How agent-ready is Workday?

Workday's overall agentability score of 37/100 indicates significant barriers for autonomous AI agents attempting to parse and operate its public website. The product scored 84/100 on machine readability, showing solid structured markup, but fell short on principles that govern agent interaction flow and observability. Homepage (37), pricing (31), and documentation (25) scores all reflect similar underlying gaps.

The most critical weaknesses appear in transparency (0/100), defaults (18/100), and clean handoffs (10/100). Workday provides no machine-readable activity logs, confidence indicators, or source citations that would help agents understand system reasoning. Error states lack structured context, and async operations offer minimal status feedback. These gaps force agents into guesswork and make recovery from failures difficult.

Control (72) and machine readability (84) represent relative strengths, suggesting the underlying page structure is sound. Improvements should focus on exposing operational metadata, enriching status signals, and making error handoffs actionable for programmatic consumers.

To improve agentability, Workday should prioritize the following changes:

• Add a machine-readable activity log or audit trail that records system actions in JSON format, paired with a visible activity feed in the UI to support transparency.
• Return structured error objects in both API responses and UI that include error codes, the specific field or resource that failed, and a clear next action (e.g., 'Email is invalid — enter a valid address').
• Emit descriptive status messages for all asynchronous operations, such as 'Uploading 3 of 10 files' or enumerated multi-step progress indicators, rather than generic spinners.
• Attach confidence scores or certainty labels to key outputs, implemented as a 'confidence' field in API responses or 'verified vs. best-guess' tags in the interface.
• Provide source references or input citations for generated content, linking outputs back to the underlying data or document IDs that informed them.
• Implement expandable drill-down patterns that pair a one-line summary with detailed breakdowns for important results, helping agents decide when to examine details.