Search term targeting refers to showing ads based on real-time search queries. This is particularly relevant for marketplaces, eCommerce sites, and any company with Sponsored Listings.
In the below example, a sponsored brownies recipe from Breyers appears when someone searches for 'brownies'.
Search-term targeting is a powerful targeting type that improves relevance for the user and leads to higher ad engagement rates.
Use Keyword Targeting to implement search term targeting. To do this, you'll want to:
- Create flights/ads targeting the search-terms that will trigger the specific ad
- Send search terms in the
keywordsfield of the Decision API request
- If there's a match, then that ad is eligible to show (please note: it may not necessarily win)
Keyword targeting is an exact match. If you target "brownie", the keyword of "chocolate brownie" would not trigger the ad
With Reserved Keys, you do NOT need to add keyword targeting at the flight/ad level. You DO have to add a Zerkel Query to the Custom Targeting section of UI or to the
CustomTargeting parameter of the Create Flights/Ad/Channel Request. You also have to pass
keywords in the Decision API request.
With Reserved Key keyword targeting, you get more operators to use, such as LIKE and CONTAINS.
For instance, if you want to target search terms containing 'brownie', you could use this Zerkel Query:
$keywords like "*brownie*"
This would capture permutations like "chocolate brownie" or "brownies for a party".
Adzerk Reporting does not break down Custom Reporting queries
If you have an advertiser who wants to always win the ad placement for a given search term, you can use priority rules to ensure this. Here are instructions for creating Exclusive Keyword Targeting Campaigns.
For instance, the recipe site could ensure that Breyer's recipe will always appear for the search of 'brownie', even if they are not the highest bidder.
It'll be difficult to create campaigns targeting every single relevant search term someone may search for. Below are some ideas for addressing this.
- Use the LIKE and CONTAIN operators via Custom Reporting to capture variations of a specific word
- Manually create a robust list composed of many different permutations (using either Keyword Targeting or Reserved Keys)
- Parse out the individual words in the search phrase and send them over as individual
keywordsin the Decision API request.
For example, a search term of
chocolate brownies could be parsed as
"keywords": ["chocolate", "brownies"] in the request.