Navigating the Shift to Agentic Commerce: Webinar Recap

Many organisations are already seeing shifts in visibility, traffic and performance driven by agentic systems, exposing weaknesses in data foundations, experience continuity and organisational readiness that will only be amplified by emerging AI models.

As agentic commerce begins to reshape how products are discovered, Inviqa CTO Kaustav Bhattacharya sat down with Graeme Hardie, Commerce’s Technical Lead Architect and Solutions Engineer, to unpack what is changing, and what brands need to get right to remain visible and in control.

Here are the key insights you need to be paying attention to:

We are in the midst of the shift to Agentic.

Graeme outlined three distinct stages organisations are moving through as commerce shifts from direct, brand-led discovery to AI-mediated and increasingly autonomous models. At each stage, long held assumptions about discovery, data and experience begin to erode.

Stage 1: Shopper to merchant

This is the model organisations have been optimising for over the past decade. Customers discover products through search engines, paid media or on-site navigation, while merchants optimise listings and funnels to support this conversion path.

This model assumes that brands largely control how products are presented and discovered, with optimisation driven by keywords, campaigns and channel-specific tactics. But this assumption no longer holds as discovery shifts away from direct interaction.

Stage 2: Agent to merchant

This is the stage where many organisations already are, whether they’re aware of it or not: customers increasingly rely on AI agents to explore options, compare products and work out what best fits their needs. 

These agents, however, don’t interpret brands in the way humans do. They rely on structured, complete and reliable product data, rather than creative naming conventions or campaign led messaging. So, data quality becomes critical to whether a product surfaces at all. 

Stage 3: Agent to agent

This is the emerging model, where intelligent agents act autonomously on behalf of both customers and merchants. Agents negotiate pricing, check stock, analyse return policies and refine recommendations while the customer is off doing other things.

At this stage, human intervention is no longer part of the moment-to-moment loop. Control shifts upstream, into the data, rules and constraints that guide how agents behave. Organisations that have not embedded guardrails into those foundations risk losing control quietly, rather than through obvious failure.

Here’s how you can prepare, and why it’s important

To capitalise on the agentic experience, organisations need complete, accurate and structured owned data. This is the deciding factor of whether agentic systems can interpret, trust and surface a brand’s products. 

“It’s often the unsexy answer, but it’s the most important answer.”

Graeme broke preparation down into three areas that consistently surface as constraints across organisations.

1.    Knowing where your data is and having it visible
Across organisations, we consistently see data fragments across ERPs, PIMs and even spreadsheets, but in an agentic context, this fragmentation of data is incredibly problematic. If it can’t be accessed, unified, and interpreted consistently, agents can’t act on it reliably.

2.    Cleansing and enriching your data so it’s discoverable
Beyond visibility, data must be cleansed and enriched so that it can be accurately interpreted by AI-driven systems. This means including materials, colours, sizes, sustainability information, product highlights and question-and-answer pairs that reflect how customers ask for products, so that these products don’t disappear from agentic or conversational discovery.

3.    Being agile
AI-based channels are changing constantly and requirements for how data needs to be structured, formatted and supplied are evolving week by week. Your data operations and tooling need to be able to adapt without months of rework, otherwise visibility gaps widen as channels evolve.

The invisible shelf, and the cost of incompleteness

One of the most immediate effects of agentic commerce is the emergence of what Graeme described as ‘the invisible shelf’.

If data completeness does not exceed roughly 95%, products may not appear at all in agentic search and answer engines. Organisations are already seeing up to a 50% drop in visibility for products that fall below these thresholds.

In this environment, data compliance becomes a growth lever rather than a constraint. Accurate pricing, availability, stock levels and structured attributes directly determine whether products are included in conversational journeys.

At the same time, a new acquisition battleground is forming. Discovery is no longer limited to search engines and social platforms. LLMs and answer engines now expect richer, more standardised product data than merchants have historically supplied.

Plus, as Kaustav points out, ROAS suffers immediately when product feeds are incomplete. Missing attributes mean your products won’t surface in Google Shopping carousels, in Bing’s AI-powered answer engine or in Google’s AI Overview panels.

This isn’t a reason to skimp on the on site experience

Despite the seemingly outward focus on how to show up in conversational search and agentic channels, Kaustav did raise an important point: you can’t ignore the on-site, owned, user experience. 

As customers increasingly begin their journey with agentic – an experience which feels natural, conversational and highly personalised – the transition to the owned experience can’t then introduce friction. If customers are met with confusing layouts, outdated UI patterns, or long forms, the experience becomes jarring, breaks the flow of the journey and erodes trust.

Agentic and owned experiences, therefore, have to work together, practically as one system, to retain customer confidence and conversion.

The time to act is now

What organisations also have to contend with is the speed of the shift and the pace of adoption. 

Like mobile, agentic AI represents a fundamental shift in how people interact with digital ecosystems. But unlike mobile, agentic AI doesn’t face hardware and infrastructure barriers, allowing agentic adoption to grow much faster.

“ChatGPT took two months to reach a million users. that changes how fast organisations need to respond.”

Beyond consumer adoption, the impact of agentic commerce is already being felt across the industry. Brands are seeing unexplained drops in organic visibility and changes in how customers arrive on their sites. If you wait to act, existing weaknesses in data, experiences and operating models will be amplified by AI driven systems.

What you need to focus on today

Agentic commerce is already reshaping how products are discovered and selected. Preparing for it requires you to make deliberate decisions, including:

• Accepting that agentic commerce is here, and planning accordingly rather than treating it as experimental.
• Treating product data as a core experience asset, not a back-office concern.
• Adapting to evolving discoverability requirements across LLMs, answer engines and agentic channels.
• Continuing to invest in owned experience quality, recognising it as the point where trust is either reinforced or lost.
• Committing to continuous refinement, combining existing data foundations with new conversational signals over time.
• Getting in touch with Inviqa. We can help you assess your organisation’s readiness for agentic commerce across data, digital experience and technology foundations. This includes identifying where visibility and control are already at risk and helping teams prioritise the changes that matter most in the next 12 to 24 months. Get in touch today to have a chat.
   

For the full discussion between Kaustav and Graeme, including additional examples and detail, you can watch the complete webinar recording here.