Talking about artificial intelligence (AI) is not talking about the future, it is talking about the present. Its presence in daily activities in different economic sectors is already a reality, and its implementation is transforming our daily lives. It is not here to replace us, but to improve our processes and help us be more efficient.
If we focus on the field of online advertising, we can see how AI is playing a fundamental role in the execution of advertising campaigns. Large advertising platforms such as Google, Meta and various programmatic DSPs have been implementing the use of AI for some time now to facilitate the initial configuration process and help optimise advertising campaigns to achieve set objectives (KPIs).
In this way, the role of the advertiser or agency is shifting from manual management to strategic supervision of algorithms. This is changing how we have historically executed online actions in the media.
The decline of manual segmentation
Since the advent of online advertising, those of us who execute campaigns have spent hours manually defining the ideal target audience we wanted to reach on the platforms.
We tended to use all the segmentation options available to us to reach our target audience as accurately as possible: age ranges, gender, specific interests, websites, detailed geographical locations beyond the country, optimal publication times, etc.
However, the use of AI in segmentation has proven to be superior for two reasons:
- Speed and volume of data: the algorithms of online media buying platforms process billions of interactions per minute (clicks, impressions, views, purchases, abandonments), something impossible for a human being. Using all this information, AI can identify patterns and micro-segments that we would never have been able to detect.
- Goodbye to the myth of the ideal audience: AI can understand that our ‘ideal target’ (female, 25 years old, fan of X) may not be the most cost-effective for our campaign. The platforms’ automated tools look for the user most likely to complete our objective (purchase, subscribe, download our app), regardless of the segmentation we have defined.
Consequently, AI improves how we find high-value users as it has more information and capacity than we can have when manually segmenting a campaign.
The key is to provide the platform with broad audience signals or first-party data, which AI can then complement and amplify. Changing the execution model: instead of restricting the audience, we give the campaign the best variety of creatives and texts, letting AI find our ideal customer in a larger universe.
Bid automation: from tactics to strategy
Bid automation becomes a key point in the implementation of AI in campaigns. We move from a model where the ideal was to manually set a maximum bid per click (CPC) or per thousand impressions (CPM) to giving the AI full control of the bid, but setting a very clear objective: a final ROAS or CPA that we will be responsible for controlling and optimising.
If we look at two key advertising tools, Meta and Google, this translates into:
- Maximum Performance Campaigns (Google Ads): in this type of campaign, we simply define a specific ROAS or CPA target, and the AI decides where (YouTube, display, search), when and how much to bid to achieve that target in the most efficient way.
- Advantage+ Shopping Campaigns (Meta Ads): in these, the system uses AI to automatically test thousands of combinations of ads, creatives and audiences, optimising our budget towards those that work best. The focus shifts from configuration to the quality of the assets (images, videos and text) that you deliver to the AI.
The future of online advertising does not lie in
