Research-Led Metadata Strategy to Enhance Sales Content Discovery

• Role: Lead UX Researcher & Information Architect
• Skills: UX Research, Metadata Modeling, Information Architecture, Stakeholder Alignment

Context

Our client was consolidating sales enablement content from 12 separate WordPress sites—each owned by a different business unit (BU)—into a single content management system (CMS). This migration presented an opportunity to standardize metadata and improve search and discovery for sales staff.

Key Challenges

• Stakeholder Resistance: BUs were protective of their existing systems, believing they already met seller needs.
• Technical Constraints: The new CMS supported faceted search but only with flat lists of values—no hierarchical taxonomies, which the enterprise typically relied on.

Approach

Grounding Decisions in User Research

• Led contextual interviews with 16 sales staff to map workflows, pain points, and content discovery behaviors.
• Designed and deployed a survey to validate interview findings at scale.
• Analyzed search logs to identify common queries and unmet content needs.
• Synthesized findings into two seller personas that directly informed metadata requirements.

Metadata Strategy & Modeling

• Defined attributes for search filters based on validated user needs.
• Designed workarounds to accommodate CMS limitations without compromising usability.
• Collaborated with developers and content owners to align technical feasibility with user goals.

Impact

• Improved Findability: Sales staff gained one central repository with effective search and filtering, replacing 12 siloed sites.
• Centralized Governance: Streamlined content workflows and reduced redundancy.
• Enduring Stakeholder Value: Research deliverables were referenced by leadership long after project completion, influencing ongoing content strategy.

Artifacts Produced

• Two detailed seller personas, highlighting information needs.
• Comprehensive research findings report and presentation.
• Metadata schema documentation tailored to CMS capabilities.

Example Artifacts

Persona: “Alan”

Both the “Alan” and “Sam” personas go beyond the usual focus on goals and pain points by including specific metadata categories and detailed content needs. That approach helped me connect what Alan and Sam actually search for with the kinds of tags and filters needed to make content easier to find. It’s a practical tool that directly informed how the metadata schema was shaped to support real user workflows.

Persona: “Sam”

Sam’s persona captures the real frustration sellers face with content that’s hard to trust, incomplete, or not tailored to their customers’ needs. Even when they find case studies or materials, these often lack clear benefits, customer stories, or localization, making it tough to build a compelling sales narrative. This persona helped highlight that improving content quality and relevance is just as crucial as improving findability. Search workflow

Seller search workflows (from findings presentation)

Key Insight: Sellers struggled to find specific information because important content was embedded within larger, inconsistently structured assets like slide decks. Experienced sellers memorized where to locate details, creating a barrier for new users and causing time-consuming trial and error. While the idea of breaking assets into smaller “content chunks” was explored, sellers preferred consistent asset structures and clearer labeling. Our research highlighted the need for improved search and preview functions in the CMS, along with user-informed taxonomy design and training, to help sellers efficiently find the content they needed. Metadata recommendations for sales enablement content

Recommendations for content metadata / re-tagging (from findings presenation)

Key Insight: Sellers’ search queries frequently included asset type names and taxonomy terms, such as “roadmap” or product codenames like “Ice Lake.” Analysis of top search queries showed strong alignment between what sellers searched for and existing Intel taxonomies. Sellers often combined multiple concepts in a single query (e.g., “ai gold deck”), blending subject areas with asset types. This highlighted an opportunity to improve the search system by automatically matching query terms to taxonomy filters, helping sellers get more relevant results without extra effort.