LinkedIn’s 2019 Ad Upgrade Continues Apace With Lookalike Launch
LinkedIn’s strong start to 2019 continues with the launch of three new targeting features, announced on Wednesday, with the most notable being Lookalike Audiences.
For those unfamiliar with the idea of lookalikes, the feature was first popularised on Facebook in 2013. If you don’t feel like counting, that was six years ago.
Lookalikes are a staple now for all major players, and even most of the newbs. As we have touched on previously, LinkedIn doesn’t like to be rushed with its marketing product development.
Watching LinkedIn roll out new marketing solutions is very similar to watching my four-year-old son getting ready in the morning for Kinder.
Painfully drawn out, but worth it in the end.
And with everyone crying.
Expanded Interest Targeting & New Audience Templates
Alongside Lookalikes, LinkedIn has also expanded its Interest targeting, which launched in January, and introduced new machine learning powered Audience Templates.
Audience Templates provide marketers with a selection of 20+ predefined B2B audiences. Including characteristics such as member skills, job titles, groups, and so on.
According to LinkedIn Marketing Solutions senior product manager Jae Oh:
"These templates include audience characteristics that you can activate with a single click. This helps you effectively reach the professional audiences that matter most to your business, without spending hours in setup."
Lookalike Audiences Pros and Cons
Creating lookalikes is normally pretty easy in any channel, although occasionally there are obvious nuances. For example, Google’s equivalent is called ‘Similar Audiences’, and is more likely to draw on search data, which in *theory* should make them more accurate.
In most cases, a marketer defines an initial ‘seed’ audience, perhaps people who have visited your website recently, and then you create a new lookalike audience of these.
To do this, a platform assesses the seed audience and then combs their records for other users with similar traits, characteristics or behaviours. And voila, you have your lookalikes.
While this a positive development for marketers ramping up LinkedIn in 2019, the performance of lookalikes can vary considerably.
Firstly, not all platforms have comparative machine learning, or user data, which essentially powers this kind of audience profiling. Secondly, if the initial seed audience is not well qualified, then the lookalikes will be weak also.
It’s a classic input/output dilemma, i.e. rubbish in, rubbish out.
Thirdly, just because Audience B is a lookalike of Audience A does not mean they will respond to a message, creative or copy in the same way. Don’t forget, the seed audience most likely knows your organisation, whereas the lookalike probably doesn’t.
On the flip-side, this can be a positive sometimes, since lookalikes are often good for prospecting. Opening an organisation up to audiences it has not reached, but who should hopefully be worth reaching nonetheless.
LinkedIn's new and expanded targeting is rolling out globally now for all users.