A subscriber cancels in the gap. While a season is running the charge has something attached to it, because the show is the reason the card is on file. The cancellation comes in the weeks after the finale, when nothing has landed and the home screen looks like every other home screen.
Most of a streaming plan is built around the other end of that pattern. There is a launch, a burst of paid media, then a wait until the next one. The money arrives at the moment retention is least at risk, but the gap where the cancelling happens is the part of the year nobody has booked.
An insurer running a video ad interrupts nothing that competes with insurance. You are in a different position. Your ad reaches someone while they are watching something, and watching something is the behaviour you need them to start on your service instead. So the ad asks a viewer to stop doing the thing you want them doing, at the moment they are least willing to stop.
Launch advertising also teaches the habit that costs you the subscriber. Every campaign tells an audience to move to whichever platform has the new thing this month. They learn it, and then they use it on you, because cancelling costs nothing and coming back costs nothing.
Somebody reading a TV column or a weekend what-to-watch list has already asked the question. They opened the page to be told. A piece of editorial that answers it does not have to win attention off whatever else is on the screen, because there is nothing else on the screen.
That is where a library does work a trailer cannot. A piece about what to put on after a finale reaches the person sitting in the gap, and it can name something in your library that is six years old. "Binged House of the Dragon in two days? Reset with something light, we've got just the thing." That line works in a TV column. It does not work as a pre-roll, because the person watching a pre-roll is watching something already.
The same page is where you can show the product itself. Downloads for a flight, or a profile that picks up where somebody stopped on another device. None of that fits inside thirty seconds of trailer, and nothing kills the vibe like a clunky interface when all you want is to unwind.
A lot of that question now goes to an assistant. Someone describes what they feel like in their own words and gets back three or four names with reasons attached. Those reasons are assembled from review and recommendation pages on domains you do not own. Foundation Marketing and AirOps, May 2026, found that 90% of the sources AI cites when answering are ones the brand does not control. The market the sample covers is not stated. Read the study.
So a piece commissioned for the reader deciding tonight is also sitting in the pile an engine draws on when it is asked the same thing in March. Your own press page does neither job. We would start from the titles the engines already quote about television in your market, then work out which of them could carry your claim without it reading like an ad.
If the job is a date in everybody's head by Sunday, buy the ad. Editorial is slower than that. It books weeks ahead, and a publisher's own editors decide how the piece reads. It is a poor fit for a platform whose problem is that nobody has heard of it at all. Where it earns its place is the gap between seasons, and the back catalogue nobody opens.
A campaign is one claim carried across nine to thirteen articles, through three deliberately different kinds of publisher. We work with a network of 600+ publishers, 500+ of them in Australia, and a growing UK network, so we choose the mix from the titles your audience already reads rather than the ones we happen to have. You sign off the titles and the claim before any publisher is approached. The publishers then write the pieces themselves, which is what makes a reader trust one and what makes a model treat it as a source.
Cashrewards had a problem of that shape. Shoppers hear cashback and assume there is a catch, and no amount of Cashrewards advertising settles a suspicion about Cashrewards. Six publications those shoppers read by choice answered it instead. Read how that ran.
The streaming version is quieter. The viewer wants to know whether there is enough on here to be worth another month. You cannot answer that one about yourself. A TV writer can. Book the launch media, then look at the weeks after it, where nothing is booked and the cancellations happen.
10 September 2026 · 6 min read
Before any engine can quote your title, the page has to clear conditions that have nothing to do with the writing. The six that decide it, the two halves of a quotable page, and the crawler distinction most robots.txt files have never made.
10 September 2026 · 6 min read
OpenAI's own documentation says a sponsored placement sits below the answer and cannot change it. That makes the ad and the recommendation two different purchases, out of two different budgets.
9 September 2026 · 6 min read
The largest kind of source in an AI answer is often one nobody can buy. Here is the subtraction that shows what is actually left to commission in your category.
See how your brand appears in AI answers, and where trusted publisher content can strengthen its visibility.
