What does Claude training look like when nobody on the team has used it?
A normal starting point, and in some ways the easiest, because there are no habits to undo. The first session is foundations: what it is, what is safe to put in, what never goes in. Then everyone does one of their own low-judgement tasks with it, a piece of research, a draft, a summary of something long, and checks the result before anyone relies on it. People leave having felt it work on their own job, which is the only thing that reliably changes behaviour.
Is it connected to our email and files, or does it only see what we paste in?
Only what you give it, unless someone has deliberately connected something. That distinction comes up in most rooms, and it is worth being clear about early, because people either assume it can see everything and get nervous, or assume it can see everything and are disappointed when it cannot. What it can reach is a decision you make, not a default you inherit.
Should staff Claude accounts allow training on conversations?
That is a decision to make on purpose, not inherit, and it is the first thing we settle in the room. The choice only exists on consumer accounts, and it changes retention: five years if you allow it, and if you decline, anything you delete is purged within thirty days, though what you keep stays kept. Business-grade accounts, Team, Enterprise and the API, are outside the choice entirely. Which way to set it depends on what your team will put in, which is the foundations conversation.
Why does the output sound nothing like us?
Because almost nobody tells it what it is doing. Most people paste a question into a chat with no instruction at all and then judge the answer, which is like briefing a new starter by saying nothing and being disappointed in them. In a session we set up a workspace for one job with the rules written in: how you speak, what you never say, what good looks like. Plain instructions in your own words, and the difference is immediate.
Can one session cover Claude alongside the tools we already run?
Yes, and it usually should, because most teams already run a mix. The habits transfer between assistants. What deserves the time is where each one fits the jobs your team actually does, so people stop treating the tools as interchangeable and start picking deliberately. We train that judgement on your work, not on a comparison chart.
How do you use Claude in your own business?
On client work every week: drafting, summarising, and assembling repeatable processes, always with a person approving anything that leaves the business. The exercises in a session are versions of jobs I run myself, which is why they are jobs rather than examples.
