The Operating System Show

Matt Dixon

Everything your business knows walks out at 5pm. You rent it back every morning. And every time someone leaves, you lose a piece of how the place runs. That's the problem The Operating System Show exists to solve. I'm Matt Dixon, fractional CTO. I talk to the founders, COOs, and executives who are figuring out how to get their companies to run without every answer living in one person's head. Every week: what broke, what they rebuilt, and where the real leverage was. New episodes every week. Subscribe so you don't miss one.

Episodes

  1. 2 days ago

    Your Team Says They Are at 140% Capacity | Kati Peterman on What a Fractional Operator Actually Fixes

    Every founder Kati Peterman meets says a version of the same sentence. I'm just not there yet. Not there yet for a COO. Not there yet for a real operator. And she has come to believe that sentence almost never means what it sounds like it means. It is rarely about the money. It is that a visionary founder cannot picture what that person would actually do all day, so a big salary looks like a bet on a job description they cannot see. Her answer is to go find the money first. Outstanding collections, billables, how clients pay and when, merchant services. Unglamorous operational leakage sitting there in almost every company. Fix that in the first weeks and the fractional executive has paid for herself, and the founder stops arguing about whether operations deserves funding. ## In this episode - What fractional actually means now that the word has gone hot, and what it buys a company not ready for a full time hire - I'm just not there yet, and why that sentence is usually about understanding rather than budget - The quick win that funds the role, starting with money the business has already earned and not collected - A better question than calling a founder a control freak: where does my founder have fear right now - Placing the right person versus installing the system, and why the operator has to actually enjoy the work - Education as the thing that makes a system stick after the operator leaves - Where she refuses to let AI go, including refunds and cancellations, and why the sales conversation is burning leads - Teams running at 120 to 140% capacity, and what it takes to get back to 70 - Who should not hire a fractional executive ## Timestamps 00:00 A show for founders who are done being the bottleneck 01:10 What fractional actually means, and why the word got hot 03:11 The first COO offer, and the gap she kept seeing 05:04 I'm just not there yet, and what founders really mean by it 08:05 Outstanding collections, billables, and money already on the table 09:10 Founders get a bad rap as control freaks 09:54 The better question: where does my founder have fear right now 14:41 Why the operator has to actually enjoy the work 16:59 Education is true empowerment, and what makes a system stick 23:12 Teams at 120 to 140% capacity, and getting back to 70 24:35 Zero AI in refunds and cancellations 26:08 The thing that is not up for debate: response time 28:50 Who should not hire a fractional executive 30:01 Why franchisee owners became the favorite client 31:40 Where to find Kati ## About Kati Peterman Kati Peterman came up as a chief of staff and integrator inside seven and eight figure companies, then founded FRX, a national bench of fractional executives placed into small and mid sized businesses. She is also COO of Empire Operating System. FRX: https://yourfrx.com LinkedIn: https://www.linkedin.com/in/katipeterman/ If your team is at 140% and hiring is not on the table, that is the conversation: https://crm.frontrangesystems.com/discovery

  2. 4 days ago

    Saying No Is Twice as Safe as Saying Yes | JL Heather on the Can't Culture Killing Your Best Ideas

