Google Ads beyond search: Shopping, Performance Max and YouTube
Search is where Google Ads starts, not where the budget ends. How Shopping, Performance Max and YouTube really work, and why every pound still answers to margin, not the platform's idea of success.
The search box is not the whole account
Most people picture Google Ads as the block of text ads above the organic results, and for years that was where nearly all the money went. It is still the sharpest tool Google offers, because someone typing a specific query has told you exactly what they want. But the platform has spent the last few years pushing budget outwards: onto the Shopping carousel, across the wider web and apps through Performance Max, and into video on YouTube. Each of these does a genuinely different job, and each fails in a different way when it is left to the platform's defaults. The mistake is to treat them as one thing because they share a login. Search harvests intent that already exists. Shopping puts your actual product, with its price and photo, in front of a shopper who is comparing options. Performance Max spreads a single budget across every surface Google owns. YouTube reaches people who are not searching at all. Run them with the same assumptions and you will overpay for the ones that create demand while starving the ones that capture it. The discipline is not to spend everywhere, it is to know what each surface is for and hold each to its own standard.
Shopping is won in the feed, not the bid
Google Shopping looks like advertising and behaves like SEO. There are no keywords to bid on directly. Instead, Google reads your product feed, decides which queries your items are relevant to, and shows the ones it thinks will win. That means the feed is the campaign. A product title that leads with the brand and model a shopper actually types, a correct GTIN, a sensible product type and a clean Google product category will pull more qualified traffic than any amount of bid tinkering on a thin, badly structured feed. This is where most Shopping budget leaks. Items sit disapproved over a missing attribute and never serve. Titles are written for the website rather than for search, so they match loosely and attract browsers rather than buyers. Out-of-stock lines keep spending because nobody reconciled the feed against live inventory. The work that pays is unglamorous: fix the feed, segment products by margin so your best-selling, best-margin lines are not averaged in with the rest, and use the search terms and negatives to stop your budget answering queries you cannot profitably win. Do that and Shopping becomes the most efficient way to sell a physical catalogue on Google. Skip it and you are paying to show the wrong products to the wrong people.
Performance Max: one budget, every surface, little visibility
Performance Max is Google's automated campaign type. From a single asset set and budget it can serve across Search, Shopping, Display, Gmail, Discover and YouTube at once, and it decides the split for you. Used well it is a powerful way to scale once the economics are proven. Used carelessly it is the easiest place in the whole account to waste money, because its defining feature is that it hides where the money went. You hand it a budget and a goal, and it optimises to that goal by whatever route is cheapest for the algorithm, which is not always the route that grows your business. The classic failure is cannibalisation. Left unchecked, Performance Max will happily spend on your own brand terms and on people who were already going to buy, then report those cheap conversions as its own success. The contribution looks excellent because it is claiming credit for demand you already owned. Containing it takes real structure: brand exclusions so it stops buying traffic you get for free, account-level negatives, disciplined asset groups, and a conversion goal that reflects genuine margin rather than raw form fills. Treat Performance Max as something to govern, not something to trust. It earns its place when it is scaling proven demand, not when it is quietly harvesting yours.
YouTube is a demand channel, not a search channel
YouTube is the surface people misjudge most, because they measure it as though it were Search. It is not. Almost nobody watches a video ad and buys in the same session, so a last-click view of YouTube will always make it look like it does nothing. What it actually does is create demand: it puts the right message in front of the right audience before they are searching, which is the part of the funnel that Search, by definition, cannot reach. If you only ever harvest existing intent, you are competing for a fixed pool of demand and wondering why it does not grow. That changes how you run it and how you judge it. Creative carries the channel, so the work is in the hook, the message and the format, not in bid tweaks. Targeting should be deliberate rather than sprayed across every placement Google will sell you. And the measurement has to be honest: judge YouTube on whether it lifts overall enquiries, search volume and revenue over a period of weeks, not on the view-through conversions the platform hands you, which flatter it. Run as a demand engine with a real read on incrementality, video widens the top of the funnel. Run as a discount last-click channel, it looks like a waste because you are asking it to do a job it was never meant to do.
The measurement that keeps it all honest
None of this works without trustworthy conversion data, and this is the step that quietly decides whether every campaign above is optimising to something real. If Google is fed browser-side conversions that fire on a thank-you page, it will optimise to whatever is cheapest to make that page load, which is rarely your best customer. Server-side tracking and offline conversion imports change the target. When you feed real revenue, or qualified pipeline from your CRM, back into the account, Google optimises Shopping, Performance Max and YouTube towards the outcomes that actually pay, rather than towards the easiest-to-generate form fill. The second half of honest measurement is incrementality, and it matters more the further you move from Search. A last-click model over-credits the surfaces that appear near the moment of purchase, so retargeting and brand-heavy Performance Max look like heroes while demand generation looks like a cost. Reading the account blended, against your own numbers rather than the platform's self-reported conversions, is what stops you defunding the channels that create demand in order to keep feeding the ones that merely capture it. Get the measurement right first, then scale. Scaling on top of bad data just buys more of the wrong thing, faster.
Run everything to one margin model
The thread that ties Search, Shopping, Performance Max and YouTube together is not a channel strategy, it is an economic one. Before any of it, you model what a customer is genuinely worth and what you can afford to pay to win one. Every surface is then judged against that single number: contribution generated per pound of media invested. High-intent Search usually earns the first and largest claim on budget because it is the cheapest demand to convert. Shopping and Performance Max take what they can prove. Demand generation on YouTube earns its share when the blended numbers show it lifting the whole account, not before. That allocation is not a fixed pie chart set once a quarter and defended in a monthly deck. It moves as the numbers move. What pays gets more; what stops paying gets stopped, not explained away. This is also why paying an agency a percentage of your media spend is a conflict of interest dressed up as alignment: it rewards spending more, when the discipline that actually protects margin is spending less on the surfaces that are not working. Media should be paid directly to Google, and the fee for managing it should not rise simply because your budget did.
Where to start, and in what order
Sequence beats enthusiasm. Turn everything on at once and you cannot tell what is working, you cannot fix the feed and the tracking under load, and Performance Max will happily absorb the budget your other campaigns should be earning. Start where intent and margin are clearest, usually high-intent Search and a properly built Shopping feed, and prove the model there while the conversion data settles. Only once the numbers are trustworthy should you widen into Performance Max, with the exclusions and structure that stop it cannibalising the demand you already have. YouTube comes when you have a reason to create demand rather than only capture it: when you have exhausted the readily available search demand, or when the product needs explaining before people know to look for it. By then you have a working margin model to judge it against and honest measurement to read the lift, which is exactly what a demand channel needs and exactly what most accounts never build before switching it on. Beyond search is not a bigger version of Search. It is a set of different jobs, and the ones who do well out of it are the ones who run each to its own discipline and all of them to the same number.