By Jon Arthurs, Managing Director, Toluna
AI is transforming how consumers discover, compare, and evaluate brands. Almost half of consumers turn to AI during their buying journeys; the share of GenAI tools replacing search engines for product or service recommendations has shot up from 25% in 2023 to 58% in 2025. Consumers are using AI for just about everything: comparing options, summarizing reviews, and creating consideration sets.
For insights teams, the challenge is clear: understanding how AI systems are learning, sharing, and ultimately shaping the consumer journey. For marketers, this means understanding how to influence how your brand turns up on LLMs.
From Category Entry Points to Prompt Entry Points
Brand research has traditionally focused on salience, differentiation, emotional associations, and situations that trigger category consideration – what we call Category Entry Points (CEPs). But in an AI-mediated journey, these CEPs increasingly become prompts. Consumers may no longer think about simply buying detergent. They might ask AI: “What is the best detergent for tough stains?” or “Which laundry brand is best for sensitive skin?” The AI tool then translates the prompt into shortlist or recommendation.
The questions insights teams need to ask, then, are:
- Which prompts lead to the brand?
- Which needs/occasions is the brand associated with?
- Which competitors are (also) mentioned? How do competitors compare?
- What language does the AI use to describe the brand? Does this language support or weaken the brand’s intended positioning?
- Which sources shape the AI’s answer?
This doesn’t replace conventional brand tracking, but it adds another layer to it: the unit of analysis expands beyond CEPs to the wider prompt-response environment around the brand – what we call Prompt Entry Points (PEPs).
AI availability is bigger than SEO
AI systems draw on a much broader range of evidence than a brand’s own website: FAQs, product pages, expert reviews, comparison sites, marketplace listings, customer reviews, media coverage, forums, social content, structured data, and operational signals (e.g. availability, complaints and service experience).
In practice, AI-mediated brand perception is likely shaped by three forms of evidence:
1.Owned evidence: what the brand says about itself through its website(s), FAQs, campaigns, and help centers.
2.Earned evidence: what others say about the brand through reviews, awards, journalism, commentary, and forums.
3.Operational evidence: whether the customer experience supports the brand promise.
This means that a wider ecosystem of content, reputation, and real-world experience now shapes how brands are represented and perceived.
Emotion as evidence
Although AI systems don’t experience emotions themselves, they can infer it from reviews, descriptions, loyalty signals, and social proof. Emotion therefore still matters, because at the heart of AI-mediated journeys are consumers who still feel, remember, and choose.
A brand may want to be seen as premium or innovative and often builds these associations through creative assets like visuals, tone, campaigns, etc. While these assets can be highly effective at shaping human perception, their intended meaning may be less visible to AI (unless supported by machine-readable signals like transcripts or metadata).
What this means for marketers: continue building emotion for people, but also ensure that emotional meaning is understandable to AI tools.
Measuring AI availability
This requires an evolution of brand measurement, starting with identifying if and how consumers are interacting with AI (beyond general AI assistants) during the category journey. Some examples may be AI-generated summaries, market-specific assistants, web chatbots, and recommendation tools.
Brands can then determine whether these tools increased, decreased, or made no impact on consumer consideration. They can also examine the role AI played: whether it introduced new brands, offered alternatives, summarized reviews, or steered consumers towards competitors. This should be paired with structured AI audits using PEPs rooted in genuine consumer needs, rather than generic prompts.
Over time, this leads to a broader AI availability metric: a way of understanding how accessible, credible, and recommendable a brand is within AI-mediated decision environments.
Moving forward
As AI systems influence discovery and consideration, then AI availability increasingly becomes part of brand equity.
This has several implications:
- Brand equity becomes a discovery asset: Alongside helping brands support human memory and preference, brand equity may now also affect AI-generated citations and recommendations.
- AI availability becomes a brand management discipline: Rather than replacing existing brand KPIs, AI availability adds a new layer to measure if the brand is visible, trusted, and accurately represented by AI systems.
- Emotional resonance should be machine-readable: Emotion will continue playing a role, only now it must be accessible to AI systems.
- Reputation and earned media are ever-important: They shape how AI understands and represents the brand.
- Organic recommendation, citation, and paid visibility are distinct from one another: Being recommended by AI is not the same as being cited as a source, and neither is synonymous to AI advertising. Knowing the difference and treating each accordingly is key.
The evolution of brand intelligence
As AI increasingly influences consumer choice, brands need to go beyond understanding what consumers think, remember, and associate (the brand) with, but also how they are represented within AI environments. This doesn’t mean that brand tracking is being replaced. It is, however, being expanded to capture the growing influence of AI on consumer decision making.