    JL Heather has a name for the thing that kills a good idea before anyone notices it died. He calls it can't culture. Not a person saying no, a system saying it for them. Twelve approvals that take six months. A budget objection, a pay grade objection, a resourcing objection, all of them sounding perfectly reasonable on the way out of someone's mouth. Nobody ever makes the decision to kill the project. It just never survives the trip. Underneath it is omission bias. Put your name on a ten million dollar bet and the downside is career limiting. Say nothing and your name is on nothing, even when a competitor ships the same idea three months later and makes a fortune. Speaking up costs about twice what staying quiet costs, so people stay quiet, and it feels prudent the whole time. ## In this episode - Getting a room full of executives to commit to one thing, and why scheduling is the first real obstacle - Compressing what normally takes three to eighteen months into a five day design sprint - Building runs at light speed now, learning did not keep up, and that gap is where the risk lives - The three ways a project dies, ranked from best to worst - Can't culture, and why it is human nature rather than bad people - Omission bias, demonstrated live on Matt with a ten million dollar hypothetical - Ask AI for an Italian plumber and you get Mario, why models pull everything back to the middle - Where AI genuinely earns its place, and the one job it should never be given - Four to five people, one room, three months, and failure criteria that count as a win ## Timestamps 00:00 A show about removing the bottleneck in your business 01:04 Getting executives to commit to one thing 03:37 Three to eighteen months compressed into five days 04:56 What kills a project that never ships 06:38 The worst case, nobody realizes what happened 08:54 Is can't culture bad, or just human 09:00 Omission bias, and the CEO of MattDixon.com 10:12 Saying no puts your name on nothing 13:34 Ask AI for an Italian plumber and you get Mario 14:33 Where AI earns its place 17:47 What AI is not, and it is not an innovator 23:26 What happens when the quarter goes sideways 25:19 Hit your failure criteria and you have succeeded 26:36 The tipping point that means you need outside eyes 29:36 Where to find JL ## About JL Heather JL Heather is managing partner and principal executive coach at Centered, where the Breakthrough Lab method takes teams from a North Star sprint through design sprints and into three to six months of coaching on the thing that actually ships. Twenty years inside organizations, working mostly in the digital space. He is writing a book on can't culture and hosts the Breakthrough Innovation podcast. Centered: https://centered.work JL's site: https://jlheather.com If ideas keep dying in your company and nobody can point at who killed them, that is the conversation: https://crm.frontrangesystems.com/discovery

  3. 3 Sept

    73% Faster With the Same Team | Steve Schroeder on the Invisible Factory Inside Your Business

    Steve Schroeder has spent almost twenty years finding the delays nobody can see. He calls it the invisible factory. In a plant you can walk the line and watch the widgets pile up at the slow station. In a digital company the same pile ups exist, they are just invisible, sitting in approval queues and handoffs nobody has ever drawn on a wall. His first value stream mapping event ran inside a large telecom during a stalled Agile transformation. They mapped the billing department onto three walls of stickies and cut lead time on new features by 73%. The executive who walked the walls at the end had no idea any of it was happening inside his own company. ## In this episode - Why a company doing all the right frameworks can still be stuck, and what the framework cannot reach - The invisible factory, and why digital work hides its own bottlenecks - Three walls of stickies, a stunned executive, and a 73% cut in lead time - What changes when AI joins value stream mapping, including rerunning twelve year old data and finding half a dozen more improvements - Why a company full of individual AI agents can still be slow end to end - The two roles that decide whether any of it survives: the value stream sponsor and the value stream manager - What separates the companies that compound from the ones whose map goes in a drawer - Utilization is the wrong scoreboard, and Messi walks most of the game ## Timestamps 00:00 A show for founders who are done being the bottleneck 01:07 What a framework cannot fix in a stuck company 02:28 A three year Agile transformation that was not working 04:59 Three walls of stickies and a 73% cut in lead time 07:12 What AI adds to value stream mapping 10:09 Why a company full of agents can still be slow 12:24 Flow Killers and Flow Intelligence, who each book is for 17:57 The moment a client's business actually changes 20:58 What survives ninety days and what goes in a drawer 24:08 The sponsor and the value stream manager 29:32 Your developers are 99% utilized and it means nothing 31:00 Who he helps, and the industry he refuses to touch 35:51 Where to find Steve ## About Steve Steve Schroeder is the founder of EliteFlow Consulting and Solara Analytics. He pioneered digital value stream mapping, bringing a discipline built for factory floors into software organizations where the work is invisible. He wrote Flow Killers, which catalogs the fifty most common things that quietly slow a technology organization, and Flow Intelligence, on marrying value stream mapping to AI. EliteFlow Consulting: https://eliteflowconsulting.com Solara Analytics: https://solaraanalytics.ai If your own business has a handoff you could not draw on a wall if someone asked, that is the conversation: https://crm.frontrangesystems.com/discovery

  4. 1 Sept

    The AI Flagged Your Best Engineer for Cuts: Alisa Reznik on Why People Decisions Stay Human

