Every ad platform lets you say who you want to show your ad to. The simplest targeting is demographic, by age and location, though you can also target by interests or occupation.
The platforms differ in this. LinkedIn, for example, lets you target by job title and industry, which is useful for professional services, while Meta works more with interests and behaviour.
In Google search ads targeting works differently, as the search terms decide who sees the ad. Here the choice of keywords and the list of excluded terms play the role of targeting.
A lookalike audience starts from an existing audience, for example the people who, according to the pixel data, bought or requested a quote on your website. The platform looks for what these people have in common, and shows the ad to others with similar characteristics who have not yet come across your brand.
Narrow or broad targeting
Narrow targeting is tempting, because it seems every part of the budget reaches the right person. Today's ad algorithms, however, often work better with a broader audience, because they learn from the conversion data who responds to the ad. In that case the creative and the message do the filtering.
- How much traffic you have, and whether there is already enough conversion data for the platform to learn from
- How narrow your target group is, for example for a professional service
- How large the budget is, because with a small budget too broad an audience gets spread thin
- Which platform you advertise on, because LinkedIn and Meta work differently
The quality of a lookalike audience depends on the source audience. If the source contains many people who only clicked, the platform looks for people like them, so it is worth starting from buyers or quote requesters.
That is why we test targeting at the start of a campaign, with several audiences in parallel, and move the budget based on the results. Broader targeting reaches new people, and remarketing brings back those who have already visited you. Both rest on the same measured events.
