Research · 6 min read
Many AI seats go quiet after the first few weeks
Use of paid AI seats peaks soon after launch, then settles lower. In a large Copilot trial, a worker's employer predicted their use better than their own work habits did.
Company
The AI Department That Stays
We work with firms of 10 to 500 people. After the build and the training, we stay on to keep things working, and your team can ask us about anything new in AI.
Who We Are
McCarthy's AI is a Canadian company based in Ontario, and we're building an in-person team in Toronto. We do AI consulting and training for firms in the US, Canada and the UK, on site or live online. Alongside that work we run an AI data lab, which helps clients license a de-identified record of their business processes to frontier labs. The company is funded by the Federal Economic Development Agency for Southern Ontario.
Why We Stay After Launch
- 01
The consultant moves on
A project ends and nobody owns what was built. When a tool changes or a workflow breaks, people go back to the old way of working.
- 02
Seats go unused
Use of paid AI seats tends to peak soon after launch, then settle lower. Without someone checking in, the bill stays the same while use drops.
- 03
Nobody to ask about what's new
New models and features keep arriving. Staff try them on their own, often on personal accounts, and nobody can say what's safe to use with client files.
How We Work
Every client starts with a free 30-minute call, then the same four steps. Audit to trained team takes 6 to 8 weeks.
- 1
Audit (2 weeks)
Interviews with leaders and the people doing the work, and a review of your tools, data and AI spend. You get ranked opportunities with the hours and dollars each saves, and a 90-day plan.
- 2
Build (3 to 4 weeks)
We connect the tools you already pay for to an AI agent and build workflows that draft work for a person to approve. The smallest high-value item ships first.
- 3
Train (1 week)
AI 101 for the whole team, then sessions for each role on the workflows we built. Every workflow gets a named owner who's trained to run it.
- 4
Support (ongoing)
A check-in at 30 days compares hours saved with the audit baseline. After that, a monthly retainer covers fixes, updates as tools change and light new work.
What You Can Count On
- You own everything we build, and it runs in your accounts
- A named owner on your team for every workflow, trained to run it
- Written docs for each build
- A 30-day check of adoption and hours saved
- Fixes and updates as tools change, on a monthly retainer
- One place to ask about anything new in AI
- Help licensing a de-identified record of your business processes, if you want it
Common Questions
Still deciding? On a free 30-minute call we'll tell you what fits.
What is the AI data lab?
It's the part of McCarthy's that helps clients license a de-identified record of their business processes to frontier labs and businesses training AI agents. We work with data partners and act for the client on consent, scope and the terms. Client details come out of every copy. Over time we plan to work with labs directly.
What happens after launch?
A monthly retainer covers fixes, updates as tools change and light new work. When something new in AI comes up, your team can ask us whether it's worth trying and how it would fit what we built.
What size of firm do you work with?
Firms of 10 to 500 people, though headcount is a rough guide. What matters more is a leader who wants this and a team with time to learn what we build.
Related Research
Training · 6 min read
Training beats tools for AI adoption
In a Danish study of about 25,000 workers, more people used AI chatbots where employers trained them than where employers only supplied a tool. Encouragement went further than either, and all three together went furthest.
Free 30-minute call
Start With a Free Call
On a free 30-minute call, you'll see one of our prompts run in your own AI tool and leave with a first step and its price.