Usage: how many people use AI and for what
Insights > Usage: AI adoption, use in allowed tools, how often AI is used, where it goes and what it is used for. Counts only, never individuals.
Insights > Usage shows how many people in your organisation use AI, in which tools and for what. The screen measures the use itself and says nothing about productivity or what that use yields. Everything you see is a count: what someone typed into an AI tool appears nowhere, and the figures carry no user ID.
Usage in the dashboardUnder Insights in the menu.Period and filters
Period sets the time frame for the figures, from Last 7 days to Last year, or a time frame with your own dates. When Department insights is on, you can also pick a department; in Standard mode the period is then at least 7 days. Under Filters you select up to 8 Tools. The same bar with period and filters also appears on Compliance and Sensitive data.
The four figures
- AI adoption is the number of seats that used AI this month, for example "12 of 40 seats AI-active this month".
- Prompts sent is the number of messages sent to AI tools in the chosen period, with the change from the previous period.
- AI tools in use is the number of distinct tools used, together with how many of them are new.
- Use in allowed tools is the share of AI use that goes through tools you have allowed.
AI adoption and usage frequency are monthly figures for the whole organisation, so the chosen period and filters do not change them. If you allowed only the business version of a tool and the extension cannot tell which account someone is using, those visits are left out of the share. The figure shows how many visits this affects.
Where and how often AI is used
Where AI use goes shows, for each allowed tool, what share of all AI visits went to that tool, largest first. The remaining tools are under Show all tools. The Not allowed or no decision yet row takes you to Compliance. Clicking a row opens the tool page, which is also where you make the decision about the tool.
How often AI is used splits your team into Daily, Weekly, Occasionally and Not active. The extension works out on the device itself which bucket the user falls into and reports only the result once a month, as +1 in one bucket. As a result, the dashboard never sees who is in which bucket. The breakdown appears from 5 active users. Someone who has the extension in two browsers may count twice, and the card says so.
Usage over time shows the number of prompts per week and the number of tools in use. With a tool filter, you see visits per week per tool.
What AI is used for
BeeSensible sorts use into seven use cases: Writing communication, Marketing & content, Summarizing, Rewriting & editing, Translating, Coding & technical and Analysis & knowledge. Click a use case to see in which tools that work takes place.
The label comes from the first message of a conversation. Our own servers, which run at European providers, assign that message to a use case and then discard it immediately, so that only the label is counted. If the message cannot be classified, the prompt counts as Unclassified.
Figures per tool
Click a tool to open its page in the register. Under Usage you find Visits, Share of all AI visits, Days active, Last seen and Installs this month. That last figure counts browser extensions rather than people, and the desktop app is not included. With one to four installs it shows "fewer than 5", so that a small number cannot point to one colleague.