    A tracking system did exactly what it was built to do. It showed one senior engineer closing seventy percent fewer tickets than everyone around him, and the CEO put him on the list of people to cut. Alisa Reznik made him go have the conversation first. The engineer was spending his days unblocking the rest of the team, which is why their numbers looked so good and his looked so bad. Alisa spent nearly two decades managing engineers, worked with more than fifty tech companies, and ran operations as a COO across offices in three countries before leaving to build her own practice, Leadify. She uses AI all day and she is clear about where it does not belong. ## In this episode The one category of work she refuses to automate, and the rule she uses to decide. The ticket dashboard that recommended firing the most valuable person on the team. Why "what is the best AI tool for my business" is the wrong question, and what to ask instead. The CTO who prepared for a hard feedback conversation, then sent the notes over Slack. Ten hours a week, and the ordinary CEO work that gets those hours back. Her test for anything you are thinking of handing over: how big is the mistake you can afford. The Friday deployment that wiped a client's CRM on Monday morning. What young founders get wrong when they treat Claude as the person who knows. ## Timestamps 00:00 What this show is, and today's guest 01:10 Why she picked "do not delegate everything to AI" 01:57 Fifty plus tech companies, and the thing that breaks the same way everywhere 04:03 Remote versus onsite, and how she runs engineering teams either way 05:41 The first process she refuses to fully automate 07:42 The tracking system that flagged the wrong engineer for cuts 10:46 AI does not save you, it amplifies what is already there 11:02 "I feel redundant, AI can do what I do" 12:48 The wrong question every engineer is asking right now 15:48 Where AI actually worked, and the ten hours a week it gave back 19:24 Moving faster versus avoiding a hard call 20:08 A better question than "what tool should I buy" 22:36 What young founders are still missing 25:10 The feedback conversation that got sent over Slack 29:00 The four zones, and where AI belongs in each 30:05 The rule: how big is the error you can afford 32:44 The Friday deploy that erased a client's data 36:30 Fixing the system and the person running it 37:36 Where to find Alisa ## About Alisa Reznik Fractional COO and executive advisor, and the founder of Leadify. Nearly twenty years managing engineers, more than fifty tech companies, and operations across three countries. She mentors at Founder Institute and writes about operations and leadership every week. Find Alisa on LinkedIn. Listen: https://feeds.transistor.fm/the-operating-system-show Be a guest: https://crm.frontrangesystems.com/discovery

  5. 27 Aug

    85% of Your Calls Don't Need a Human: Timothy Bramlett on Building an AI Receptionist

    Timothy Bramlett bootstrapped an AI receptionist by sitting in a dental office day after day and cloning the tool they already hated paying for. He walked away from HP and Capital One, learned the hard way that building first is the wrong order, and now runs Zinng, an agent that answers the calls a small business misses. He also tells the story of one after hours call that changed how he thinks about what an agent can handle. ## In this episode - Why building the product first is the wrong order, and what he does instead - How to start when you cannot find anyone to talk to: clone an existing tool and niche down - Narrow and deep versus wide and shallow, and why complexity kills small products - Why 85% of calls to a small business never needed a person - Speed to lead in a dental office, and what happens after four rings - The internal tooling he built with Claude Code, and why his engineers still do not trust AI - How AI moved the bottleneck from writing code to testing it - Running several agents at once, and where that breaks down - Why "good enough" beats a better sounding voice - Telling every caller it is AI, up front, on purpose - The after hours call he says changed his mind about what an agent can handle ## Timestamps 00:00 Who Timothy is 01:23 Why he was never any good at being an employee 02:50 The first thing he built, and why it went nowhere 03:42 "That is absolutely not my philosophy" 05:39 Build the channel before you build the product 07:45 When you cannot talk to customers, clone and niche down 10:26 Narrow and deep versus wide and shallow 12:34 Why 85% of small business calls do not need a human 14:34 What it is actually worth to the owner 15:31 Speed to lead, and the four ring handoff 17:30 The internal tooling he built with Claude Code 17:59 Why his engineers use AI and still do not trust it 19:20 AI moved the bottleneck from writing to testing 21:00 The anti AI crowd and who gets left behind 22:04 Can you hand AI a whole job yet 28:00 Running several agents at once 29:14 Local models as the backup plan 31:23 What is next for Zinng 32:26 Why good enough beats better sounding 33:38 Tell them it is AI, up front 35:00 The after hours call that changed his mind 36:31 Where to find Timothy ## About Timothy Timothy Bramlett is a software engineer turned bootstrapped founder. He spent years at HP and Capital One before leaving to build his own products. He co-founded Stammer, a white label AI platform for agencies, and now builds Zinng, an AI phone agent that answers every call a business misses, flags the emergencies, and works around the clock. Zinng: https://zinng.ai Instagram: https://www.instagram.com/timothybramlett/ Stammer: https://stammer.ai Listen: https://feeds.transistor.fm/the-operating-system-show Be a guest: https://crm.frontrangesystems.com/discovery

