From brand discovery to brand agents:

How to prepare your brand for agentic customer experience

From brand discovery to brand agents
  • Publication
  • 10 minute read
  • June 03, 2026

As a customer, have you done this before? You're not sure you're getting the best deal on your home insurance. You ask ChatGPT: "Show me the best home insurance for a family in Melbourne." Three recommendations come back. None is your current insurer. You switch.

Now think about that from the brand side. What would it take to be one of the three recommended: accurately represented, competitively positioned, and ready for the customer to act?

A "Brand Agent Readiness" layer can help. It builds on capabilities most organisations already have, such as structured data, SEO, content management, and APIs, so the starting point is often more accessible than it appears. And because the landscape is evolving rapidly, an insights-first approach lets you build only what your data steers you towards.

AI agents are a new front door to your brand

Customers are shifting from searching to asking. Instead of browsing a list of search results, they ask an AI assistant (ChatGPT, Google AI, Copilot) and receive a recommendation directly.

Gartner predicted search volume would drop 25% by 2026 due to AI chatbots, and estimated that 60% of brands will use agentic AI to deliver streamlined 1:1 interactions by 2028. Based on what we're seeing, this shift is well underway.

The challenge: AI agents don't experience your brand the way humans do. They don't browse your homepage or feel your brand story. They ingest structured data, product attributes, pricing, and reviews – and use it to build a recommendation.

Search engines rank pages. Large language models rank knowledge. If your brand's knowledge isn't structured and machine–readable, you may not make the shortlist. 

Journey shift diagram

What agents look for

AI agents pull from your website and external sources, including comparison sites, review platforms, Wikipedia, and social media. They extract entities, attributes, relationships, and trust signals. From that, they determine a recommended action.

The key challenge: How can you influence what an agent recommends to your customer?

You can't control external agents directly. But you can shape what they find. Structured, specific, verifiable information performs better than vague marketing language. Data and clear evidence are easier for AI systems to retrieve, compare, and trust.

Not all agents are the same

We see six types, each with different levels of brand control:

Customer agents (ChatGPT, Claude, Google, Amazon) – capture intent, shortlist providers. No control. Focus on discoverability.

Emerging intermediaries (AI shopping assistants, fintech apps, vertical startups) – act on behalf of customers. Limited control. Monitor your representation.

Aggregator agents (comparison platforms, marketplaces) – compare price and features. Partial control.

Brand agents (your websites, apps, contact centres) – guide decisions, convert. Full control.

Platform agents (Adobe, Salesforce, Microsoft) – power data and personalisation. High control.

Core systems (policy, pricing, fulfilment platforms) – execute and fulfil. Full control. API–enable everything.

Build for human experience and agent comprehension

Your brand still needs to connect with people through design, storytelling, and emotional resonance. That doesn't change. But now it also needs to be comprehensible to machines.

Human experience: Visual, emotional, persuasive. Designed for attention, trust, and conversion.

Agent comprehension: Clear, structured, specific. Designed for retrieval, comparison, and reliable action.

This doesn't mean two websites. It means ensuring your existing content is structured for both human browsing and machine interpretation.

From brand discovery to brand agents: Three horizons

Agents find and interpret your brand today. Build discovery foundations: schema markup, structured product data, crawlable content. SEO still matters. If agents can't read you, they can't recommend you.

Some agents navigate journeys and transact on customers' behalf. Build actionability foundations: APIs, low–friction flows, machine–readable pricing. If agents can't act, the customer journey stalls at discovery.

Brands may deploy their own agents for proactive service, automated renewals, and personalised recommendations. Explore experimentally. If you don't own the agent relationship, someone else will.

Horizon Matrix.png

Brand agent readiness: a new capability

Delivering on these horizons requires building an emerging part of your technology stack. We call it the Brand Agent Readiness layer – capabilities that make your brand discoverable, understandable, and actionable to AI agents, while connecting to your existing enterprise technology.

What are the three dimensions?

Measurement and insights: always–on agent analytics. Build for low observability now; design for more as the market matures.

Inbound: be found, understood, and chosen by agents you don't control. Includes AI–optimised discovery, structured brand understanding, agent orchestration, APIs, and trust/identity/delegation.

Outbound: your agents serving your customers. Concierge services, contact centre consistency, proactive recommendations. Mostly experimental today.

There is no single product to buy. This is a conceptual architecture to build towards, implemented through API gateways, content management, structured data platforms, and emerging agent access tools.

Getting started with brand agents: what steps should you take?

Ask ChatGPT, Google AI, and Copilot about your brand. What do they recommend? What do they get wrong? This takes minutes and almost always surprises.

Are your product attributes explicit and comparable? Is your schema markup current?

Schema markup, structured product data, crawlable content. Experiment with emerging formats like llm.txt and content knowledge graph where relevant.

Where do CAPTCHAs, pop–ups, or authentication flows block agents? Where could APIs help?

Build for experimentation, not static reporting. This will look very different in 6–12 months.

 

Creating new business value: discoverability, trust, and retention

Brands that build structured, machine-readable foundations now will expand into a fast-growing discovery channel. They'll build trust with the agents that advise customers, not just the customers themselves.

The brands that wait may find agents have already formed opinions, based on whatever data was available.

Contact Us

Damien Brown

Partner, Customer Technology & Adobe Lead, PwC Australia

Contact form

Bill Bovopoulos

Managing Director, Customer Technology, PwC Australia

Contact form

Follow PwC Australia