only that the first installation went well.

Conventional reporting sees none of this because each transaction opens as a new job with its own paperwork. The channel is credited once, at the lowest value the relationship will ever have, and every subsequent order appears to have arrived from nowhere and is quietly filed under repeat business. Over a few years that understates the productive channels by a considerable margin. Ten years of a dealership's best customers can sit in that undifferentiated pile, invisible to every report the business runs. The channel that found them is credited with one straight lift.

Matching on the household keeps the whole sequence attached to one acquisition. A product column then separates lifts, bathing, ramps and service so the pattern is legible: which sources produce customers who come back, and which produce a single job and silence. That is a different ranking from cost per install, and it is the one that should be setting a dealer's budget. It also tells a dealer something about which products to lead with, because the entry job that produces repeat orders is not always the one with the best margin on the day.

Families in this market almost always get more than one quote. A surveyor visits, measures the staircase, produces a price, and then the household discusses it for a week or two while two competitors do the same thing. Whoever can fit the job soonest frequently wins it, and the difference between winning and losing is often a matter of days rather than of price. The survey itself costs the dealer a technician's morning either way. A quote that arrives third, on a Friday, after the family has already agreed with somebody else, has cost as much to produce as the one that won.

Because a booked survey is countable within hours and an installation is not, the survey becomes the metric. Campaigns get optimised toward whatever produces the most appointments, which is generally the cheapest and least committed enquiry, and the surveying diary fills with visits that were never going to convert. The dashboard improves while the installation schedule does not. The surveyors notice long before the marketing does, because they are the ones driving to the appointments. What they cannot do is put a figure on it.

Ranking sources on invoiced installations rather than on surveys booked corrects it, and exporting both dates makes the lag explicit. A source producing enquiries that convert in ten days and one producing enquiries that convert in fifty can carry the same cost per survey, and in a trade where the fastest fitter usually wins, they are not remotely equivalent. Speed is a purchasable advantage in this trade, but only if somebody can show which sources are worth being fast for and which are filling the diary with browsing.