Module 4: Google and YouTube
Meta finds people who might want your product. Google catches people already looking for it. This module covers the other half of a buyer's toolkit: Search ads that buy existing intent, Performance Max used with your eyes open, and YouTube ads where the first five seconds decide everything.
Search: buying intent that already exists
Every search ad starts with someone typing what they want. That makes Search the most honest inventory in advertising — and the least forgiving. A person searching "emergency plumber open now" is ready to spend money in the next hour. A person searching "how to fix a leaking tap" is holding a wrench, not a wallet. Your first job is sorting keywords by how close the searcher is to buying, then pointing your budget at the closest ones first.
Build your keyword list from buying language: "buy," "price," "cost," "near me," "best [service] in [city]," brand and competitor names, and the exact product or service names you sell. Park the research-stage searches ("what is," "how to," "DIY," "jobs," "salary") on a negative keyword list from day one, and keep adding to it weekly — every irrelevant click you block is budget returned to searches that can convert. A small account with thirty tight keywords beats a sprawling one with three hundred loose ones.
Then make the ad and the landing page answer the same question. If the search was "vegan leather weekender bag," the headline should say that, and the click should land on that bag — not the homepage, not a general collection. This is message match, and it decides your click price and your conversion rate at the same time. Organize the account so each ad group holds one tight theme with its own matched page. When a client asks why a competitor seems to pay less per click, the answer is usually this: their ads promise exactly what the searcher asked for, and their pages keep the promise.
A replacement-window company gives you a fixed Search budget for the month. Which keyword group deserves the money first?
Performance Max, with your eyes open
Performance Max runs one campaign across Search, YouTube, Display, Discover, Gmail, and Maps, using the assets and signals you feed it. It can be genuinely strong — especially for e-commerce with a product feed and steady conversion volume. But you hand over real control to get that reach, and a professional knows exactly what they traded. You choose the budget, the conversion goal, the locations, the audience signals, and the assets: images, video, headlines, descriptions. You do not choose which keywords trigger your ads, which placements carry them, or how the budget splits between channels.
Because the system optimizes toward whatever conversion you define, your tracking setup is the steering wheel. Feed it junk and it will efficiently buy more junk. A lead-generation client whose form counts a conversion the moment someone hits submit — including bots, job seekers, and people who never answer the phone — will watch Performance Max proudly deliver more of exactly that. The fix is upstream: import qualified outcomes (the leads that became calls, quotes, and sales) and optimize toward those, or don't run the campaign type yet.
Work it with guardrails. Apply brand exclusions so it doesn't spend your budget winning searches for your own name and claiming the credit. Give it assets worth distributing — weak creative in Performance Max just fails in more places. Keep a standard Search campaign running beside it for your highest-intent keywords, where you keep full control of the message. And read its results in slices: if all the conversions come from branded search, it isn't prospecting, no matter what the dashboard calls it.
A roofing client's Performance Max campaign delivers cheap form fills, but the sales team says almost none of them pick up the phone or want a roof. What's the buyer's move?
YouTube and video ads: the five-second hook
On a skippable YouTube ad, your viewer gets a choice at five seconds, and most video ads lose them right there. So the ad is really two ads: a five-second hook that earns the next twenty-five, and a payoff that makes the hook true. The hook has one job — make the right person think "this is about me." Name the audience or the problem out loud in the first sentence, show the product or the result early, and put the brand in before the skip button appears, so even a skipped view leaves something behind.
If the hook works and the rest doesn't, people stay five seconds and drift. Watch the numbers as a story: the view rate tells you how often the hook earns attention; average watch time tells you whether the payoff holds it; click-through and conversions tell you whether attention turned into action. When viewers stay but don't click, the ad entertained without selling — the offer or the call to action needs work, not the opening. Test new hooks against the same payoff video before you reshoot anything else; the hook is where the leverage is.
Audience signals on YouTube work differently than on Meta. You're matching content and intent: channels and topics your buyer already watches, searches they've made, sites they've visited. Start with what a customer does in the week before they buy, and build from there. And treat your viewers as an asset — people who watched most of an ad are a warm retargeting pool you can follow up on YouTube, Search, and Meta with the next step of the argument. A video view that never converts still built you an audience you own a little more each week.
Your YouTube ad loses most viewers at the five-second skip. But viewers who stay watch about 80% of the video, and they convert well. Where's the problem?
What you can now do
- Sort search keywords by how close the searcher is to buying, and point budget at buying language first — price, cost, near me, brand names — before any informational query gets a dollar.
- Run Performance Max with your eyes open: name exactly what you control — budget, conversion goal, locations, signals, assets — and what you handed over: keywords, placements, and channel splits.
- Build a YouTube ad as two ads: a five-second hook that names the audience or the problem and shows the brand before the skip button, then a payoff that makes the hook true.