One team, one day. Everyone leaves using AI on their own work.
Every team asks me the same thing on the fit call: what does the day actually look like for us. This page is my answer for one team. The company is made up. The shape is the one I run.
The day is about the people in the room. How AI and large language models work, how to brief them so they sound like you, and how to use the AI already sitting in the tools you pay for. Nothing gets built for the team. The team learns to do it.
They have the seats. They write emails with them.
A Series B billing platform. 9 AEs, 5 CSMs, two managers, one VP of Revenue.
- HubSpot
- Gong
- Slack
- Notion
- Claude seats for the whole team
Not a real customer. I picked a stack most GTM teams will recognize, because the day is about the tools you already own.
“We bought the seats six months ago. People use it to write emails and summarize calls. That is about it. Nobody has shown them what else it can do, and I cannot, because I am in the same spot.”
The VP, on the fit call, in roughly those words. I hear a version of it on most fit calls. The seats are not the problem. Nobody has shown the team how the thing works.
The morning is the same for every team. The afternoon is only yours.
This is the part that makes the day theirs. The fundamentals do not change from company to company. The stack, the seats and the tasks do, so I collect five things before I show up, and the afternoon is built from them.
Every tool the team touches in a week, and which ones already have AI features switched on that nobody uses.
Who is an AE, who is a CSM, who manages. The afternoon is built per seat, so I need the split before the day.
The five things each role does every week that they would hand off tomorrow if they could. Prep, follow-up, notes, renewals, the pipeline review.
The ICP, the positioning, the pricing page, three good call recordings. The team writes the context files from these during the morning.
Where does my data go. Can I trust it. Will it make my job go away. I collect these anonymously before the day and answer them first.
For this team the five tasks came back almost identical across the AEs: prep, follow-up, CRM notes, the Monday pipeline update, and finding out what changed at an account. The CSMs added renewals and QBR prep. That list is the afternoon.
Fundamentals in the morning. Your own work in the afternoon.
Nine to five, on-site, seventeen people in the room, laptops open from the second block. Nobody watches me build anything. Everyone builds their own.
- 9:00FOUNDATION
What an LLM actually is
In plain words, no slides of architecture. What it is good at, what it is bad at, why it makes things up, and where your data goes when you paste a contract into it. The anonymous questions get answered here, first, before anything else.
- 9:45FOUNDATION
How to brief it so it sounds like you
The difference between a question and a brief. Context, role, what good looks like, what to leave out. Everyone rewrites one of their own prompts live and sees the output change in front of them.
- 10:30FOUNDATION
The foundation that makes it stick
The context files: the company, the ICP, the pricing, how the team talks, written as files the AI reads every time. One shared place for them. A folder system so a brief made on Tuesday is findable on Friday. The team writes the first versions in the room.
- 11:30YOUR STACK
The AI already in your stack
Tool by tool. What HubSpot's AI does and does not do. What Gong's call summaries are good for and where to stop trusting them. How Slack and Notion fit. The point is one map of which tool does which job, so nobody has five tabs doing the same thing badly.
- 12:30
Lunch
The context files get saved into the shared folder. The afternoon runs on them.
- 1:30YOUR SEATS
Every seat on its own work
Three rooms, one per role. AEs build their call prep and their follow-up on real accounts from their own book. CSMs build their renewal read and their QBR prep on real customers. Managers build the Monday pipeline read and a coaching note from one call recording. I move between rooms. Each person leaves with three workflows set up, not watched.
- 3:30YOUR SEATS
Show and tell
Each room shows one workflow to the other two. The AEs usually want the CSMs' renewal read. The managers usually want everything. The best ones go into the shared folder with a name on them.
- 4:15THE PLAN
The next 30 days
Who runs what, where the team shares what worked, and the date of the champion check-in. Written on the wall with the stack map.
Three workflows each, on their own accounts, by five o'clock.
Call prep that reads the account, the last three calls and the ICP before they do. A follow-up draft in their own words, ten minutes after the call ends. A weekly read of their pipeline in plain language.
The hour of tab-hopping before a call becomes ten minutes of reading and fixing.
A renewal read ninety days out, built from usage, tickets and the last calls. QBR prep from the quarter's data. A health note on any account in under a minute.
They stop finding out about risk from the customer.
The Monday pipeline read with the questions to ask each rep. A coaching note from one call recording, in the team's own language. A view of which workflows the team is running and which died.
They inspect the work instead of chasing the summary.
- ✓The context folder: company, ICP, pricing, how the team talks, as files the AI reads
- ✓The stack map: one page of which tool does which job, and which AI features are on
- ✓Three workflows per person, set up on their own accounts and saved in the shared folder
- ✓The anonymous questions, answered in writing so the next hire reads them too
- ✓The 30-day plan with an owner on every week
The day is where it starts. The month is where it sticks.
Most AI training dies the Friday after. The plan is short on purpose: one channel, one owner a week, one call at the end.
Everyone runs their three workflows on real work. One Slack channel for what worked and what came back wrong.
The misses go back into the context files, not the prompts. The managers read the channel and say which workflows to keep.
Each room picks workflow four and builds it without me. The stack map gets corrected where it was wrong.
A one-hour ask-me-anything with the VP, the champion for this one. What stuck, what died, what the team ran into, and what they want to learn next.
Whether the managers belong in the room for the afternoon.
In this example they get their own room, and the pipeline read they build is the one that keeps the other workflows alive. I have also seen a rep stop experimenting because their manager was two seats over. I would run it with the managers in the room and watch the reps for the first hour. If the room goes quiet, I would move them out. If you have run it either way and know, I want to hear it.
This example is the company session, the full day, which includes the one-hour ask-me-anything with the champion a few weeks after. The half-day team workshop runs the morning and one seat's afternoon. The virtual cohort spreads the same day across a few live sessions with the building done in between. The three formats are on the workshops page.
What is your team doing with the seats you already pay for?
Fifteen minutes on a call and I will tell you what the day would look like for your stack and your seats, and whether a workshop is the right shape for it.
Book a fit call