
AI That Pays Back · 28 October 2026
How Do You Calculate the ROI of AI in a Business?
AI ROI = value created minus total cost, divided by total cost. Here's how to measure it honestly.
5 min read
Toro Insights
Practical notes on agentic workflows, AI that pays back, agentic coding and GTM engineering. Every post is built on systems I run, with real figures and the mistakes left in.
For owners and operators
What agentic workflows are, real examples by function, and how they're built, monitored and handed over.
Why AI rollouts fail to pay back, how to measure AI ROI, readiness assessments and lean before agents.
Lean before agents, Knowledge · Skills · Wisdom and ZPD for teams, plus the monthly Agent Log: what my agents did and what broke.
For builders and growth teams
Agentic coding in real teams, testing and reviewing AI-written code, strategic programming and the new team shape.
The GTM engineer role, AI content and SEO/GEO pipelines, and the Verqos engine as a worked example.
Want the pipeline behind these posts for your own site? See GTM engineering. Want your business to show up in AI answers? See AAIO.
Category: AI That Pays Back

AI That Pays Back · 28 October 2026
AI ROI = value created minus total cost, divided by total cost. Here's how to measure it honestly.
5 min read

AI That Pays Back · 21 October 2026
AI not delivering value? The problem is usually the process, not the tool. Here's how to fix it.
5 min read
The Agentic Playbook
A PDF of the workflows I run in production, each with what it does, what stays with a person, and the before and after figures. Use it to pick your own first workflow.
Proof from my own production lines
Posts here are planned from search data and drafted by Verqos, the content engine I built for my own sites, working from my own run logs. Anything with pricing, legal or competitor claims comes to me before it publishes, and every figure comes from systems I run.
The content engine behind both
The same pipeline, productised. It plans, drafts, checks and publishes, with a human review queue for anything legal, pricing or competitive.
Hundreds of hours saved
Scheduled agent tasks
My weekly LinkedIn and SEO reviews run as scheduled agent tasks with written quality guardrails. I read the output, not the raw data.
24/7 automated agents running
It's a weekly email with one workflow from my own businesses: what it does, its numbers and what broke that week. It's short, it's written from real run logs, and you can unsubscribe in one click.
I do, with help from Verqos, the content engine I built for my own sites. It plans topics from search data and drafts each post from my notes and run logs, and it checks the draft against written rules. Anything with pricing, legal or competitor claims comes to me before it publishes.
New articles go up on Wednesdays and Fridays, plus the Agent Log on the last Wednesday of each month. The Friday agent log email goes out every week.
From the systems I run: Campercation, TrustedVillas, Verqos and the scheduled agent tasks behind my own business. Any industry figure is linked to its source. If there's no reliable number, the post says so rather than inventing one.
Yes, that's the point of writing them up. If you'd rather have one built, fixed-price and handed over to your team, book a 20-minute call.
Real numbers from systems in production, every Friday.