LinkedIn recently announced the retirement of its lookalike audiences targeting feature. For companies and marketers relying on lookalike audiences, this change will significantly impact their campaigns, requiring strategic adjustments.
LinkedIn’s Details
LinkedIn specified that lookalike audiences will only be available until February 29, 2024. In their announcement, LinkedIn suggested two alternatives for marketers to replace lookalike audiences: predictive audiences and Audience Expansion. We’ll explore these two options and include a third option with one caveat, it’s outside the LinkedIn environment.
Predictive Audiences
In essence, predictive audiences are similar to lookalike audiences but with two key differences. First, it sources data from Lead Gen Forms, LinkedIn contact lists, and conversions data generated via the Insight Tag.
Second, LinkedIn combines “your data source and LinkedIn’s AI to automatically generate a new custom audience to use in your campaigns.”
Audience Expansion
Audience Expansion increases the reach of your ads by looking for user attributes that are similar to your original audience. Here is an example directly from LinkedIn:
“If your campaign targets members with the skill Online Advertising, Audience Expansion might expand your audience to members who list the skill Interactive Marketing on their profile. This means you can discover new quality prospects and automatically drive them into your marketing funnel.”
Beyond LinkedIn
One criteria sorely missing from the two options LinkedIn now offers is intent. Sure, you may reach users within LinkedIn with many similar characteristics to your existing customer base but what if they are not ready to buy at this time?
Enter CollegeApp, which connects educational organizations to their database of over 240 million adults who’ve expressed intent in registering in a continued education program. This is powerful if you’re a higher ed institution, because they focus on changing the way institutions advertise by tapping into audiences who’ve expressed intent in enrolling.
Pros & Cons
Cons
Having to Learn a New Process
The biggest drawback of this change has more to do with comfort than performance. If you’ve been using lookalike audiences in your campaigns and are happy with the result, it’s understandable that you would be concerned that it’s going away. As the old saying goes, “If it ain’t broke, don’t fix it.”
Bias in AI
Predictive audiences rely on LinkedIn’s internal systems and machine learning. A problem already seen with machine learning models is bias. Here is how the Harvard Business Review explains it:
“AI systems learn to make decisions based on training data, which can include biased human decisions or reflect historical or social inequities, even if sensitive variables such as gender, race, or sexual orientation are removed.”
When using lookalike audiences, you at least had control of creating the initial list to be used as the source material. You could potentially review and manually adjust that list to make it a fair representation of your target audience.
With predictive audiences, all sources are generated within the LinkedIn ecosystem and then its AI model “works its magic” to build the target audience. If bias is present in the source material or the AI itself, you can’t do much about it.
Loss of App Integration
With lookalike audiences going away, so is the LinkedIn Lookalike API. This means “customers leveraging a Marketing Partner, such as HubSpot, to create a lookalike audience will no longer be able to.”
Ads on social lead to landing pages that create email lists, which are then segmented and used to improve ads on social platforms. LinkedIn’s step is breaking this cycle and making it tougher for companies to create seamless experiences across a buyer’s journey.
Pros
Easier for New Organizations
If you’re a new company, you may not have enough data for lookalike or predictive audiences. It’s good to know that you can select attributes to match your target customer and LinkedIn’s Audience Expansion can give you more reach. If you have never used lookalike audiences, you won’t miss them and this would still be better going in than having no feedback at all.
Next Steps
As of this post, you have less than a month before LinkedIn discontinues lookalike audiences. Take full advantage of this time to test new campaigns using predictive audiences or Audience Expansion. Experimentation through A/B testing can help determine the most effective strategy for your needs.
However, due to these methods working only off internal LinkedIn activities, you may want to go outside of that environment. LinkedIn could potentially measure criteria you don’t find important. It’s their algorithm that determines what to include and its weight.
LinkedIn’s algorithm is designed to improve LinkedIn’s metrics, not necessarily your own. The predictive audience could yield solid impressions and even clicks but if the audience is not interested in buying, you won’t get the conversions. A custom, intent-based audience puts you back in control. You can make manual adjustments and the list itself is made up of people with purchase intent in the industry you work in.
For additional assistance with your ad campaigns, BranDimensional is here to help. We make it our priority to become an extension of your brand to represent competitive advantage in current and emerging markets by staying up on trends and technological developments. Contact us today to set up an initial free consultation with one of our experts.