  6. 25 Aug

    Irritation Is Operational Data: Gwen Taniguchi on Getting Your Business Ready to Sell

    Gwen Taniguchi is a Certified Exit Planning Advisor and partner at Peek Advisory. Her diagnostic tool is one nobody talks about: the low grade irritation inside a team. "Why do I always have to fix this?" is not venting, it is a map of where the business depends on one person. Revenue hides operational dysfunction, and Gwen follows the irritation to find it. ## In this episode * Why exit planning is really about optionality: sell, hire a CEO, or step up to founder, your choice * The quarter test: surviving two weeks without the owner proves nothing * How founder dependency starts on purpose and grows insidiously * Why SOPs alone fail when one person holds the tribal knowledge * Delegating ownership instead of tasks, and what "done right" has to mean * Where AI belongs in operations and accounting, and where it never should go * What buyers actually pay a higher multiple for in diligence ## Timestamps 00:31 Who Gwen is and what exit planning really means 02:42 Irritation as operational data 05:43 The quarter test, what exit ready operations look like 07:35 How founder dependency starts 11:41 The collections story and exception culture 13:49 Gwen's three part approach: leadership, ownership, execution 14:51 Tools: ClickUp and AI agents 16:40 Should AI replace whole departments 20:58 Delegating ownership, not tasks 25:17 What buyers pay a higher multiple for 26:34 When to start exit planning 27:48 What is next at Peek Advisory ## Find Gwen Peek Advisory: https://www.peekadvisory.com LinkedIn: https://www.linkedin.com/in/gwentaniguchi/ ## The Operating System Show Listen: https://feeds.transistor.fm/the-operating-system-show Be a guest: https://crm.frontrangesystems.com/discovery

  7. 18 Aug

    They Don't Want to Buy Your Job: Peter Boolkah on What Founder Dependency Costs at Exit

    Peter Boolkah spent fifteen years inside McDonald's, which is about the most ruthlessly systematized business on the planet, and then spent twenty one years coaching owners out of the trap he watched them build. His line, and the title of his book, is that if the business cannot run without you then you do not own it, it owns you. The part of this one that will sting is the exit math. When a buyer sees that the founder is load bearing, they do not walk away. They discount. ## In this episode * Why a twenty year old could run a three million pound restaurant, and why the same kid could not be trusted with a company car * How founders train the company to depend on them in the years before the first hire * The first employee, and the loop where the owner ends up doing both jobs * What a private equity buyer actually looks at, and why founder dependency turns into a price cut * The split deal, roughly 40 percent up front and the rest tied to numbers you no longer control * Why getting a company ready to sell starts about five years out * Everybody overcomplicates it before they learn to simplify it * The 25 year goal, and why anything shorter keeps the founder in the middle * Do not be lazy with AI, and what lazy actually costs ## Timestamps 00:00 Intro 01:11 Fifteen years inside McDonald's, starting at sixteen 01:50 Taylorism, and a twenty year old running a three million pound restaurant 04:19 You do not own a business, your business owns you 04:54 How the dependency gets trained in from day one 07:32 The first hire, and why it often makes things worse 09:08 Influence instead of control 10:41 Choosing Scaling Up, and leaving after sixteen years 13:17 What a buyer looks at in due diligence 14:46 Why founder dependency becomes a discount 15:56 The 40 percent up front and the earnout behind it 17:40 Getting a business ready to sell, five years out 18:51 Complexity fails, simplicity scales 21:29 Steve Jobs, the first tenure and the second 23:45 Consistency as a strategy, McDonald's and Chick-fil-A 26:22 What he looks at on the first deep dive 28:00 What do you actually want out of your life 30:14 The 25 year goal 31:57 Why he wrote Your Business Sucks 36:42 Do not be lazy with AI 40:05 Where to find Peter ## Peter Boolkah Known as The Transition Guy. Fifteen years at McDonald's, then twenty one years coaching owners across five continents, taking companies through to sale and IPO. His book is Your Business Sucks, on Amazon. He is at boolkah.com ## A note on the book Peter mentions a 99 cent Kindle price near the end. That was a launch offer at the time we recorded, so check Amazon for the current price. ## One more thing Most of what makes a business hard to sell is the stuff living in one person's head. That is the work I do at Front Range Systems: https://crm.frontrangesystems.com/discovery

