In a post dated September 16, 2026, OpenAI described how administrators and business leaders can use analytics in the ChatGPT Admin Console to connect AI usage to business value across ChatGPT Work and Codex. According to OpenAI, the Usage view brings together active users, credits, and token usage to show where adoption is growing and where spend is concentrated. A task classifier in the Insights section groups a sample of messages into use cases and tasks—such as software engineering (feature development, code maintenance) and sales & revenue (account research, planning)—shown in an Overview tab and a Use cases table with credit, message, and active-user breakdowns. Task details include Models, Reasoning, and Speed breakdowns, plus a Plugin leaderboard and Skills view intended to help admins target training. For engineering teams, the Outcomes view tracks Codex contributions to merged commits and lines of code alongside code-review activity. OpenAI states that an Admin plugin can compare adoption and spend and generate reports, including a leadership deck, while an Admin API lets teams automate reports and combine analytics with business-system data. OpenAI notes that usage data is a starting point, and that business owners must supply context on what changed and whether results improved. All screenshots use illustrative demo data.
- Post published by OpenAI on September 16, 2026
- Analytics cover ChatGPT Work and Codex in the Admin Console
- Includes Usage, Insights (task classifier), and Outcomes views
- Admin plugin and Admin API can generate and automate reports
- All screenshots use illustrative demo data
What it means for you
OpenAI is adding dashboards that show which teams use ChatGPT and Codex, what they use it for, and how much it costs. These tools are aimed at organizations with enough seats to justify an admin tracking adoption and spend. For most small businesses, the real point buried in here is the method, not the product: usage numbers alone don't prove value—you have to compare a task before and after AI and see if time, quality, or revenue actually moved.
Try this
Pick one repetitive task your team already uses AI for, write down how long it took and how good the output was before, then measure the same thing this month. That comparison is the 'business value' link—no admin console required.
Who should care
Businesses on ChatGPT Work or Codex with enough users that someone is responsible for managing seats, spend, and rollout decisions.
Skip this if
You're a solo user or small team on a standard ChatGPT plan without the Admin Console—these analytics features don't apply to you.
Sources: OpenAI — read the original