The new rules of discovery: is your brand ready for the agentic commerce era?

Build commerce for people...and machines 

Tomorrow’s eCommerce experience needs to work for two audiences: human customers, and the machine layer acting on their behalf. AI tools can only recommend what they can see, understand, and ultimately, trust. If product information, availability, pricing, or brand context is inconsistent or invisible to crawlers, a business may be absent from an AI-generated answer, even when it meets the search intent of the user on the other end. With Gartner forecasting $15 trillion in AI-intermediated purchases by 2028, the cost of failing to meet these discoverability conditions is clear. 

These imperatives place renewed importance on structured, governed product catalogues. For non-branded searches in particular, product context must closely match customer intent if a brand is to earn visibility and consideration. With around 95% of current recorded search terms registering as brand-agnostic, humility and a willingness to embrace the current promiscuity of eCommerce drivers will breed success and longevity.  

In addition, semantic and agentic search can also shorten the path from inspiration to purchase. Combined with composable commerce, brands can engage audiences effectively with tightly defined experiences, such as a limited-edition colourway available for a single week only, for example, without rebuilding the entire storefront. Discovery becomes more urgent, more contextual, and more commercially focused, and from a marketing execution point of view, both efficient and lucrative.  

Technical performance is now a marketing metric 

Optimising for AI discovery should not be considered simply a content, or ‘traditional’ marketing challenge. It also exposes the underpinning fundamentals of the digital experience: load speed, accessibility, crawlability, dwell time, and bounce rate. These are sometimes framed as technical concerns, but their impact is highly commercial. Page speed, for example, directly affects the efficiency of every single pound spent on traffic. Google’s mobile research tells us that bounce probability jumps around 32% when load time goes from just one to three seconds, increasing to an eye-watering 90% jump at five seconds. Just 0.1 seconds of improvement was associated with a massive 8.4% increase in page conversions and 9.2% uplift in average order value, as evidenced in a joint retail and lead generation study by Google and Deloitte. The message is clear: you can spend budget getting people to your storefronts, or can you can make sure more of the people you’re already paying to acquire actually engage and convert. In short, take care of the basics, and the material gains will pay for them many times over. 

A practical readiness agenda for marketers 

So, all of this adds up, but how do brands tackle the discoverability shift? Not by frantically trying to transform every lever at once. Think about responding to an initial series of macro maturity measures and goals, and how priority pilots can be run, analysed, and layered up into wider transformation programmes. Assessing brand capability (and capacity) to respond to questions like the below create a practical framework of factors. 

• Can we categorise intent? 
• Can we respond intelligently? 
• Can we create personalised experiences at scale? 

Final takeaways 

In an ever-evolving and noisy eCommerce landscape, it’s easy to lose sight of the most impactful and critical efforts. Make AI discovery a measurable growth investment, not another innovation initiative. Storefronts are not disappearing, but powerful evolving forces are changing the route customers take to reach them for the long term. To survive and thrive, brands must be visible and useful to AI systems, well-positioned at the moment of intent, and ready to deliver a distinctive experience when customers arrive. Brands that move to do so now can capture emerging demand more efficiently, defend market share, and create a durable advantage before competitors beat them to it.