Get the business out of your head and into a digital system your team runs, your buyers trust, and machines get right.

Spent a day with the team at IWBI back in July. Eight hours, one room, everybody bringing whatever they were actually stuck on.

The morning was spent fixing access. We got them into an AI-friendly version of their own site on GitHub, which sounds like a boring morning but helped them in a meaningful way. Once a machine can read your site, everything after it gets easier.

From there it opened up. Their designers got onto some new design tools and started organizing the process and generation. A strategist got linked into Claude Code, and by the afternoon we were running experiments to generate new pages on the site.

That day is what became the Install.

The shape of it now is one day, at your place, on your machine. You get Claude Code installed and authenticated, your site sitting in a GitHub repo you own, one change you wrote in plain English and shipped yourself, and one agent running on a task you actually have.

Your hands are on the keyboard for it, and that's the point.

Most of what I sell keeps me in the loop. This one is built to take me out of it -- some people finish the day and run with it, some finish and decide they'd rather not live in a terminal.

If your team's been talking about AI without getting past the talking part, happy to walk through what a day like that looks like.

https://shanegring.com/install

Something that still gets me every time: the more successful the company, the more workarounds you tend to find inside it.

Been under the hood of operations with thousands of customers and real revenue, and the inside is usually held together with more tape than you'd ever guess from the outside.

The part I keep noticing is how it got there. Every piece of tape made sense the day it went on. A workaround while somebody was out. A process that lived in one person's inbox because that was faster than writing it down. Nobody was being sloppy, they were shipping.

It accumulates faster than anyone tracks.

So when I meet a smaller team that still fits in one person's head, what I see is the cheapest moment there'll ever be to get some of it written down. Nothing has calcified yet. Nobody's built a workaround on top of a workaround.

If you're running something small and it feels held together, you're in good company. Most things look like that underneath.

The ones who edge ahead pick the single piece of tape that's starting to drag on them, and deal with that one.

Watched a lot of people learn new tools over the years. The ones who keep the skill are the ones whose hands were on the keyboard.

Demos are the trap. Watching an expert fly through something feels like progress, and a week later nothing's changed, because watching doesn't build the muscle. Courses pile up the same way -- bought, bookmarked, gathering dust.

So when I run setup days now, I build them so I barely touch the keyboard. Four of the five working blocks are driven by the client. Their machine, their site, their team in the room. I sit next to them and talk them through it, and by the afternoon they're describing a change in plain English, reviewing what the AI wrote, and shipping it themselves.

If I drive, they've watched a demo. Demos don't stick.

One day, your hands, a working system by dinner. Details in the first comment.

A client of mine runs a software company doing about $10M a year. One of their best lead generators turned out to be a spreadsheet.

Their product helps coaches manage training programs. Thousands of customers, genuinely good tool. But plenty of coaches out there still live in Excel, and no amount of marketing was going to talk them out of it.

So instead of fighting the habit, they bought a small brand that makes Excel templates for exactly that crowd.

Then it sat for about a year. Good team, smart play. The blockers were small: design / copy / a couple of tech decisions / nobody owning it.

We came in and got it moving. Took what they had, rebuilt the site, wired up Shopify, and pulled in brand assets that'd been parked in a folder for months. Most of the work was deciding fast and clearing the little stuff that stalls these things out.

Ten days, start to finish. It went live on a Friday, and watching the weekend roll in was a blast: 30 organic leads by Monday. They posted about it once to their audience, and two days later they were sitting at 500+ signups.

Twenty days earlier, that offer didn't exist.

Two things worth stealing from it. Take a look at what your customers refuse to quit using, because that's often a door rather than a problem. And whatever's been sitting on your shelf for a year is probably a lot closer to done than it feels.

A company I work with added a COO and a CMO in the same quarter without making a full-time hire for either one.

Both roles are acting. I hold one, a colleague I've worked with for years holds the other. It's a lean organization at the stage where the founder had been carrying operations, sales, and marketing personally, and the math on two executive salaries wasn't there yet.

People ask how that actually works, so here's the shape of it.

We step into the real seat, with the title and the accountability that comes with it. We run the function, build the systems, and document everything as we go. The whole arrangement is built toward a decision point, where the company either grows into full-time versions of these roles or keeps this setup because it fits.

What the founder gets back is room. Relationships, advocacy, funding, the parts only a founder can do.

It has limits worth naming. This fits a transition or a build phase better than steady-state operations. And it works when the operator has enough time in the seat to carry real accountability, not a few hours a month.

If you're carrying three jobs and the hiring math isn't there yet, there's a middle path between doing it all yourself and a six-figure executive search.

Happy to talk through whether it fits your stage.

My favorite test for whether a website is being cared for: what happens when the owner wants one word changed?

The usual answers. A ticket into a portal, "please allow 5 to 7 business days." A favor owed to whoever built it. Or nothing at all, because it's too much hassle to bother -- so the wrong thing just stays up.

That hassle is most of why sites go stale. Small changes have to be frictionless or they stop happening, and a site nobody changes drifts away from the business it describes, one unfixed word at a time.

The bar I hold my own service to: a change request is an email. Headline swapped, photo replaced, line rewritten. You send it, it gets done, no portal in between. Because the moment changing your own website feels like filing paperwork, you'll stop changing it, and the drift starts.

Run the one-word test on your own setup this week. If the answer made you tired, details are in the comments.

Ran the site scan on a business the other day and it came back a 52.

The technical side was sharp. Clean structure, schema done right, a couple of things most small sites haven't gotten to yet. Great bones.

Then the content score came back low, and that was the whole story. A handful of thin pages standing in for a real content layer. The site was selling expertise it wasn't showing anywhere.

That gap shows up more than you'd think, and it's the one most tools walk right past. They hand you a grade for the plumbing and stop there. Whether there's anything worth reading on top of it is a different question, and it's usually the one costing you.

If you want to check your own without a tool: open your services page and count the sentences a stranger couldn't have guessed before landing there. That's your content layer. Everything else is furniture.

More on how the scan reads a site in the comments, if a second opinion sounds useful.

One of the first things worth doing inside a business is a comparison nobody enjoys: where the time went last month versus where the money came from.

Almost every team can tell me what they're busy with. Fewer can say which of those busy things pay for the place.

When you lay the two side by side, there's usually a gap. Hours pouring into work that's loosely tied to any dollar, while the offer funding everything gets whatever attention is left over.

Nobody plans it that way. Work accumulates and commitments stack up.

Here's the version you can run this week. Two columns on one page. On the left, the five things that ate the most hours last month. On the right, the five sources that brought in the most revenue. Then draw lines between the ones that connect.

The unconnected items on the left are your answer. Some of them you'll protect anyway, and that's a decision instead of an accident.

Takes about an hour of honest accounting. From what I've seen, it moves the next quarter more than most planning sessions do.