Autotargeting helps advertisers quickly launch ads and save time selecting keywords. It also finds queries that the user missed, which might increase an ad's effectiveness.
We reviewed the numbers and can say that autotargeting is just as accurate as a human specialist. The percentage of impressions for irrelevant queries is the same for keywords selected by a human and machine responses (roughly 15%).
The new settings can lower this percentage and allow advertisers to choose which categories they need the algorithm's help with, preview and evaluate selected keywords, and get statistics on all queries.
Choosing query categories
In the ad campaign settings, you can now choose separate categories for autotargeting. By default, autotargeting is enabled for all categories. Categories for ads in search can be selected at the group level. When filtering for impressions, autotargeting and keywords have equal priority.
There are now five query categories for autotargeting to choose from:
- Targeted. The ad completely matches the user's queries. For example, the advertiser "Backpack world" with the ad [Backpacks for laptops] will have the targeted query [buy backpack for laptop].
- Alternative. The advertised product can replace what the user is looking for. The ad still matches the query. For example: [buy bag for laptop].
- Broad. The query is broader than the ad promotion. For example: [buy backpack].
- Related. Advertised goods and services can supplement the product in the query. For example: [buy laptop].
- Queries mentioning competitors. The advertised product is searched for. Example: [buy backpack for laptop "Backpack world"].
Query preview
We also added a query preview for advertised products in each category so that you can evaluate your keywords and remove any you don't need. These can be viewed on the ad level.
Statistics on all queries
You can now view statistics on all search queries in different categories. Analyze which query categories you pay for with parent keywords, and which with autotargeting. This saves time working with semantics and helps you evaluate which categories are the most effective.
Learn more about autotargeting in search in Help.