Skip to main content
Haney Strategy

The Friday run · a real workflow, recreated

Fri · 6:00 am

Mail

0

threads

Calendar

0

meetings

Transcripts

0

calls

Client folders

0

files touched

Client sites

0

new posts

What running a Claude workflow can look like.

An example of a real workflow I run with Claude every Friday, for client update briefs.

Book an Introductory Call

Weekly brief · draft

0.0

Run complete · 12 minutes

The run is automated. The output stays gated, pending human review.

5 sources · 1 draft · grade 9.2

Email draft · Fri 6:12 am

9.2

Email draft · Fri 6:12 am

9.2

Waiting for Jim's review

Meanwhile, in most companies

AI, in a browser tab.

Somebody has ChatGPT open in a browser tab. They ask a question, copy the answer, and paste it into an email. That is AI inside most companies right now: a smart person, a good tool, and a lot of copying and pasting in between.

Now set it beside the Friday run at the top of this page. Nobody types. Nobody pastes. Five sources get swept, a draft writes itself and grades its own work against a bar I set, and the finished piece sits in a slot waiting for the one thing a machine should never own: a person's sign-off. The browser tab is using AI. The Friday run is what running with it looks like.

And if the browser tab felt familiar, here is the comfort.

In January 2026, a survey of 5,000 knowledge workers at companies of a thousand people or more found:

55%

use AI every week

15%

have found a valuable use for it at work

Source: Section, The AI Proficiency Report, January 2026

Even at that size, the browser tab is the norm. Running with AI is just newer.

Claude Team Training

In four weeks, your team runs its real work with Claude.

And proves it to you.

Pick the team. One manager, about ten people, one set of real work: the client updates, the quotes and proposals, the service follow-ups, and the reports that team produces every week.

Four weeks later, all of it runs with Claude, Anthropic's AI assistant. Built by your people, in plain English, with me in the room, in a few working sessions built around the team's normal week. Documented, so it holds after the four weeks end. And every workflow ends at a person: Claude does the assembly, your people keep the judgment.

How the four weeks go

01

The big session. Everyone builds a real piece of their own work, and leaves with it running.

02

The build sessions. The team builds its top workflows live with me, and runs them between sessions.

03

The demo. Your team demos their own work, running, to you. That is the finish line.

The proof

The Friday run up top is a recreation. The original is one motion of my Friday mornings.

I started where almost everybody started, in ChatGPT: email, copy, novelty. The shift came when I put Claude on real work and the work came back complete: proposals, business plans, and eventually the code behind this site. Today I run my whole practice with Claude, every single day: the research, the proposals, the client dashboards, the writing, the build. The Friday run is just the motion small enough to fit on a page.

MIT

AI & Digital Transformation certificate

26

Years across sales, marketing, technology, and AI

31

Months of hands-on AI build, daily

As of July 1, 2026

200+

MSP leaders trained on AI at IT Nation Connect, 2024 and 2025

The price

One offer. One price per team.

$6,500

Per team

About ten people

Four weeks

Your team demos it

The number covers the team, the four weeks, and the finish line. Bigger team? We split into cohorts, same price per team. We'll shape it on the call.

The intro call

Bring the team you'd start with.

It's free, and it's specific. Tell me which team you have in mind and what their week looks like. Sales, service, or the back office all make a strong first team. We'll map their four weeks together, on the call.

Book an Introductory Call

Questions

What owners ask before they start.

What is Claude Team Training?
Claude is the AI assistant built by Anthropic; Haney Strategy is an independent practice, not affiliated with or endorsed by Anthropic. Claude Team Training is a four-week engagement for one of your teams: one manager, about ten people, and the real work they already produce every week. The team builds that work into named workflows that Claude carries out: gathering, assembling, drafting, and checking, with a named person reviewing every output before it goes anywhere. It ends with the team demoing their own work, running, to you. The training is the means. Work that runs is the point.
Do my people need to be technical?
No. If someone can describe a piece of their work in plain English, they can build it with Claude and me in the room. The whole engagement runs in plain language, and every workflow is documented the same way, so it holds up long after week four. The hard part is not the typing. It is knowing which work to hand over, how to gate it, and how to make it stick. That is the part I bring. Anthropic's own free courses are the prework; I never charge you for what the vendor gives away.
What Claude accounts or licences does my team need?
A paid Claude plan for each person on the team, bought straight from Anthropic at their published prices. That is the whole shopping list. Which plan fits depends on the work your team does, so it is one of the first things we settle on the intro call, and it lands in writing with the ground rules: a company workspace rather than personal accounts, so the admin and data controls belong to your business. I sell no licences and take no commission on any of it, which keeps the recommendation honest. Its only job is fitting your team.
What about our data?
That is the first thing we settle, and we settle it in writing. The engagement starts with ground rules: what Claude may see, what it may never see, and who signs off on what. The recreation at the top of this page shows counts and timings and never a word of content. That is deliberate, and it is the same discipline your team learns. The ground rules also cover the data itself: where it lives, how long it is kept, and whether it trains the model, set against Anthropic's published commitments, in writing.
Why Claude?
I stay current on all of it. I go deep on one. My advisory work stays neutral across the AI market, and Google's Gemini still makes my blog images. Claude is where my own work gets done every single day, and that is the honest reason it is the one I teach.
How is this different from a workshop?
A workshop puts skills in the room in a day. This engagement lives inside one team's real work for four weeks and ends with that work running, demoed to you. One builds skills. This one rebuilds the work. Different jobs, and they stack.
What happens after week four?
Your team keeps everything: the workflows, the documentation, and the weekly rhythm the manager uses to keep the work honest. It is yours, built by your people, running inside your company. Where it goes from there is a conversation the demo usually starts on its own.