← Back to Blog

The 2-Person Company Playbook: Running 111 Workflows With Zero Employees

By Ethan Wilmoth March 15, 2026 16 min read

The Economics of Two

MEWR is run by two people: Ethan (CEO, operations, strategy) and Sarah (COO, product, customer relations). We generate $146K in monthly recurring revenue. We run 111 workflows across three news agencies, five revenue tools, and dozens of content automation pipelines. We have zero employees, zero contractors, and zero support staff.

This doesn't scale if you're hiring humans. But with AI and automation, two people is enough. This is the playbook.

The Thesis: Traditional companies scale linearly (more revenue = more people). Automation companies scale logarithmically (more revenue = same people). The math is different. The operations are different. The mindset is different.

The Core Architecture: What AI Does

We don't have a content team because Claude generates content. We don't have a business intelligence team because workflows analyze data. We don't have a customer support team because our products are self-serve. AI fills these roles.

The Three Agent Tiers

Tier Role Examples Human Input
Tier 1: Execution Autonomous workflows that run without intervention RSS feed monitoring, content summarization, file processing <1% (scheduled, no decision required)
Tier 2: Analysis Workflows that score, evaluate, or flag content for review Quality scoring, bias detection, opportunity identification 5% (human reviews flagged items)
Tier 3: Strategic Workflows that require human decision-making Customer onboarding, pricing changes, product launches 30% (humans make the call)

80% of our workflows are Tier 1 (execution). They run automatically. Ethan checks logs weekly, looking for anomalies. Sarah reviews one dashboard (5 minutes/day) to see if anything broke. The rest handles itself.

15% are Tier 2 (analysis). The workflow flags something, sends it to Slack, and one of us reviews. Average time: 3 minutes per flag. Frequency: 10–15 flags per day. Total time: 45 minutes/day split between both of us.

5% are Tier 3 (strategic). These require actual human judgment. How do we price a new product? Should we launch another news agency? Do we hire or keep automating? We spend 8–10 hours per week on these decisions, mostly asynchronous (we message back and forth).

The Operational Rhythm: How Time Is Allocated

Ethan's Weekly Schedule (50 hours/week)

Sarah's Weekly Schedule (40 hours/week)

Total human input: 90 hours/week for a $146K/month business. A traditional company of this size would have 10–15 people.

The Leverage Ratio

$146,000 monthly revenue ÷ 90 weekly human hours = $1,622 per human hour. Compare to a traditional startup: 10 people × 40 hours = 400 human hours, $365 per human hour. MEWR's leverage is 4.4x.

Operational Principles: How to Stay Lean

Principle 1: Automate Ruthlessly

Before hiring for a task, ask: can a workflow do it? Can Zapier/n8n handle it? Can a script run it? Hiring should be the last resort, not the first.

Real example: We were manually posting to LinkedIn. 30 minutes/day. Sarah wrote an n8n workflow. Now Claude generates captions, the workflow posts to LinkedIn, and we just approve/edit. Time spent: 5 minutes/day. Automation savings: 25 minutes/day.

Principle 2: Default to Self-Serve

Every product we build is self-serve. No onboarding calls (except for high-ticket items). No customer hand-holding. If a user can't figure it out, we improve the UI, not add support staff.

This sounds harsh, but it's liberating. Sarah spends 30 minutes/week on support questions. A traditional SaaS company would hire a support person at this volume.

Principle 3: Build Systems, Not Tasks

We don't do one-off things. We build systems that handle entire categories of work. One person doesn't handle "content creation." A system does. One person doesn't monitor news. 24 agents do, continuously.

This requires 3–6 weeks of upfront investment to build a system (e.g., the RSS-to-Slack-to-analysis pipeline). But then the system runs forever, no additional labor required.

Principle 4: Measure Everything

We track: workflow execution count, success rate, cost per execution, revenue per customer, churn, support tickets, time spent on manual tasks. We review this dashboard weekly. If a number trends in the wrong direction, we investigate immediately.

The dashboard is an n8n workflow that pulls from Stripe, n8n, Slack, and our internal systems. Takes 1 hour to set up, saves us 5 hours/week in manual reporting.

