What Happened
On October 26, 2026, Amazon Connect Customer added scoring for interaction outcomes including customer satisfaction, churn risk, purchases and whether a need was resolved through self-service. Each metric has a default definition that businesses can customize. Results are available in analytics dashboards and through APIs. No version number was specified in the announcement. [1]
OpenAI also described OpenAI Signals, which examines how people use ChatGPT for work across 165 activities in nearly 100 countries and tracks changes over time. The announcement does not specify an API or a version number for Signals. [2]
Why It Matters to Businesses
Connect’s change lets teams assess whether interactions achieved business outcomes, rather than relying only on efficiency measures. Signals offers broader context on work-related ChatGPT usage, but it is not a measure of an individual company’s productivity or return on investment. [1][2]
Kimbodo Engineering Perspective
Outcome definitions matter more than the dashboard displaying them. A churn-risk score, a completed purchase and a self-service resolution require different validation methods. We would treat default metrics as starting points, document any custom definitions and compare scores with observed outcomes before using them to direct customer-service decisions.
How We Would Implement It
- Choose a small set of outcomes tied to specific decisions, such as follow-up prioritization or self-service improvement.
- Record each metric’s definition, owner and permitted use; test defaults against historical interactions before customizing them.
- Retrieve Connect results through the available APIs for governed reporting, while retaining dashboard access for operational review. [1]
- Monitor score changes and compare them with subsequent purchases, resolutions and customer feedback.
Risks, Costs and Security
Scores may be incomplete or misleading if interaction data is missing, definitions change or a prediction is mistaken for a confirmed outcome. Restrict access to customer-level results, minimize exported personal data and log API access. Integration, validation and ongoing monitoring add costs; neither announcement provides pricing details for these uses. [1][2]
Where Kimbodo Comes In
Kimbodo builds and operates this in production for businesses — see our AI Application Development practice, or Estimate My AI Application.
