In recent weeks, we’ve been following with interest the increasing noise around the idea of GenAi Engine Optimisation (the idea that we optimise our assets for GenAI tools like ChatGPT, Claude, Grok, etc.), and reflecting on how these topics can often become the latest emperor’s new clothes. Focus is drawn to the shiny new thing, when maybe it would be more effective to focus on the things we have always known to be best practice.
These waves come and go, and decision makers in marketing, ecommerce and customer development need to keep an outward perspective, anticipating the next wave, and assessing whether it is going to be consequential, transformative, etc. Will we have to completely change our operating approach, or will it just be another area where simply good execution is required?Another of these potential waves we should be keeping an eye on is the idea of Agentic AI. Potentially, it’s not just a tweak to e-commerce, but a full-on redefinition of how brands and consumers connect (I say this with the normal healthy cynicism required). I’m talking about agentic commerce – shopping that is powered not just by humans clicking buttons, but by AI agents acting on behalf of humans.
What is Agentic Commerce?
Agentic commerce is the idea that we move beyond search → browse → buy. Instead, intelligent autonomous agents anticipate needs, navigate options & transact, all aligned to human intent, but acting with independence. According to McKinsey & Company, it “represents a seismic shift in the marketplace … AI agents now anticipate, personalise and automate every step of the process.”
In practice this might look like: a shopper’s AI agent pops up, knows their preferences, picks the product, compares across merchants, adds to cart, completes checkout while the human is busy doing something else less boring instead. Or perhaps the agent negotiates offers, handles subscriptions, optimises for loyalty or value. It’s commerce, but with a different lead actor.
Why this matters for brands?
If you’ve been in marketing, e-commerce or CRM long enough, you recognise the guardrails: funnel models, customer journeys, click-throughs, cart abandonment. We probably need to consider if/how Agentic commerce will disrupt these. Here’s what could change:
- The decision-maker shifts. It’s no longer solely the human at the keyboard. It’s their agent. That means brands must ask: how do you win with the agent, not just the human? McKinsey suggests brands will need to rethink “the full stack of engagement; not for the people you’ve worked so hard to understand, but for the agents now acting on their behalf.”
- The experience flips. Instead of brands guiding the shopper through their site or app, the agent navigates across brands and channels often without us being aware. For example, one report described the move as: “agentic commerce flips the customer journey on its head: Instead of users visiting a site or app, autonomous agents do the leg-work.”
- Data and control get contested. If agents transact on behalf of users across multiple retailers, where does your insight come from? Who controls the connection? For large platforms, this is already contentious. Week by week we’re seeing skirmishes: e.g., Amazon has pushed back against an AI browser agent that automates purchases on its platform, citing risk to its customer experience and data model.
- New monetisation, new threat. McKinsey projects that in the US B2C retail space alone, agentic commerce could deliver up to $1 trillion by 2030, with global potential possibly $3-5 trillion. That’s a massive opportunity — but also a threat if you’re not ready.
Brand relevance gets recalibrated. When agents act, brands become part of an ecosystem of choices, protocols, integrations and permissions. If you’re not the brand the agent favours, you might get skipped entirely. This is very similar to the argument about GEO.
Key Trends and Movements
Several signals suggest this is becoming more than hype.
- Partnerships are forming between retail and AI-platforms. For example, Walmart Inc.’s tie-up with OpenAI to create “AI-first shopping experiences” underscores how some the retail giants see this as strategic.
- Platform conflicts are emerging. The Amazon vs Perplexity AI case, where Amazon is said to have demanded that Perplexity stop enabling its browser agent to make purchases — highlights how disruptive agentic commerce can be to existing business models.
- Infrastructure innovation is underway. New protocols: Agent-to-Agent (A2A), Agent Payments (AP2), Model Context Protocol (MCP). These enable agents to act autonomously, transact, reason and interact with each other and with systems.
The technology readiness is ramping. With generative AI and agentic AI moving past experimentation into structured deployment, organisations are under pressure to move from ‘could do’ to ‘must do’.
What should you consider now?
Here are three actions to get ahead.
- Map your agent-interface risk and opportunity. Ask: in a world where an agent can buy without a human browsing our site, what does our brand look like? Who is our ‘agent customer’? Which of our propositions would still win when the shopper is an AI?
- Design for agent-trust & delegated authority. If a user hands over their agent to transact, trust matters. How will you ensure it chooses you, values you, and remains aligned with your brand? Think about consent, agent-permissions, transparency and experience optimisation for non-human users.
- Experiment in live settings. It’s not enough to theorise. Build a small pilot: allow an assistant or agent prototype to transact a simple purchase journey, negotiate, optimise or repurchase. Measure what parts of your funnel work when the human is barely present. Document what changes.
Agentic commerce could develop into something meaningful that might impact, and if it does it could change how value is created, how brands compete, how platforms dominate, and how customers buy.
Keep a watchful eye on agentic commerce. Because when your customer is no longer clicking your website, but instead delegating to an AI agent, you’ll want to ensure your brand still gets chosen.Now, they need integration.
Cognitive Union is a progressive, boutique learning and performance consultancy. We work with forward-thinking businesses. Transforming their people. Shaping their culture. Helping them embrace change and take on the world. Find this blog useful? Sign up to our email newsletter (bottom of this page) where you can receive articles like this and other insights (not publically published), and you can also follow us on LinkedIn.