Nadya Anjani
Senior Product Experience Management Business Analyst
How an AI Recommendation Agent Transforms Product Experience
WPP Open AI - Innovation in Action
Being part of WPP, we’re part of an incredible momentum where we are redefining how intelligence and creativity intersect. Through WPP Open AI, our agencies are pioneering real-world applications of AI across creative, data, and commerce ecosystems.
Within this framework, VML is integrating AI into Product Experience Management (PXM) - transforming manual, repetitive data review into intelligent, scalable optimization.
One key outcome: the PXM Recommendation Agent, a practical AI module designed to analyze, diagnose, and recommend precise, ready-to-implement improvements to product data.
The reality of PXM today
Every PXM Manager knows the challenge too well: thousands of SKUs, each with multiple attributes, descriptions, and digital assets — all requiring consistent, optimized, and localized data.
Manual review is time-consuming. Inconsistent copy weakens SEO. Vague attributes confuse customers. And while the intent is always to deliver rich, conversion-driven product experiences, the process often becomes an uphill battle of spreadsheets and spot checks.
Thousands of SKUs, each with multiple attributes, descriptions, and digital assets - all requiring consistent, optimized, and localized data.
The missing link: from analysis to action
Most digital transformation initiatives stop at visibility: we can now see data quality scores, missing attributes, and empty fields. But few systems bridge the gap between knowing what’s wrong and knowing exactly how to fix it. This is where the PXM Recommendation Agent enters the picture.
How the Recommendation Agent works
The Recommendation Agent performs five key steps in one automated flow:
- Scan: Extracts product titles, descriptions, attributes, and linked assets.
- Assess: Evaluates completeness, tone, accuracy, and SEO quality.
- Recommend: Suggests targeted improvements for every detected issue.
- Prioritize: Ranks issues by business impact (High / Medium / Low).
- Enhance: Generates a new version of the same file, enriched with all approved recommendations ready for syndication.
This turns what used to be weeks of manual work into minutes of intelligent optimization. This last step transforms the process from insight delivery to execution readiness.
Why it matters
When implemented, this AI-driven layer doesn’t just clean up data — it enhances discoverability, scalability, and conversion.
- Time efficiency
AI reduces manual review time by up to 90%. What once took 20 minutes per SKU now takes seconds — freeing analysts to focus on strategy, not data clean-up. - Improved scalability
Whether you manage 5,000 or 50,000 SKUs, the same level of quality control can be applied across brands, languages, and markets — instantly. - Higher data quality
Consistent, complete, and SEO-optimized product content ensures every channel reflects the best version of your brand with data quality improvements in completeness up to 50%
How to implement?
This can be as simple as a one time enrichment across your product portfolio, where an export of product data in the right structure is enough to enrich the product data directly to prove the value. Adding more input to create better output, think of SEO strategy or brand guidelines, still keeps implementation efforts to days or a few weeks.
Once value is proved VML can to embed this in your standard way of working, to embed this in the core processes whenever a new product is introduced, existing products change or requirements from customers change.
From reactive to proactive PXM
The future of PXM isn’t about more dashboards - it’s about smarter, more actionable insights that close the loop between data diagnosis and data improvement.
AI agents like this Recommendation Agent turn product data into a living system - one that learns, suggests, and evolves with your catalog. This is how WPP Open AI translates intelligence into impact - transforming product experience management from reactive maintenance into proactive optimization at scale.
