Case study
How No Limit Email went from $4.5K to $12K retainers with the same team
| Before AI | Now | |
|---|---|---|
| Retainer | $4,500/month on average | $12,000/month on new retainers |
| Team per client | Same | Same |
| Briefs written per hour | 1 to 2 | 40 to 50 |
| Time on QA, reporting, strategy, onboarding | Baseline | 60% less |
| AI spend | None | $1,000 to $2,000/month, whole company |
The agency
No Limit Email is an email and retention agency for US ecommerce brands. The team is fully remote, spread across Latin America, Canada, the UK, and South Africa. Each client gets a strategist, a project coordinator, copywriters, and designers.
The problem
Strategy quality depended on who we hired. Most strategists we hired were okay, and some were weak on the fundamentals. The best thinking in the agency lived in a few people's heads, and a new hire took months to get close to it.
Every morning started with a scavenger hunt. Before any real work, each strategist and account manager checked ClickUp for mentions, overdue tasks, and what was due that day. Then Slack for client questions and feedback, Figma for design comments, and Google Docs for comments on briefs and roadmap sheets.
Work fell through the gaps between people. One email goes through a brief, then copy, then design. When a copywriter changed the copy but not the brief, the designer never found out. When the brief changed, the copywriter often didn't hear about it. Clients noticed.
Client feedback didn't stick. A client would tell us something once, and three weeks later someone else on the team would make the same mistake.
Briefs were slow. A strategist could write one or two in an hour.
What we built
In January 2026 we moved the agency onto one shared AI system built on Claude Code. The whole team works inside it. It holds every client's knowledge, our playbooks, and our processes, and we've built 110 custom commands on top of it. These are the ones that did the most work.
Our strategy playbooks, built into the AI
Our retention playbooks, strategy frameworks, and years of client learnings live inside the system. A strategist hired last week works from the same thinking as our most senior people.
Start Work
A team member starts their day by running one command. The AI checks ClickUp, Slack, Figma, and Google Docs for them and hands back a prioritized list of what needs attention.
Briefs and copy
The AI drafts briefs and copy from the client's knowledge base, the roadmap, and our playbooks. Briefs went from 1 or 2 an hour to 40 or 50.
Learning loops
When a client gives feedback on a brief or on copy, it gets saved to that client's record. Every future brief and email for that client already accounts for it, so clients stopped repeating themselves.
Downstream sync
When a brief or copy changes, the AI updates everything that depends on it. Then it tells the team member which task to comment on so the next person knows. Changes don't get lost between strategy, copy, and design anymore.
QA, reporting, and onboarding
Each has its own AI workflow. Together they take 60% less time than they used to.
Teach Me
This command trains strategists on our methods and quizzes them. The team keeps improving without us running training sessions.
Usage tracking
Using AI is part of every team member's KPIs. A monthly report shows who is using which commands and who isn't.
What changed
We raised prices instead of adding clients. An agency doing simpler work at smaller retainers could use the same system to take on about a third more clients per person. We went the other way. Every strategist now works at a senior level, and the AI helped us build more advanced services like deeper analytics, advanced reporting, and holdout testing. That's what let us sell a bigger retainer.
New retainers went from about $4,500 to $12,000 a month, with the same size team on each account.
We make fewer mistakes. Learning loops and downstream sync fixed the two things that frustrated clients most: the same error happening twice, and a change that never reached the designer.
AI costs stayed small. The whole company spends $1,000 to $2,000 a month on AI subscriptions and API usage. The retainer increase on a single client is about $7,500 a month.
How much the team uses it
From January to September 2026:
- About 16,000 pieces of work saved in the shared system. Activity went from about 20 a month in January to between 3,000 and 4,400 a month from May on.
- More than 2,000 briefs created.
- About 800 client reports produced.
- 110 custom commands built.
- About 15 team members working in it daily, plus automated agents running jobs in the background.
What this means for your agency
I built this system as an agency owner, without hiring engineers. If your agency has uneven strategist quality, mornings lost to checking tools, handoff mistakes, or client feedback that gets forgotten, the same approach will work for you.
We now install this system in other ecommerce agencies with 10 or more people. It starts with an AI Ops Diagnostic: a scorecard of your agency, the workflows costing you the most hours, and a plan. Then a two to four week sprint builds it and trains your team. Agencies can keep us on a retainer to keep it growing.