Evergreen
Playbooks
Evergreen, plain-English guides to putting AI to work in a real business. No hype, no jargon, sources on everything.
- PlaybookSeptember 10, 2026
The Letter on the Counter You Do Not Understand
Nine days of dread, four minutes of work. What it says, what it wants, what the deadline is, and what happens if you ignore it.
Read· 2 min read - PlaybookSeptember 9, 2026
Reply to the Reviews You Have Been Ignoring
The reply is not for the person who wrote it. It is a sample of how you behave when something goes wrong, posted where every future customer reads it.
Read· 1 min read - PlaybookSeptember 8, 2026
Six Questions to Ask a Contract Before You Sign It
This is not legal advice. It is the list of things to ask the person who gives it. Six specific questions, each answered with the exact line it came from.
Read· 2 min read - PlaybookSeptember 7, 2026
The Receipt Pile, as a Spreadsheet
You will not do them all in January. One photo of six receipts, one prompt, and a four column table you can hand to your bookkeeper, with the unreadable ones flagged instead of guessed.
Read· 1 min read - PlaybookSeptember 4, 2026
Talk the Quote. Do Not Write It.
The quote has sat for three days because you do not want to write the email, not because you do not know the number. Say it out loud once and get a sendable estimate back.
Read· 1 min read - PlaybookSeptember 3, 2026
Photograph the Job Notes You Are Never Going to Type Up
Three weeks of handwritten notes in the glovebox, and the follow ups inside them are worth real money. One photo and one prompt turns them into typed notes and a task list.
Read· 1 min read - PlaybookSeptember 2, 2026
The Email You Will Regret by Thursday
One prompt, sixty seconds, before the angry email leaves your outbox. Keeps the deadline and the consequence intact, removes the sentence you would have winced at.
Read· 1 min read - PlaybookSeptember 1, 2026
What the Free Version of ChatGPT Actually Gives You
Nobody selling you AI tells you where the free version stops. Here is the honest map, tested this week: what is unlimited, what is capped at three a day, and the one silent limit that will quietly give you a wrong answer.
Read· 1 min read - PlaybookAugust 9, 2026
5 Jobs AI Can Quietly Take Off Your Plate This Month
Not a list of tools. Five specific jobs, the kind that leak money or eat your evenings, that AI systems handle reliably today: missed calls, review replies, invoice chasing, no-show reminders, and turning one piece of work into a week of posts.
Read· 2 min read - PlaybookJuly 26, 2026
"I Don't Know Where to Start." Start Where the Money Leaks and You Can Count It.
The last and most honest reason people stall on AI is that the whole thing feels like a fog with no edges. The fix is a single rule that cuts through it: start where the money is leaking and where you can count it. Missed calls, dead customer lists, unpaid invoices. Pick the one that bleeds most, fix just that, and measure it.
Read· 3 min read - PlaybookJuly 25, 2026
"What if It Says the Wrong Thing to a Customer?" A Human Is on the Final Call.
The fear of AI embarrassing you in front of a customer is the most responsible objection on the list, because your reputation is the whole business. The answer isn't to promise AI never makes a mistake. It's how a good system is built: guardrails on what it can say, and a human on the final call, so nothing reaches a customer unreviewed.
Read· 3 min read - PlaybookJuly 24, 2026
"I Tried an AI Tool and It Flopped." You Tried a Tool, Not a System.
If an AI tool burned you once, that's not a reason to write off AI. It's evidence of the single biggest split in the research: buying a system that someone installs and runs succeeds far more often than a tool you're handed and left to figure out. You didn't try AI and fail. You tried the version that fails most, the DIY one.
Read· 3 min read - PlaybookJuly 23, 2026
"My Business Is Too Small for AI." Small Is the Advantage.
Owners assume AI is for big companies with big budgets and IT departments. But this week's research says the opposite: the giant companies are the ones failing at AI, precisely because of their size. The moves that make AI work, pick one problem, decide fast, wire it into simple workflows, are all easier when you're small. Small isn't the disqualifier. It's the edge.
Read· 3 min read - PlaybookJuly 22, 2026
"AI Will Make My Business Feel Cold." It Does the Opposite.
The fear that AI will strip the humanity out of a business is a good instinct, because the personal touch is often the whole reason customers choose a small business over a big one. But used right, AI does the robotic work, the reminders and the follow-ups, so the owner has more time for the human parts, not less. It protects the warmth instead of replacing it.
Read· 3 min read - PlaybookJuly 21, 2026
"I'm Not Technical Enough for AI." You Don't Build It. Someone Installs It.
A lot of owners count themselves out of AI because they picture themselves wrestling with software they don't understand. That picture is wrong. The version that works for a small business isn't something you build or operate. It's something someone installs for you, and then it runs on its own. You press zero buttons.
Read· 3 min read - PlaybookJuly 20, 2026
"It's Too Expensive for a Business My Size." The Cost Isn't the Tool.
The most common reason a small business owner waves off AI is price. But the sticker on the tool is the wrong number to look at. The real number is the money already leaking out of the business every week, unnoticed, while you decide whether the fix is worth it. Line those two up and the math flips.
Read· 4 min read - PlaybookJuly 19, 2026
What the 5% Do Differently
All week the data said the same thing from four directions: most AI projects fail, and they fail in the wiring, not the technology. But a small group wins. Here is what that 5% actually does differently, and why every one of those moves is easier for a small business than for the giant companies the failure stats came from.
Read· 5 min read - PlaybookJune 17, 2026
MCP, Explained: How to Wire One AI Into the Tools Your Business Already Uses
There is now a single standard, called MCP, that lets one AI assistant securely plug into your email, your spreadsheets, your notes, and the other tools you already run. It went from one company's experiment to shared industry infrastructure in under two years. Here is what it is, a real workflow it makes possible, and how to start without handing over the keys to everything.
Read· 3 min read - PlaybookJune 11, 2026
Fable 5 Ultracode: The Complete Guide (With Exact Prompts)
We run Claude's Fable 5 model with the ultracode setting every day, and the combo can turn one instruction into hours of multi-step work. This is the complete field guide to Fable 5 ultracode: what it actually is, what it costs, when it is worth it, and the exact prompts that get the most out of it. Field notes, not theory.
Read· 4 min read - PlaybookJune 9, 2026
Where to Let AI Act in Your Business (and Where Not To)
AI has gone from answering questions to taking actions. The useful skill now is not picking a tool. It is knowing which work is safe to hand off, and which is not. Here is a simple test.
Read· 3 min read
Stay ahead of AI in five minutes a week
Each week I send the one development in AI that actually matters for operators, in plain English, with sources. No hype, no jargon, no firehose. It is free, and you can unsubscribe in one click.