Every Q4, someone in a budget meeting points at the video line and asks what it’s actually doing. Nobody has a clean number, so upper funnel video advertising gets cut. For the next couple of months, that looks like the smartest call of the year. Cost per lead drops, leads hold, finance is happy.

Then spring comes, lead volume slides, and nobody traces it back to that meeting. We see this pattern often enough that we built our Performance Video Network (PVN) around one rule: every program launches with untargeted holdout markets, so video has a number before anyone asks for one.

What happens after you cut upper funnel video advertising?

In our experience, search gets more expensive and converts worse, starting a few months after the cut. The timing varies by category and sales cycle, but the sequence is consistent.

Video is what puts your name in someone’s head before they need you. When a window starts leaking, they search for the brand they already recognize. Search catches that demand, but video created it.

Cut video and the familiarity you built keeps converting for a while. Attribution credits whatever the buyer touched last, usually search, so the cut looks smart.

Then the familiarity runs out. Branded search volume slips, so more of your search spend shifts to costlier non-brand terms and blended cost per click climbs. Appointment rates drift down, usually inside normal variance. Within two quarters, lead volume falls and cost per acquisition (CPA) rises. The post-mortem blames search, seasonality, or the landing page. Video was cut two quarters ago, so nobody connects the two.

Restarting video after stopping has its own cost, which is a separate conversation. The point here is that the cut’s price arrives late and under a different name.

So video stays on, at some level, but how much?

How should you split budget between video and search?

Fund them as one program and let one team move the money between them week to week.

Most brands buy linear, CTV, YouTube, social, and search as separate line items, each with its own budget, its own report, and its own owner. When the quarter tightens, video and search fight over the same dollars, and video loses because its report looks weaker.

Plan them against one audience and one scorecard and the fight goes away. CTV and linear put your name in the living room. YouTube and social follow the household while they research. Search harvests the demand video created, and branded terms clear cheaper when that demand exists.

Planning them together also exposes overlap. Where the platforms allow it, you can cap exposures per household across CTV and YouTube so no single channel dominates. Linear can’t be capped at the household level, so it gets planned around instead, with its reach set first and digital filling the gaps. Either way, you stop paying more than once to reach the same house.

How do you prove your CTV ads are actually driving sales?

Hold some markets back from video and compare revenue against them. Finance trusts that number because it doesn’t depend on platform attribution. It compares markets that saw the ads against comparable markets that didn’t, using your own sales data.

Two things make it credible. First, the markets have to be matched on baseline sales, seasonality, and size, or the comparison means nothing. Second, the test has to run long enough to catch the lag described above. A four-week holdout tells you whether video moves sales now. Only a longer one tells you what happens when you stop.

We built PVN so that every screen runs as one program with one budget, measured against holdout markets, so video never walks into a budget meeting without a number.

If you want to know what your current setup is costing you, we’ll audit it. Bring your channel mix and your last six months of data. We’ll show you where last-click reporting is hiding video’s contribution, where you’re paying more than once to reach the same household, and what a holdout test in your markets would look like. Get your free audit here.