A list of YouTube channels can look like a great audience plan. It isn't one. It tells you what you want the ad to run alongside, not who is watching or whether they’re in a position to buy.

I’d start by writing down what you want to learn. Are you testing whether a certain group responds to the video, or whether the video works better next to a certain kind of content? Pick one for the first test. It makes the results a lot easier to work with.

Who sees it versus what they’re watching

Audience targeting starts with people. That could mean a relevant in-market segment, a custom segment, or an eligible segment built from your own data. Google’s custom segments guide makes a useful distinction: entering a website URL as an audience input doesn't put your ad on that website. And using search terms to help define an audience on eligible Google surfaces doesn't turn the video into a Search ad.

Content targeting starts with what the ad appears alongside. Where the campaign supports it, you can choose specific YouTube channels or videos as placements, or use topics or video keywords for a broader context. In eligible Video campaigns, Google can match one of the selected content types rather than requiring all of them. So if you want to test a list of placements, adding a broad topic may change the test. Check Google’s Video targeting guide for the controls that apply.

Check what the campaign is built to do

If the job is exposure, start by looking at Video reach options. If the job is getting people to watch, look at Video views. Reach and views answer different questions. Neither, by itself, tells you whether you’ve found buyers.

If the job is leads or purchases, I’d evaluate Demand Gen with video assets. Google says the old Video Action campaigns were upgraded to Demand Gen by May 2026. Don’t use an old Video Action tutorial as a checklist for a new campaign.

For a YouTube-only Demand Gen test, check the ad group’s channel controls and select the eligible YouTube inventory you intend to test. Selecting YouTube Shorts, for example, is a choice about inventory. It isn’t a list of individual videos. If you need specific placements, verify that the selected setup offers positive placement targeting before promising where the ad will run.

Also look at optimized targeting in Demand Gen. When enabled, the selected audience can be a starting signal rather than a boundary. That may be fine for finding more converters. It’s a different test from asking how one selected audience responds.

A small test you could actually read

Hypothetical example: A company has a short walkthrough of its scheduling software for repair shops. It wants to learn whether the video earns attention in the context of shop-management tutorials. I’d start with a Video views setup that supports the intended placements and build an eligible list of relevant channels or videos. I wouldn’t throw in a broad automotive topic just to increase delivery. That could bring in people watching car reviews, which is a different context from running a shop.

If the list barely serves, that’s a delivery problem to investigate, not proof the message failed. Check placement eligibility and whether there’s enough inventory before changing the video. If it serves but viewers don’t take the next step, look at the opening and the offer before deciding to widen targeting.

If the real job is qualified demos, I’d write a separate brief. Use a campaign built around that action, define what counts as a qualified demo, and check it against the lead record. Don’t compare a cheap view with the cost of a booked meeting and call one campaign the winner.

Write down the boundaries before launch

  • Question: Are we testing the people, the viewing context, or the video? Pick one primary variable.
  • Setup: Record the objective, campaign type, chosen formats and inventory, audience or content inputs, exclusions, and any expansion settings.
  • Evidence: Decide what would make the test usable. For a views test, review delivery and view metrics. For a lead test, record the conversion event and who checks lead quality.
  • Next move: Agree on what would justify widening the test, revising the video, or waiting for more evidence.

Once it runs, look at the Where ads showed report where available. Google notes that it doesn’t list every placement, and some YouTube inventory appears in aggregate. Use it to spot a delivery mismatch, not to prove that every impression landed exactly where you expected.

If you already have video and want help deciding who should see it, where it should run, and what to check after launch, that’s part of my YouTube Ads management work. If you’re still deciding whether the business needs a video campaign or a Search test, start with Demand Gen Ads vs. Search.

Platform details checked November 11, 2026 against the Google Ads Help pages linked above. Available controls vary with campaign setup and eligible inventory. The example is hypothetical.