Shorts Campaigns Run Apart From the Rest of the Plan

Shorts inventory clears at a fraction of in-stream CPM, and dropping it into a shared campaign lowers your average cost per view while telling you nothing about whether anybody acted. We run Shorts on its own campaign, its own budget line and its own conversion window, without exception. In the first report, Shorts usually accounts for the majority of impressions and a much smaller share of measured action. That ratio is the finding. You should be able to see it rather than have it averaged into a comfortable blended number. The first two seconds: The thumb is already moving. During a flight we treat the share of impressions surviving past the second second as the working metric and re-cut against it weekly. Openings that begin with a logo, a drone shot or a spoken brand name lose that window in every category we have run. The first frame has to state the viewer's situation, the standing charge line on a bill, the deadline on an application portal, the van that did not arrive, before the advertiser appears anywhere. Vertical is not a crop: The bottom fifth of the frame carries the handle, the caption and the action button, and the right edge carries the interaction column. Anything placed there was never read by anyone. We re-block the shot for 9:16 rather than re-export it, and we decline to run a landscape asset pillarboxed with a colour bar, which reads as a mistake to the viewer and gets priced as one by the auction. Sound: Shorts plays with sound on. Muted-first design is a habit imported from feed advertising and it costs money here. We still burn in subtitles, sized for a phone held at arm's length, but the audio carries the argument. A voice that starts talking on frame one identifies itself as an advertisement immediately; we open on a sound from the situation and let the sentence land a beat later. What the low CPM conceals: Frequency behaves differently in a swipe feed, since the same person can be reached repeatedly inside one session. We cap by day rather than by week and watch the session-level distribution rather than the weekly average. Audience composition skews younger and more mobile than the rest of your plan, which means a Shorts audience insight does not transfer into in-stream targeting, and we will not hand you one dressed up as a general finding. We will not: Absorb leftover quarterly budget into Shorts because it can spend it quickly. Build trend-format creative for a regulated energy supplier, where the compliance review kills it in week two and you have already paid for the edit. Use trending audio without a cleared licence, whatever the platform's library appears to permit for organic posting. Report view counts as the outcome of a paid flight. Measurement: A same-audience comparison against your existing in-stream buy, run for at least four weeks, so the question answered is whether Shorts added action rather than whether it added impressions. Where volume allows, a geo split as well. Where neither is possible because the budget cannot power a test, we say so at the outset and run the flight as an explicit bet rather than pretending the report will settle it afterwards. Who this suits: Advertisers with an audience under forty and a short, concrete promise: an enrolment date, a fixed price, an appointment window. Education providers reaching applicants directly rather than their parents. It suits almost no B2B procurement cycle, and we would rather tell you that now than after the invoice.