What each channel costs, answered in full.

No, and it matters most on this page. AdBuyMCP is in its design-partner phase: no buying rail has reached spend-tested or report-verified on our own evidence ledger, so live spend runs inside a design-partner engagement rather than from a self-serve top-up. The whole loop — brief, seven compiled plans, creative, approval, launch, delivery, measurement — runs today in a sandbox that needs no credentials and no card, and the numbers on these pages are what a plan would carry.

They measure different things and both apply. The plan floor is enforced when AdBuyMCP builds a plan: a line that cannot clear it is dropped and its budget redistributed, with a note saying so. The market minimum is what the seller will actually accept for a campaign, which on Sky AdSmart is around £3,000 against a £1,000 plan floor. The floor stops a pointless line surviving into a plan; the seller's minimum stops a pointless campaign reaching air.

Three things, all named per channel. Data segment uplifts, which on a targeted business audience add roughly £2.50 to £3.00 per thousand impressions — the sellers' cost, not ours, and shown on the plan before approval. Creative production, whether generated here at published rates or produced elsewhere. And clearance time on the channels that need it, which is a lead time rather than a fee but is the constraint more often than money is.

Against a UK agency, favourably on fee and not on scope: agencies charge 10% to 30% of spend, most commonly 15% to 20%, with £500 to £1,000 minimum monthly fees — and they are also selling strategy, creative judgement and someone accountable, none of which this is. Against the platforms, it depends: The Trade Desk's observed take rate is about 20% and DV360's around 10% to 15%, while Meta and Google's own managers are free. The comparison that matters is the whole waterfall, not the headline percentage.

No. Sky AdSmart is a booking-workflow rail here: we generate the brief with the right household attributes, drive the clearance step and hold the order state machine, but we do not make the booking and it does not report the delivery. No addressable-TV delivery statement has been ingested through it. The supply page grades it accordingly.

Because they measure different things. £1,000 is the point below which an addressable line stops being worth carrying inside a multi-channel plan, so the planner drops it and redistributes the budget. £3,000 is what the seller will actually accept for a campaign. Both apply, at different moments.

Not in live mode, and this is worth being exact about. Live activation on the DOOH rail requires hand-picked unit identifiers, because venue and postcode semantics have no verified translation into vendor ids yet, and the other DOOH connector has no live client at all. The compiled plan, the venue selection and the sandbox loop all work. The live buy needs more than a human to pick the units: on that rail live creative submission and report pulls both fail closed as well, and no DOOH rail has reached spend-tested or report-verified in any case.

If search is your only channel, no. Google and Microsoft both provide free management interfaces and a percentage on top buys you nothing you cannot do yourself. The case for paying a layer starts when search is one of three or more channels and you need the audience deduplicated across them, the fee waterfall in one place, and a causal test that spans all of it rather than per-channel attribution that double-counts.