Elevate the power of your product data with AI

Sharp’s digital transformation journey

A standout moment of the morning was Jason Cort’s detailed walkthrough of Sharp Europe’s multi-year Martech transformation. The story highlighted how a legacy-heavy tech stack was turned into a modern, data-driven engine for growth.

Six years ago, Sharp recognised that its Martech stack had reached “end of life.” What followed was a staged, structured, transformation:

• A new enterprise CMS supporting 14 languages across 19 countries
• A Magento-based e-commerce platform for their extensive reseller network
• An internally developed dealer portal
• A complete replacement of a home-grown PIM system that had become so opaque that no one fully understood it

After a full RFP, Sharp selected Inriver as both PIM platform and implementation partner, establishing a single source of truth for product data across regions and channels.

Jason emphasized that although significant progress has been made, optimizing content and structure is the next step to ensure product information keeps pace with how AI ranks, reasons, and recommends.

Why AI pushes product information into the spotlight

Karl and Steve highlighted a major shift in how AI tools gather signals. Instead of relying solely on a brand’s own site, models now draw from a broad surface area that includes:

• Reseller and retailer sites
• Marketplaces and comparison platforms
• Third-party content
• Social media and user-generated reviews

This creates a profound new challenge. Karl summed up the risk of misalignment across those touchpoints:

“If it sees your product information differ, your AI confidence score drops, and your products become less likely to be recommended.”

For modern brands, this elevates product information from a backend operational resource to a frontline competitive asset.

Every attribute, every specification, every description either supports AI-driven visibility or quietly erodes it.

Consistency vs. controlled variation: How should brands think about content?

One of the most striking ideas from the breakfast was Steve’s concept of being “consistently inconsistent.” On one hand, technical data must be highly consistent. AI models and search engines rely heavily on structured information, including:

• Dimensions
• Weights
• Materials
• Compatibility details
• Variant attributes

Missing, incomplete, or conflicting data points in these fields directly impact ranking, reduce recommendation quality, and create uncertainty for AI.

On the other hand, descriptive and emotional copy benefits from intentional variation. If every channel publishes the same product description, search engines can limit how many of those pages appear in results. Steve made the point clearly:

“Google needs inconsistency, if it sees the same copy everywhere, it will limit the number of sites representing your product in the results. You can achieve this while reaching different audiences by writing differently for lifestyle, technical and value-driven consumers.”

Balancing these two requirements, rigid technical accuracy and flexible storytelling, is now central to modern product information strategy.

Structured data as the hidden engine of AI visibility

The conversation then shifted to what truly enables AI tools to “understand” products: structured, complete, and accessible data.

Steve explained that different AI tools have different preferences. Some show a strong preference for structured information, while others place more weight on natural language or visual inputs:

“Gemini really likes structured data, whereas ChatGPT prefers natural language. Omitting text about non-existent features is natural, but tabular data benefits from every attribute having a value even if it’s ‘Not Applicable’’

Empty fields in specification tables can signal uncertainty, which reduces AI confidence and visibility. This is driving concrete changes inside many organizations, including:

• Expanded attribute taxonomies
• Systematic enrichment of missing data
• Stronger PIM governance and ownership
• Additional product information pages where product complexity demands it

Karl added a note of caution:

“You might have the perfect product for a customer but if the data isn’t consistent, AI won’t show it.”

For many brands, structured data has moved from operational detail to essential infrastructure for growth.

Content variations for different search intents

AI-powered search is increasingly intent-driven rather than keyword-led. Steve described how this shift plays out in real customer behavior: “Somebody might be searching by lifestyle, somebody by technical data, somebody else for a family product”

To match that range of intent, product content strategies must include content variations tailored for:

• Technical shoppers
• Value-focused customers
• Lifestyle or inspiration-driven buyers

These variations strengthen SEO, support omnichannel discovery, and align with how AI tools refine recommendations.

This layered approach to content naturally introduces the “good inconsistency” that search engines and AI engines reward, while structured attributes stay aligned underneath.

Keeping pace with the new world of AI search

Jason also described a reality that many brands now face: Even with strong internal data governance, external representations of products can drift over time.

“AI search scrapes the whole internet, not just your website or your partners’ sites. It’s social media too.”

Sharp has responded by strengthening relationships with channel partners, consolidating product data through Inriver, expanding oversight through Digital Shelf Analytics and creating cleaner, more consistent syndication workflows.

“Our house is pretty clean now inside, now the next thing is to extend that out,” Jason said.

The priority is now to project that internal clarity outward, reducing gaps between the single source of truth and the many destinations AI uses to learn about Sharp’s products.

Human creativity still matters: Using AI without losing your voice

Towards the end of the session, the focus moved to creativity, brand voice, and the practical use of AI in content creation. Many organizations remain cautious about over-automation, and the panel agreed that balance is key.

Steve summarized it succinctly:

“AI isn’t your creative director. Your marketers still need to write the base copy, and AI helps you scale them. What is important is that this team become proficient in prompting to ensure brand alignment and tone.”

Jason shared how Sharp’s agencies now work with AI in their creative process:

“We use AI as part of the creative process, but it doesn’t replace the team.”

Today, AI is proving most valuable for:

• Translations
• Alt text generation
• Filling gaps in supplier data
• Variant-level copy
• Creating channel-specific adaptations

But the emotional resonance. the “story” of a product. still depends on human writers.

What London taught us

The London event closed with a shared recognition: product data has moved from a supporting role into a clear growth lever in an AI-driven market. Where product data once focused on internal operations, it now:

• Influences whether AI recommends a product at all
• Shapes how features and benefits are interpreted
• Guides how retailers and partners present assortments
• Impacts conversion confidence at the moment of decision
• Determines which products are surfaced, compared, and ultimately chosen

From Sharp’s multi-year transformation to Inriver’s industry insights, the morning highlighted a universal truth: In the age of AI, great product data isn’t just important; it’s everything.

Structured, consistent, channel-ready product information has become a foundation of visibility, trust, and long-term commercial growth.

One attendee summed up the sentiment of the room while finishing their coffee:

“Product data used to be something we managed. Now it’s something we compete on.”