Please please please run this 15-minute huddle if you're serious about fixing delivery: 1. Yesterday's numbers Every client marked green, yellow or red. Skip anything green. 2. Blockers "What's blocking you?" on every yellow and red. 3. Root causes Every red metric gets a written root cause before the call ends. We ran it every morning at WKND, one pod at a time. Run it for two weeks and watch problems get caught on Tuesday instead of exploding on Friday.

Every part of the 8-figure agency I was COO of ran off a written playbook. I'm giving away the EXACT blueprint we used... It's quite literally the playbook you'd hand a new hire on day one. And this is the biggest bottleneck with most agencies... The founder builds the whole thing in their head... And it falls over the SECOND they get sick or hire someone new. - client fires land on the founder's phone - delivery slips and nobody flags it - new hires shadow for weeks - margins look fine until someone checks That said... I'd rather give you the written system WE used to get out of chaos. Inside: - the Pod Model, so every client has an owner who isn't you - team-led SOPs that actually get followed - 5 leading metrics that show a fire before it starts - profit-sharing that pays for the right behavior Want it? Comment "SCALE" and I'll send it over. Must be following for the DM.

11 places to get margin back for agency owners: 1. Legacy clients never repriced (20+) 2. Scope creep (one client at 126% of hours) 3. Overage clauses nobody enforces ($225/hr) 4. Rush jobs priced like normal work 5. Clients off autopay (the founder becomes collections) 6. Volume discounts for your 2 biggest clients 7. Hours that never hit a client account (28%) 8. Scoping that only counts labor x rate 9. SOW work missing from the pricing model (83 hours) 10. Capacity planned at 40 hours a week 11. Revisions caused by incomplete briefs

Ranking every client communication habit in tiers: S TIER - A chase log. Every chase becomes a fix. - A production health slide. On every client call. - A 3-hour Slack promise. Then reply in 1. A TIER - Something the client can see by day 7. - Never 3 business days of silence after the first invoice. B TIER - A client dashboard. Only if it says what's burning. C TIER - "Let me know what you think." Every deliverable becomes an approval request. - NPS surveys. A 6 means they've been unhappy for months. If you want to learn how I set up the S tier, bookmark the article below.

I grew my lead gen agency to 7 figures with Google Ads and then built a DTC brand to $27.6M in a year. I still stop work at 4:30 every day. If I started a new agency tomorrow, here's my calendar: (save asap) 1. Personal time first, not last 2. Top 1-3 priorities in the morning 3. Meetings pushed to the end of the day 4. Slack 3 times a day, never all day 5. My team brings 1 problem, 3 solutions and 1 recommendation 6. Every block color-coded 7. Next week planned before it starts Same order every week, and I'd probably be stricter about it the second time around.

This is the check I'd run before an agency adds another editor and $1,500 more per client. Paste the full article into Claude with one of your own deliverables. Ask it where the days went. Bookmark it for your next hiring conversation.

"If you were COO of my agency, what's the first thing you'd do?" Delete the SOP library. Why? Because most agencies under $3M have the same problem: the founder wrote the SOPs, and nobody reads them. The founder spends a weekend documenting everything, the team checks out, and the process still lives in people's heads. That's a two-week fix at 5 people and a 3-month cleanup at 30. Playbooks fix it. When the team creates the process, it's a playbook, and they'll actually use it. Here's how I set them up in every agency I work with: 1. Only build one if it saves 20-30 minutes per client. 2. Score every idea on urgency and frequency. One agency's list went from 50+ down to 4. 3. The person doing the work writes it. You review. 4. Put the value stream at the top, so everyone sees where the task sits. 5. Number every task, one action each. 6. Add key points under each task. This is where the tribal knowledge goes. 7. Write the process time next to every step. 8. Add visuals for anything easier to show than say. 9. Put the playbook inside the task in your PM tool, where the work happens. 10. Ship one a week. In a year you have 52.

I've built or run 5 businesses since 2014. No one wants to hear this, but every one of them was won or lost on operations. Here's what each stage taught me: 2014 to 2016: a B2B lead gen agency. I learned Google Ads and landing pages from YouTube. We did over 7 figures in revenue in year two. I sold it to my business partner for 7 figures, because running a call center stopped being fun. 2016 to 2019: the lost years. One month off turned into two years. I traveled, earned nothing and blew all of the money. Then came a supplement company and about five other ideas that never took off. 2019 to 2020: a DTC brand. $500K in the last three months of 2019, then $27.6M in 2020. One Christmas a manufacturer lost containers of our orders in transit. My bank account went from $2.5M to $129K in 90 days. 2022: COO at WKND Digital. I walked into over 100 people and no systems. By the end of 2022 we were around 40 people. That same headcount went on to do more revenue than we did in 2022. Late 2024: Agency Operators. $382,428 in year one, at 87% margins, with one person. None of these stages replaced the last one. The lessons stacked. If you take one thing from this: I had the vision every time. Execution is what nearly broke me.

