An affinity audience, now called an affinity segment in Google Ads, is a group of people Google believes have a strong, long-running interest in a topic, such as home cooking, running, travel or technology. It describes who someone is over months and years, not what they are about to buy this week.
How affinity audiences work
Google builds these segments from patterns in how signed-in users behave across its services and partner sites: the videos they watch, the sites they visit and the apps they use over time. It then offers them as ready-made audience segments with names like “Food and Dining: Cooking Enthusiasts” or “Sports and Fitness: Golf Enthusiasts”. You cannot see who is in a segment, only its name and a rough size estimate.
Affinity segments are mainly used on YouTube, display, Demand Gen and similar campaigns where the aim is to reach new people before they start searching. In Performance Max they can be added as an audience signal, a hint rather than a hard limit. In search campaigns they can be added in observation mode, so you see how people in the segment perform without restricting who sees the ads, and then adjust bids or messaging.
The difference from an in-market audience is timing. In-market segments are built from recent behaviour that suggests someone is actively researching a purchase, such as comparing car insurance quotes. Affinity segments are broader and slower moving. Someone in the golf affinity segment may never buy a new set of clubs this year; someone in-market for golf equipment probably will soon.
Why it matters
For a UK business trying to build awareness, affinity segments are an affordable way to reach people whose interests match the product, before competitors reach them through search. A cookware brand, a London cookery school or a specialist food delivery company can put a video in front of people who already watch cooking content, at a lower cost per impression than most search clicks.
The catch is that broad interest does not mean intent. Affinity campaigns rarely produce cheap leads straight away, and judging them on cost per conversion in the first month usually leads to them being switched off before they have done their job. They work best as the top layer of a plan that also includes search and remarketing.
Common mistakes
- Choosing an affinity segment for a search or lead campaign and then wondering why the leads are poor quality.
- Picking the segment whose name sounds closest without checking how large and how broad it is. “Shoppers” covers almost everyone.
- Using targeting mode in search campaigns when observation was intended. The targeting setting decides whether a segment restricts reach or only reports on it.
- Measuring awareness campaigns only on last-click conversions, which credit the final search rather than the video that started the journey.
How to act on it
Decide first what the campaign is for. If the goal is reaching new people who fit your customer profile, test two or three affinity segments that genuinely match, alongside a custom segment built from the search terms and websites your customers use, which is often sharper. Run them on video or Demand Gen with creative that introduces the business rather than asking for a sale.
Judge results on reach, view rate, cost per view and any later lift in branded searches, as well as conversions over a longer window. In search campaigns, add relevant affinity segments in observation mode and look at their conversion rates after a month. Planning audience layers like this is part of how I run YouTube advertising campaigns.