Principle 5: Communicate Asynchronously

We're not in real-time communication. Ethan and Sarah message, they don't call. We don't have daily standups. We document decisions in Notion. This saves 10+ hours/week and makes everyone more thoughtful (you write better when asynchronous than when reacting to verbal discussion).

The Delegation Model: How AI Agents Replace Humans

Our rule: if a task is repeatable and well-defined, delegate it to an agent (AI workflow).

Content Generation

Human role: Define the direction, set constraints, review 20% of outputs for quality.

AI role: Generate 80% of content. Headlines, blog posts, social captions, newsletter edits, video descriptions. Claude does this. Takes 2 hours/week human oversight.

Data Analysis

Human role: Ask the question, interpret the answer, make the decision.

AI role: Pull the data, run the analysis, generate a report. An n8n workflow pulls from databases, feeds to Claude, generates a markdown report. Takes 0 hours/week (totally automated, reports land in Slack daily).

Customer Onboarding

Human role: Review onboarding funnels, fix friction points, adjust messaging.

AI role: Respond to FAQs, send welcome sequences, guide users through signup. Beehiiv automation + custom n8n workflow handles this. Takes 1 hour/week to monitor and tweak.

Bug Detection

Human role: Receive alerts, triage, decide on fix or workaround.

AI role: Monitor workflows, log errors, flag anomalies. Our monitoring system (n8n + custom checks) alerts Sarah when something breaks. Takes 2 hours/week to investigate and fix.

Pattern: Humans stay in the decision-making layer. AI handles execution and analysis. This creates leverage.

Tools That Enable This (The Tech Stack)

Total monthly infrastructure cost: ~$3,827/month (mostly Stripe fees). That's 2.6% of revenue. Scales automatically.

The Cost Advantage: A traditional 10-person startup spends $300K–$500K/year on salaries alone. MEWR spends $46K/year on infrastructure. That's a 10–100x cost advantage, which compounds in market capability.

The Hard Parts: What Actually Requires Humans

Not everything can be automated. These are the human bottlenecks we haven't solved.

Strategic Decisions

Do we launch a new agency? What should we charge? Should we hire or stay lean? These require human judgment, taste, and risk tolerance. AI can gather data, but we make the call.

Customer Relationships

High-value customers need relationship management. Sarah handles this. It's 3–4 hours/week but pays for itself (these customers have higher retention, higher LTV).

Creative Direction

AI generates content, but humans decide: is this voice right for our brand? Does this feel authentic? Should we publish this or pull it? Takes 2–3 hours/week to review and edit.

Bug Investigation

Workflows sometimes fail in weird ways. When they do, Ethan investigates: Why did this webhook timeout? Why did Claude hallucinate? This is detective work. Takes 1–2 hours/week.

Total "irreducible" human work: 10–15 hours/week. Everything else is automation.

Scaling Without Hiring: The Next Level

What happens when revenue doubles to $300K/month? We don't hire. We automate more.

We could reach $1M/month revenue with the same two people. It requires building new systems and refining existing ones, but not hiring staff.

The Secret Sauce

It's not genius. It's not luck. It's ruthless prioritization of automation, a deep understanding of your own workflows, and the discipline to say "no" to tasks that don't scale. Build systems that run without you. Trust them. Improve them. Scale them. That's it.

The Operational Mindset

Running a 2-person company isn't about working harder. It's about working differently.

Traditional mindset: I need help with X. Hire someone. Manage them. Onboard them. Pay them.

Automation mindset: I need help with X. Can a workflow do it? Can an existing tool do it? Can I build a script? Only hire if all three are no. (Spoiler: they rarely are.)

Every hire is a decision to stop automating. We've made that decision zero times outside co-founders. Not because we're ascetic, but because we haven't hit a problem automation couldn't solve.

You don't need our tech stack. You need this mindset. Start there.

By Ethan Wilmoth, MEWR Creative Enterprises LLC
Two people, 111 workflows, $146K/month revenue, zero employees. The operational framework for automation-first businesses.

← Back to Blog