Dentsu on How AI Is Reshaping Search

AI Max & AI in Search

How are you advising clients on when to use AI Max in their search campaigns?

We are encouraging clients to test AI Max as soon as they can, wherever they can! With AI Mode now in play in the UK, and Google’s intention to bring ad inventory to AI Mode, advertisers should be doing all they can to learn about the latest format to put them in a stronger, more competitive position for the inevitable rollout and scaling of the AI Max/AI Mode features.

As with any new campaign type, especially those powered by AI technology, there can be a natural hesitancy to test, with concerns around performance and brand safety. A robust list of negatives is a must, and for advertisers that would prefer to just dip their toe in, there is the option to run “part component testing”, which means excluding the asset and URL opt-ins. However, this approach is thought to reduce the likelihood of showing for AI Mode queries due to ad rank being poorer.

Further tips for testing closely align with Google’s best practice, following A/B, pre-post and incrementality testing methodologies and ensuring that campaigns have sufficient budget to run effectively (>$50 per day). Auditing and enhancing the content across a brand’s digital estate is also key for maximising the full potential of AI Max.

What role will first-party data play in strengthening advertiser performance in an AI-driven search environment? 

First party data, whether collected directly or via loyalty programmes, website visits, app usage or other channels, are all forms of consumer signal. Signals are everything because they reflect real user behaviour and intent.

Therefore, AI driven search advertising should be built on a strong foundation of first-party data because it can deliver more relevant, more personalised and more measurable campaigns, while also staying compliant and future-ready.

With that in mind, having “data strength” is more important than ever to enable AI to recognise patterns, predict outcomes and optimise in real time. Advertisers need to break down data silos to leverage the benefits of their proprietary data to deliver outcomes for their business.

Those that invest in connecting, enriching, and activating their first-party data will see the greatest performance gains as AI becomes even more central to search and digital media.

What would be the one thing you would suggest advertisers, agencies and clients could do immediately to prepare for the future of Search advertising?

Preparing for the move away from keyword driven search to context driven. We’re seeing this with the latest search product releases and there are no signs of this slowing down. Search won’t just be a case of “what people search”, but more a case of “what people want, need and mean”, even before they ask it. This is another reason leaning into personalisation and first-party data is crucial to search delivery in the future. The first step to this journey is to start testing these formats so you have a data driven baseline in which to understand their relative effectiveness for your business.

What skills and capabilities do you think your teams will need to future-proof your search offering as AI evolves?

• AI literacy and proficiency: search marketers will need to be able to use AI tools for end-to-end campaign management, data analysis, experimentation and governance. Crafting AI prompts will become a core requirement and developing skills to interact with AI agents will be a big part of the briefing and activation process
• Strategic prowess: skills will shift from activation to become more routed in strategy. The focus will be on innovation and bespoke solutions, continuing to harness the element of human connection to deliver on business outcomes.
• Facilitating increased cross functional workflows (data, creative, tech, media etc): deliver integrated AI-powered campaigns. In particular, Paid Search and SEO will become even more intrinsically linked as content on a website will more directly impact how and where an ad serves.

Ultimately, those that can understand the whole search landscape, interactions within and around it and are able to deploy AI effectively will lead the way.