Google Ads
The Google Ads Campaign Types, and What Each One Is For
Google Ads presents itself as one product. It is really a set of quite different advertising businesses sharing a login, and the most expensive mistakes in paid media come from using one of them to…
In short
Every campaign type sits on one side of a line: capturing demand that already exists, or creating demand that does not. Capture is cheaper per customer and limited by search volume; creation is more expensive and effectively unlimited. Most businesses should exhaust Search before spending seriously on anything else — and the common failure is reaching for a newer campaign type when the actual problem was tracking.
Google Ads presents itself as one product. It is really a set of quite different advertising businesses sharing a login, and the most expensive mistakes in paid media come from using one of them to do another one's job.
Here is what each campaign type is actually for, and when it is the wrong choice.
The distinction that governs everything: capture or create
Every advertising format sits on one side of a line.
Demand capture reaches people already looking for what you sell. They have a problem, they have named it, they are searching. Your job is to be the answer they find.
Demand creation reaches people who are not looking. They may fit the profile perfectly, but nothing has prompted them today. Your job is to interrupt usefully.
Capture is cheaper per customer and strictly limited by how many people are searching. Creation is more expensive per customer and effectively unlimited. Almost every business should exhaust the first before spending seriously on the second — and a great many buy the second while the first is still half empty.
Search campaigns
What it is: text ads on Google results pages, triggered by keywords.
What it is for: capturing existing demand. This is the highest-intent inventory Google sells, and for most businesses with a defined product or service it is where paid media should start and where most of the budget should stay.
When it is wrong: when nobody is searching. If you have invented a category, or you sell something people do not know exists, there is no search volume to capture and Search will simply spend very little while you wonder what is broken.
The thing people get wrong: treating keywords as the whole of targeting and ignoring the search terms report, which is where you find out what you actually bought.
Performance Max
What it is: one campaign running across Search, Display, YouTube, Gmail, Discover and Maps, with Google allocating budget between them.
What it is for: scaling an account that already has reliable conversion data, a strong product feed, and enough volume for automation to learn from.
When it is wrong: as a first campaign, or as a fix for an account that is not converting. It gives you less visibility and less control than Search, and it will absorb whatever budget you give it. If tracking is imperfect, Performance Max will optimise confidently towards the wrong thing and the reporting will make it hard to see.
The thing people get wrong: running it alongside Search without watching what happens to the Search campaign's volume. The two overlap, and you can end up paying Performance Max prices for traffic Search was already capturing more cheaply.
Display campaigns
What it is: image and text ads across roughly two million sites and apps in Google's Display Network.
What it is for: remarketing — putting your brand back in front of people who already visited. Cheap reach for an audience that already knows who you are.
When it is wrong: as a source of new customers for most small and mid-sized businesses. Clicks are cheap because intent is near zero, and cheap clicks make the reporting look busy while the sales pipeline stays empty.
The thing people get wrong: judging it by clicks. A large proportion of Display clicks on mobile are accidental. Judge it by conversions or not at all.
Video campaigns
What it is: ads on YouTube, in various skippable and non-skippable formats.
What it is for: demand creation, and the one format that can genuinely explain something complicated. If your product needs a demonstration before anyone understands why they want it, video is the only place in the Google stack that can do that at scale.
When it is wrong: as a direct response channel for a business with an unexhausted Search opportunity. Video sells things people were not looking for, at a cost that reflects that.
The thing people get wrong: measuring it on last-click conversions, which will always make it look worse than it is, and then either killing it prematurely or defending it with metrics nobody can tie to revenue. Decide before launch which it is: brand investment measured on reach and recall, or performance spend measured on conversions.
Shopping campaigns
What it is: product listings with image, price and merchant name, drawn from a feed you supply through Merchant Center.
What it is for: retail. If you sell physical products online, this is usually the highest-return format available.
When it is wrong: for services, or for anything without a clean product catalogue.
The thing people get wrong: treating it as a bidding exercise. In Shopping, the feed is the campaign — titles, product types, attributes and image quality determine which searches you appear for, far more than bids do. A day spent fixing product titles typically beats a month of bid adjustments. The Shopping and video guide covers this properly.
Demand Gen campaigns
What it is: visually-led ads across YouTube, Discover and Gmail, aimed at audiences rather than keywords.
What it is for: demand creation for consumer products with strong visuals, particularly where you have good first-party audience data to build from.
When it is wrong: business-to-business with long sales cycles and small addressable audiences, where the targeting cannot get precise enough to be efficient.
App campaigns
What it is: automated promotion of a mobile app across Google's properties.
What it is for: app installs and in-app actions. Narrow and self-explanatory.
When it is wrong: any time you do not have an app.
How to sequence them
The order that works
Which campaign type, and when
Capture demand that already exists before paying to create demand that does not. The common failure is jumping to step four because growth stalled, when the problem was at step one.
Search, on tight keywords
The highest-intent inventory Google sells, and the easiest to diagnose when something breaks.
Needs: conversion tracking you have tested yourself
Shopping — immediately, if you are a retailer
The feed is the campaign. Fixing product titles beats a month of bid adjustments.
Needs: a clean product catalogue in Merchant Center
Remarketing on Display
Cheap reach to people who already know who you are. Poor as a source of new demand.
Needs: enough traffic to build meaningful audiences
Performance Max
Only once Search is genuinely capacity-limited — winning the impressions worth winning and unable to spend more there profitably.
Needs: real conversion volume and trustworthy tracking
Video or Demand Gen
When capture is exhausted and growth has to come from creating demand.
Needs: budget that can absorb a longer payback
Each step assumes the one before it is working. Jumping ahead because growth stalled is how accounts end up paying Performance Max prices for traffic Search was capturing more cheaply.
Choose by the question you are answering
A quick way to decide:
- Are people searching for this? If yes, Search first.
- Do I sell physical products? Shopping, and fix the feed before anything else.
- Have people visited and not converted? Remarketing.
- Do I need to explain something before anyone wants it? Video.
- Is Search maxed out and still profitable? Then, and only then, expand into automated multi-channel formats.
The formats are not ranked by sophistication, and the newest one is not the best one. They answer different questions, and the expensive mistake is answering a question you do not have.
If you want a structured read on which parts of an existing account are weakest before adding another campaign type, the Google Ads scorecard covers it in twenty questions.
Frequently asked
Which campaign type should a new advertiser start with?
Search, in almost every case where people are actively looking for what you sell. It is the highest-intent inventory Google has, it is the easiest to diagnose when something goes wrong, and it gives you the conversion history that every automated campaign type later depends on.
Should I run Performance Max and Search at the same time?
You can, but watch what happens to the Search campaign's volume. The two overlap, and Performance Max can absorb traffic that Search was already capturing more cheaply — which shows up as flat overall performance at higher cost. Segment your reporting and judge them separately rather than looking only at the blended figure.
Is Display advertising worth it for finding new customers?
Rarely, for most small and mid-sized businesses. Display clicks are cheap because intent is close to zero, and a meaningful share of mobile Display clicks are accidental. Its genuine strength is remarketing — reaching people who have already visited your site and know who you are.
What matters most in a Shopping campaign?
The product feed. Titles, product types, attributes and image quality determine which searches you appear for, far more than bidding does. A day spent fixing product titles typically returns more than a month of bid adjustments.
- Campaign Types
- Google Ads
- Performance Max
- Shopping
Next step
Score your account
Twenty questions across tracking, structure, landing pages and bidding. You get a score and a prioritised list of what to fix first.