Elon Musk admits he built the Tesla Model 3 backwards: "I literally automated, accelerated, simplified and then deleted." I see agencies doing the same with Claude. One deliverable at a $1.2M agency: 8 hours of work, 23 days on the calendar. Claude saves 4 hours. The 23 days stay. So I wrote out his 5 steps for an agency: > Question it: 3 more editors, or better B-roll search > Delete what doesn't protect quality, speed or margin > Simplify: one QA checklist per deliverable > Speed it up: cut the days work sits waiting > Only then automate Full breakdown is in the article:

There's one question I ask almost every agency owner I work with. It tells me if they're underpriced before I've opened their P&L. "How much does it cost you to deliver $1,000 worth of client work?" Here's how to answer it for your own agency: 1. Pull every client's monthly fee and the hours your team spent on them last month. 2. Cost those hours at loaded labor. Plan capacity at 32.5 hours a week, and price your margins at 24. 3. Divide. A $500K agency owner I asked found it cost him $720 to deliver $1,000. 4. Watch for red flags: Close rate above 40% = you're probably too cheap. Rates set 2-3 years ago = a legacy client dragging the pod down. Gross margin under 50% = you're below the floor. Your biggest account = check it first. One $80K a month owner's two largest clients were his least profitable. 5. Set the floor at 50% gross margin and target 60%. Every client under it goes on a list. 6. Work the list over 60 days: reprice, rescope or sunset. 7. Raise prices for new clients too. One $500K agency went up 40% and lost one client. Most agencies I audit have 10-20% of their clients sitting close to zero or negative gross margin. The owner usually finds them the same way, one question in.

I'm an ex-8-figure agency COO. Here's the exact hiring system I install in agencies. One $1.7M agency cut new-hire ramp from 90 days to about 45 with steps 6 and 10 alone: 1/ Write the scorecard before the job post. Mission, targets, key initiatives, skills, responsibilities, values. No more hiring on vibes. 2/ Write the job post like a sales letter. A $120K team lead role that reads like your accountant wrote it gets nobody good. 3/ Pay at the 75th to 90th percentile. Pay average and you get average. 4/ Ask your team first. Half your hires should come from referrals. Pay the referrer 5% of the hire's year-one salary once they pass 90 days. 5/ You interview last. One $1.5M owner went from 8-10 hours per hire to under 90 minutes once his team ran the first rounds. 6/ Run Week Zero. Logins, tools and a buddy get sorted before day one, so nobody burns their first morning waiting on Meta access. 7/ Load accounts slowly. Zero clients in week one. 8/ Meet daily for the first 2 weeks, 3 times in week 3, then weekly. 9/ Train with Toyota's method: I do, we do, you do. 10/ Spend 2 hours with every new hire on where the agency is going. I did it with every single hire as COO. Skip these and you're back to 30+ hours of hand-holding on every new hire.

A $1.2M agency got its QA person out of daily review within 90 days. Client ads used to go live with typos every week, and the owner's plan was one more review layer. The setup: 1. His QA person's checklist, built into every stage of production 2. Everyone starts with 100% of their work reviewed 3. Hold a 90%+ pass rate for 30 days and review steps down 4. 75%, then 50%, then 10% To copy this: > Write down the checklist your reviewer already uses > Put it inside every stage > Track each person's pass rate for 30 days The checklist is the part you own. AI comes last.

"What does a COO actually do in an AI agency?" > opens the editors' queue before approving a hire > finds B-roll split across 4 folders with no naming > spots 8 edits vs 4 between two editors on one title > sets output targets strategists can hit while the pod catches up > writes one QA checklist per deliverable type > hands it to designers AND reviewers > drops review once someone holds 90%+ for 30 days > maps 23 days of waiting on 8 hours of work > sits with three leads before the AI rollout

I'm an ex-8-figure agency COO. An $80k/month agency tripled its scripts with Claude. Its editors still did 2-3 edits a day. Across the 27 agencies I've worked with ($36M+ in revenue between them), AI breaks the same 7 places: - More scripts, same edit queue - $1,500 more per client for one editor - Mediocre work, just faster - QA automated before a checklist exists - 23 days of waiting on 8 hours of work - A rollout nobody was asked about - AI added before the system was fixed Full breakdown in the article below:

A $2.4M agency owner I work with promoted his best creative strategist to team lead after 3 months, and 47 days later he was back running the team himself. He told me he'd delegated. He'd abdicated. So I told him to rewind and run a 30-60-90 day transition. Works like this: Step 1: Scorecard first (define the measurable criteria for the role before you promote anyone) Step 2: Show the gaps (when longer-tenured people see the objective gaps they'd need to close, the politics go away) Step 3: Days 1-30 (weekly 1-on-1s, plus joint client calls where the new lead runs the meeting and you observe) Step 4: Days 31-60 (you shift to reviewing their decisions and talking through the reasoning) Step 5: Days 61-90 (you only step in for escalations) Step 6: Criteria per phase (nobody moves to the next 30 days until they've hit them) For those 90 days, you're coaching, not doing. That's how you build a leader who can actually hold the role, not someone set up to fail in 47 days.

Just chatted with a $1.2M agency owner He has a huge problem He built his team a new AI system No heads-up or 1-on-1s Strategy roadmaps, content templates, ClickUp task logging His team only finds out tomorrow morning He's presenting it all at once I told him to scrap it "Talk to your three leads first," I said He thought I was crazy "Look, if your leads hear it from you 1-on-1, they'll back it. But if they hear it in a team meeting, they hear 'they're figuring out how to do this without me.'"