Skip to main content
AI

The Future of Shopping: A Brand Plan for Agentic Commerce

By AJ San Pedro 5 min read
The Future of Shopping: A Brand Plan for Agentic Commerce

For as long as there has been e-commerce, the last step of the journey looked the same: a person comparing a few tabs, reading a few reviews, and clicking “buy”. That last step is starting to move. AI assistants now shortlist products, compare prices, summarise reviews, and in a growing number of cases, complete the purchase. Marketers call it agentic commerce. For brands, it raises an uncomfortable question: what happens when the first “customer” to read your pitch is a machine?

The shift is already measurable. McKinsey’s Agentic Commerce Consumer Survey, fielded in June 2026 with around 21,000 consumers across 11 markets, found that 53% had used AI for shopping in the previous three months. Among US AI users, the share who had completed a purchase with AI rose from 10% to 17% in a single year. Respondents were 3.2 times more likely to rank AI search as more influential than traditional search engines.

Retailers are bracing for it too. In BCG’s October 2026 survey of 150 leading retailers and brands, 68% of retailers expected agentic commerce to have a significant or transformational effect within two years. Referrals from external AI agents made up less than 1% of website visits in late 2025, but BCG estimates that share could reach 5% by the end of 2026, and the real influence on traffic can be up to four times higher than referrals alone show.

What actually changes for brands

Three things change at once. First, the shortlist is built before a person ever sees your website. Second, the agent reads structured facts, reviews, and third-party sources more carefully than your hero banner. Third, the human is still in charge of trust. BCG found that just 4% of consumers would let an agent complete a purchase entirely on its own. People will happily hand over the legwork. The final judgement they keep for themselves.

So agentic commerce belongs on the strategy agenda, next to pricing and positioning, rather than in an IT ticket. The brands that win will be easy for machines to understand and easy for people to believe.

Are most brands ready?

Not yet. BCG’s 2026 survey of US consumer packaged goods companies found that 59% reported no meaningful agentic AI capabilities. Among retailers, only 18% considered their marketplaces ready for agents, and only 18% were deploying agentic commerce at scale. The gap is an opportunity for brands that move first, because agents tend to reward whoever gives them the clearest, most complete answer.

A five-part plan for agentic commerce

When a client asks us where to begin, we sketch the same five building blocks on a whiteboard, roughly in this order.

  • Make your product and service data agent-ready. Agents compare specifications, prices, availability, delivery times, and return policies. If those facts are missing, inconsistent, or out of date, you drop off the shortlist. McKinsey cites Shopify data showing that AI searches using a structured product catalogue convert at roughly twice the rate of searches that rely on scraped or stale data.
  • Give agents reasons, not slogans. “Premium quality” means nothing to a model comparing ten options. Clear claims with proof do: what the product is for, who it suits, how it compares, what customers say, and which certifications back it up.
  • Earn the outside sources agents trust. Agents lean heavily on reviews, comparison sites, expert coverage, and editorial content. This is the same earned-media logic we described in our communications plan for AI search: AI trusts what others say about you.
  • Decide your data stance on purpose. Retailers are already choosing sides. In BCG’s survey, 42% planned a collaborative approach to agent data access, 32% a selective one, 6% planned to block agents, and 20% were undecided. Brands need the same clarity: which data you share, with which platforms, under which privacy and consent rules. Handled well, your data policy becomes part of the brand promise.
  • Protect the human relationship. As agents handle more of the comparison, the moments that remain human matter more: the unboxing, the service call, the community, the creator who explains the product honestly. Those are where loyalty is built, and where agents later find the reviews that recommend you.

What to track

Classic web analytics will undercount agentic influence, so we add a small set of measures:

  • Agent shortlist share: how often your products appear when assistants are asked your category’s key buying questions.
  • AI referral traffic and conversion: visits and orders from AI platforms, tracked separately from organic search.
  • Data completeness: the share of products with full, current attributes, pricing, and availability across every feed.
  • Description accuracy: whether AI describes your brand, products, and policies correctly, and where it gets facts wrong.
  • Incrementality: whether agent-driven sales are genuinely new or simply moved from other channels. BCG notes that brands increasingly want proof of incremental sales before investing heavily.

Where brands are already tripping up

  • Chasing a technical hack. Markup and feeds matter, but no single tag makes an agent recommend you. Clear facts and credible third-party proof do.
  • Letting stale data speak for you. An agent that finds an old price or a discontinued product will simply move on, and it will not tell you why.
  • Ignoring the margin maths. BCG points out that intermediaries have historically charged for placement, with online travel agencies taking roughly 15% to 25% of each booking. Plan for agent-routed demand to carry costs too.

A first step you can take this month

Pick your ten most important products or services. Ask the major AI assistants the questions your buyers ask before they choose. Record which brands appear, which sources are cited, and how your offer is described. Then fix the three biggest gaps in your product data and plan how to earn a place in the outside sources the agents keep citing. You will probably be surprised by at least one answer. Most teams we run this with are.

In a market where the first reader is a machine, the clearest brand tends to win the shortlist. Being trusted by the person who makes the final call still decides the sale.

If you want to know how AI shopping assistants see your brand today, and what it would take to make the shortlist, start a project with us.

Sources

Want these ideas working for your brand?

Start a conversation