Do this: Paste this week's five worst customer messages into your AI assistant and get the one policy line that would have prevented three of them.
PASTE THIS
You are helping a small business find the cheapest fix for repeated customer complaints. Below are five real messages from this week. Return three things: the causes that appear in more than one message, ranked by how many messages each explains; one sentence to add to our quote, booking form or FAQ that would have prevented the top cause; and the one internal step to change so it stops repeating. Do not recommend new software. Messages follow:
You'll get: A ranked cause list, one line you can publish before Friday, and the internal step that kills the repeat. No purchase, no integration, about two minutes.
The one-page AI payoff log
Five columns, one row per tool, filled in every Friday:
The rule that makes it work: a tool with nothing in column three for two Fridays running gets cancelled, no debate and no pilot extension. Keep the log beside the profit and loss, and open it at renewal.
Apollo's chief economist ran the numbers on the S&P 493, the index minus the seven firms selling AI. Margins flat. Revenue per worker flat. No productivity gain. Not an argument against the tools. An argument against buying them as sold: seats, licences, three-year roadmaps. Point one at your own messy data, count what comes out, or stop paying.
Apollo's Torsten Slok looked at the S&P 493, the S&P 500 minus Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla, and found profit margins and revenue per worker flat, with no evidence AI is lifting productivity. Deloitte's research points the same way, The Register reports. So what: this week, take the AI tool with the biggest bill, write down what it produced since last Monday, and cancel it on Friday if the answer is a feeling rather than a number.
Anthropic's filing puts a price on your renewal
Anthropic filed to list on Nasdaq as ANT, with an $80.7 billion net loss on $16.3 billion of revenue for the first half of 2026 and 80% of revenue billed by usage, CNBC reports. The prospectus carries nearly $800 billion of compute commitments over ten years. So what: before your next AI renewal this quarter, get a written cap on mid-term price rises, because public margins get defended on your invoice.
A 175% price rise sent Tesco to open source
Tesco is moving store and edge virtualisation from VMware to Proxmox, and suing VMware over a contract dispute, The Register reports. During negotiations, Tesco says VMware offered a 175% increase. So what: if your renewal quote jumped this year, do not sign it this week. Price one alternative and check whether your perpetual licence still covers upgrades.
AI search is eating the clicks your site needs
AI Overviews and AI Mode pushed clicks off publisher pages: New York Times search traffic fell from 44% to 35.5% in three years, and one UK group measured click-through down 89%, The Verge reports. So what: this week, compare your Search Console clicks for the last 90 days against last year, and move next quarter's content budget toward email if the line is down.
This week's question: a software vendor emailed a price rise eight months into a three-year contract. Is that allowed? Only if a clause says so. Look for language permitting unilateral changes to fees, and the notice period attached to it. If the price sits in a signed order form, a mid-term rise is a change to the agreement and needs a fresh signature. Reply in writing, ask which clause they are relying on, and request the increase be deferred to renewal. If they refuse, ask for a 30-day exit and a full export of your data.
What would make you cancel an AI tool first?
A. It produced nothing measurable
B. A price rise at renewal
C. The team quietly stopped using it
D. A data or privacy concern
Reply with one letter — results next issue.
Anthropic's IPO paperwork runs to roughly 240,000 words. How many of its pages are given over to risk factors?
Answer tomorrow — and no, we are not printing it here.
YESTERDAY'S ANSWER
The UK government announced a £100 million investment this week in a 'world first' national facility to defend against AI-enabled biological and chemical threats. What is the facility called?
The AI Biosecurity Lab. The £100m investment announced by the UK government establishes the world's first national AI biosecurity laboratory to assess biological and chemical risks (BBC Technology, https://www.bbc.co.uk/news/articles/cqdjv8xllvyyo).
About 5 minutes to read. One thing you can use today every issue.
Every Friday the desk answers one reader's AI question in depth. Reply with yours — the best one gets the full writeup.
Know an owner drowning in their inbox? Forward this — the 2-Minute Win is worth 2 minutes.
Forwarded this? Subscribe here
Sent by AI for Everyday Business.
8735 Dunwoody Place STE R, Atlanta, GA 30350