AI isn’t killing search — it’s evolving it
Opinion
As search evolves rapidly in the age of LLMs, here are five steps for marketers to future-proof their discovery strategy.
The search landscape is evolving faster than ever. The expedited launch of AI Mode by Google in the UK at the end of July, as well as its recent move to index Instagram posts, are just two examples of how the platforms are responding to a move away from traditional search methods.
According to Salesforce, 76% of Gen Z have discovered products on social platforms and 39% have made purchases. This demographic has preferred social search over traditional search since 2017.
Gen Z might be the earliest adopters, but they are by no means alone in their usage. For example, the majority of ChatGPT’s user base is over 25.
The rise of large language models (LLMs) further sharpens this evolution, introducing a technology that is not only accessible and intuitive but also feels at once very natural while entirely futuristic.
In this new world, here are five practical steps marketers can future-proof their brand discovery strategy.
Havas Media Network launches ‘generative engine optimisation’ tool
Assemble a discovery team
LLMs and user behaviour are changing rapidly.
The most impactful action brands can take right now is to gather a diverse, cross-functional group of passionate people, ideally spanning SEO, PR, social and analytics across a mix of seniorities, to help navigate this.
A clear set of objectives rooted in progress (not results) is required to guarantee momentum during this period of analysis and discovery; rigid success metrics risk stifling experimentation.
Measure what you can, model what you can’t
One of the most unnerving parts of LLM and social search-influenced journeys is how many touchpoints now occur invisibly.
Urchin tracking module parameters should be used to track LLM website referrals. This data can then be analysed to identify which LLMs are driving visits, where they land and what they do next. Trending behaviour can help marketers understand what visitors are trying to achieve and help hypothesise their previous journeys within LLMs.
If presence tracking tools aren’t available, these journeys can be mimicked through conversational search. It won’t be exact, but it will help identify recurring patterns of competitors, partners and publishers that are showing up in the journey (personalisation will need to be disabled here).
The future of branding: Influence LLMs to protect your share of market
Invest in multi-modal content
Multi-modal content (content that is more than just text on a website) is key to growing a brand’s presence in LLMs and social search.
Through a brand’s existing channels, they will have an idea of where their audience is and where some content gaps might be; this is their jumping off point for creating multi-modal content. Marketers will need to reset expectations for this content; it’s unlikely to drive last-click conversions or even referrals.
Content for social search should be evergreen and discoverable. This will make it far less spiky in performance than trending, topical, viral content. Content for LLMs, on the other hand, is about visibility across largely untraceable journeys.
Stay adaptable
To succeed in today’s search landscape, brands and their organisations need to remain adaptable as the technology is changing by the day. Measurement capabilities are evolving and being debunked, tactics are successful one week and defunct the next.
Adopting a growth mindset and encouraging it throughout the team will enable marketers to change direction when necessary and will set up frameworks for evolving workflows instead of fixed ones.
There won’t just be one first-mover advantage — there will be many and it’s key to maximise your chances.
‘End of an era for search as we know it’? Publishers grapple with gen-AI search
Show up in conversations, not just clicks
AI is bringing about a rapid shift in how people discover, evaluate and choose brands. But this isn’t a cause for panic; it’s an opportunity.
As we’ve seen before, marketers have always adapted to changes in data, technology and how consumers seek information. The evolution we’re witnessing now, from search dominance to fragmented discovery across platforms and conversations with LLMs, may feel unfamiliar, but SEO’s core principles still apply: adapt, adapt, adapt.
AI isn’t killing search. AI is evolving it.
The brands that thrive won’t be those clinging to search volume reports or waiting for stable tracking solutions. They’ll be the ones building curious teams, encouraging experimentation and developing multi-modal content that meets users where they are — whether that’s a TikTok feed, a Reddit thread or a ChatGPT prompt.
Matt Allfrey is head of SEO, EMEA, at PMG