Short-form for brands nobody can buy online.
Many premium brands we talk to cannot sell you anything directly. The sale belongs to a dealer, a distributor or a showroom two steps down the chain. Most conclude short-form is a toy for e-commerce. Wrong conclusion. The purchase stayed in the store. The decision left it years ago.
The shift: the purchase stayed in the store, the decision did not.
A 3D Cloud and Provoke Insights survey of US furniture buyers (2025) found 46 per cent went to a website first, then bought in store. Provoke Insights (2024) is blunter: 71 per cent prefer to buy a sofa or sideboard in a shop. In major appliances, NielsenIQ's Home Appliances Outlook 2026 (2025) puts online at roughly 30 per cent of global sales; seven in ten still leave a store. The till did not move. The mind did.
Google's messy middle research (2020) describes what happens before the till: people loop between exploring and evaluating until something tips them. Creator content and reviews shorten it by six days on average (Google/Material, 2023). McKinsey's State of the Consumer (2026): 34 per cent of Gen Z say social media plays a key role in purchase decisions, against 16 per cent of boomers. The dealer sees the last ten minutes of a decision that took weeks. Short-form is the other weeks.
Reframe: you are not selling online, you are booking the showroom visit.
The job of short-form for a dealer-sold brand is not conversion. It is to make yours the name asked for in a multi-brand showroom. Not product films, then. Proof: a drawer runner under a phone torch, a spirit poured in a real bar, a tiler on why one porcelain does not chip. What survives the feed's organic test earns the recall. Paid then puts it in front of the postcodes where your dealers are.
Every video needs a next step without a basket: find a dealer, book a consultation, request a sample, ask where to see it. The dealer-sold brands in our portfolio run on this: the video starts a route, it does not end one. Our record is above 100 million organic views for one brand; the number a dealer-sold brand should watch is enquiries.
Short-form does not sell the sofa; it decides which name gets asked for in someone else's showroom.
What the platforms actually give you in 2026.
Less than the decks imply, more than most brands use. Meta retired the Store Traffic objective in 2023; the rest sits under Awareness: store sets, map cards with the nearest location, offline events through the Conversions API. TikTok never had a store visits objective. It offers district-level targeting, Lead Generation with native forms that sync to a CRM (TikTok, January 2026) and an offline Events API with a 28-day click window (TikTok, 2025).
Neither platform will optimise towards a showroom visit. Both take dealer leads and matched sell-out as signals and let you target the towns where you have a door. One condition: the brand owns the lead form and the dealer locator, not the distributor. If the enquiry lands in a dealer's inbox and never comes back, the platform learns nothing. Nor do you.
How to prove it when the till is not yours.
Use geography, the one thing dealers and platforms share. Vaver and Koehler at Google (2011) showed that randomising regions into test and control measures incremental sales without tracking anyone, and that offline response lags, so the test must outlast the spend. Meta's open-source GeoLift applies synthetic control to the same idea. For a brand with 40 dealers: switch content on in half the territories, read distributor sell-out by region for a quarter. The read is sell-through, not views.
Where a geo test is impossible: TikTok's InMarket partnership (2024) measured store-visit lift across 126 retail and dining campaigns and found significant lift in 121. NielsenIQ's retail media incrementality analysis (July 2026) is honest about the rest: retailers rarely support randomised or geo-lift tests, so credible measurement is an ensemble of methods, and ROAS measures efficiency, not causality.
Where this argument stops.
Two places. First, when the dealer network captures the demand and the brand never sees it. No sell-out by region, no log of where enquiries came from: no experiment, only a view counter. Second, when the purchase cycle outlasts the budget cycle. Lewis and Rao (2015) analysed 25 large advertising experiments: the median test could not tell a 10 per cent return from zero without a campaign 61 times larger. A kitchen bought once in fifteen years proves nothing inside a quarter. Then we say so before the brief and build for recall and dealer pull, not a number nobody can defend.
Sold through dealers and showrooms?
Tell us about your brand and your goal. We will come back with a plan and the first step.
Book a call