Pillar 04 — Operations

Margin doesn't disappear overnight. It leaks.

Operational drag is invisible until it isn't. Every week your team does things the slow way, does the same thing twice, or waits for an approval that shouldn't exist — that's margin leaving the building.

Operations is the least glamorous of the four pillars, and it's the one that pays the most. The businesses I've worked with that struggled with growth almost always had an ops problem masquerading as a revenue problem. They weren't undercutting themselves on price — they were burning their margin on rework, hand-offs, and decisions that took three times as long as they should.

The framework I use comes from the Theory of Constraints: every system has exactly one bottleneck at a time, and improving anything other than the bottleneck is wasted effort. Most businesses work on ten things at once and wonder why nothing improves. The session starts with diagnosis — finding the actual constraint, not the loudest symptom.

From there, we match the bottleneck to one of five patterns that cover the vast majority of SMB operational problems. Each pattern has a proven fix that can be implemented in two to four weeks without hiring anyone new or buying expensive software. We design the minimum viable version in the room.

AI has changed what's possible here. Automations that used to require a developer can be wired up in an afternoon. Documentation that would take days to write can be scaffolded in an hour. The ceiling on what a small team can operationally manage has risen significantly — if you know the patterns.

The patterns

Five bottleneck patterns. One of them is yours right now.

P1

The Intake Bottleneck

Symptom

Work piles up at the point where new requests enter the system. Team is always behind before they start.

Root cause

No defined intake process. Everything comes in differently — email, Slack, phone, walk-in — and gets triaged by whoever is available.

The fix

Single intake channel, standard fields, automatic routing. Takes one afternoon to build. Eliminates the chaos that masquerades as a staffing problem.

P2

The Tribal Knowledge Problem

Symptom

One person knows how to do something critical. When they're out, things break. When they leave, institutional knowledge leaves with them.

Root cause

No documentation culture. SOPs aren't written because writing them feels like overhead. They're never written because it never feels like the right time.

The fix

Loom-first documentation: record it once, write the outline after, publish to a searchable wiki. 20 minutes to capture what took years to learn.

P3

The Meeting Tax

Symptom

Calendars are full, decisions are slow, and nobody can point to why. High-performers start leaving because they can't do deep work.

Root cause

No distinction between synchronous communication (decisions, creative work) and asynchronous communication (updates, status, approvals).

The fix

A communication protocol: what belongs in async (written, with a decision deadline), what belongs in sync (genuinely blocked, genuinely time-sensitive). Most meetings become documents.

P4

The Approval Gridlock

Symptom

Things sit waiting for sign-off. The person doing the signing becomes the bottleneck. The team stops trying to move fast because nothing moves fast.

Root cause

Unclear authority. Nobody knows what they're empowered to decide. Default behavior is to escalate everything, which trains the team to wait.

The fix

A decision matrix: what can be decided by role without approval, what needs a single approver, what needs consensus. Published and followed.

P5

The Metric Blind Spot

Symptom

Nobody knows if last month was good or bad until well into this month. Decisions get made on gut feel because the data is stale or missing.

Root cause

Metrics exist in siloed tools — accounting software, CRM, spreadsheets — and nobody has synthesized them into a weekly picture of business health.

The fix

A single weekly number dashboard: 5–7 metrics, updated by one person, reviewed every Monday. Not a BI platform — a shared doc with formulas.

The rollout

How to implement operational changes without breaking what's working.

Diagnose
Week 1

Find the real bottleneck. Interview the team. Map the workflow. Look for where work piles up, where errors happen, and where decisions stall. Most businesses identify the wrong bottleneck because they confuse symptoms with causes.

Design
Week 1–2

Match the bottleneck to the right pattern. Design the minimum viable fix — not the perfect system. The goal is an improvement that can be live in 2 weeks, not a 6-month ERP implementation.

Pilot
Week 2–3

Run the new process with one team or one project type. Collect feedback. Measure the before/after on the one metric that matters. Adjust before full rollout.

Document
Week 3

Write the SOP. Record the Loom. Not because it will be perfect — because it makes training the next person 10× faster and makes the process improvable by someone other than its inventor.

Roll out
Week 4

Full deployment with a 30-day review date already on the calendar. Communicate the why before the what — people follow processes they understand. They route around processes they don't.

Iterate
Ongoing

The first version is never the best version. Build a lightweight feedback loop: a monthly 30-minute retro with the people who do the work, not the people who designed the process.

The stack

The tools that cover 95% of SMB operational needs.

Tool selection is secondary to process design. A bad process in good software is still a bad process. Build the process first. Pick the tool that fits it.

Process documentation
Notion, Confluence, Slite

Pick any. Consistency matters more than features. A Notion wiki you actually update beats Confluence nobody touches.

Task & project management
Linear, Asana, ClickUp

For most SMBs under 25 people: Linear is the cleanest. For teams already in Microsoft 365: Planner is enough. Don't over-engineer it.

Automation
Make, Zapier, n8n

Any process you do the same way more than 5 times a week is an automation candidate. n8n for self-hosted control, Make for ease of use.

Metrics dashboard
Google Sheets, Notion, Airtable

Start with a shared spreadsheet. Fancy BI tools add complexity before you've established what you actually need to see weekly.

Communication protocol
Slack + Loom or Teams + Stream

The tool is secondary. The protocol is primary. Document what belongs async vs sync and enforce it for 30 days. The culture shifts.

Next session

Bring your biggest bottleneck. We'll diagnose it in the room.

Free, in-person, monthly across Washington with an online option. Every attendee who brings a real ops problem gets time on the floor.