  8. 11 Aug

    30 Points of Engagement in Six Hours: Omar L. Harris on Culture Operating Systems

    Omar L. Harris ran a 900 person organization in Indonesia and nearly sank it with a handful of bad hires. To dig out, he wrote himself a manual. That manual became a book, the book became eight more, and the whole thing became a framework for the part of a business most founders treat as an afterthought. Matt builds operating systems out of software. Omar builds them out of people. Same word, opposite raw material, and the conversation goes straight at where the two meet. ## In this episode * Why a resume is a marketing document, and what actually predicts how someone shows up * The difference between a personality test and a character test, which is what you choose under pressure * Six hours with a team, fourteen teams, and a 30 point jump in engagement six weeks later * Why culture built with a team sticks and culture handed to a team does not * The three things that send any team back to square one * The person who knows everything, and the incentive quietly paying them to stay that way * Where AI stops being useful and human judgment takes over ## Timestamps 00:00 Intro 00:49 Nine books, and a novel that took twelve years 01:34 Rebuilding a 900 person organization in Indonesia 02:25 Grow With W.H.O.M., eleven years in the making 03:08 Why pedigree and references predict less than behavior 04:29 Character assessment versus personality assessment 05:17 Walking away from the corporate job, twice 06:27 What carries across every market he has worked in 08:26 What a culture operating system actually is 10:43 Forming, storming, norming, performing, and how to speed it up 11:43 The three things that send any team back to forming 14:30 Fourteen teams, six hours each, 30 points of engagement 16:19 What happens after the first engagement ends 17:20 Where AI stops and judgment starts 20:24 Who should not install a culture operating system 22:09 Tribal knowledge and the incentive behind it 24:29 The assessments he is certified on and how he uses them 26:10 Where to find Omar ## Omar L. Harris Founder of Intent Consulting and author of nine books, including Leaderboard: The DNA of High Performing Teams and his latest, Grow With W.H.O.M. Former general manager for a pharmaceutical business in Indonesia, with earlier roles in Brazil and Turkey. W.H.O.M. stands for Work Ethic, Heart, Optimism, Maturity. The book and the free character assessment are both at growwithwhom.com. He is on LinkedIn as Omar L. Harris. ## One more thing If your business runs on what one person happens to remember, that is a systems problem before it is a people problem. That is the work I do at Front Range Systems: https://crm.frontrangesystems.com/discovery

  9. 4 Aug

    From 7 to 150 People: How Sean Griffith Scaled SimpleTexting and Now Ships Features Overnight with AI

    Sean Griffith joined SimpleTexting at seven people and helped grow it to about 150 in under four years, through an acquisition, without a dedicated sales team for most of that run. He is now CEO and cofounder of Truffle, a candidate screening platform for small and mid sized businesses. We cover what actually made that scale work, why EOS was the unlock at 50 people, and the overnight release pipeline his team runs now that would have taken three or four people a month before. ## In this episode * Hiring conservatively and putting the cash into SEO instead of ad spend * Settling on an operating model at 40 to 50 people, then running it loosely on purpose * The overnight pipeline: product catalog, help center articles, customer email, and a feature video, all queued for approval in the morning * One front end developer turning a product brief into a shipped feature in a couple of days * Why Truffle surfaces signal and never makes the hire or pass decision * Assessments built so an AI assisted candidate cannot game them * Why they deliberately did not build an AI interviewer * The real cost of a founder reading resumes at 9 PM ## Timestamps * 00:00 Meet Sean Griffith, CEO and cofounder of Truffle * 01:17 The problem that started it: 2,000 applications per role * 03:18 What it actually took to scale SimpleTexting * 04:50 Choosing an operating model at 40 to 50 people * 06:06 Scaling pre AI versus scaling with AI * 07:21 The overnight feature release pipeline * 09:00 Values, L10s, rocks, right person right seat * 11:41 Running EOS your own way instead of by the book * 12:45 What Truffle does * 15:14 Signal versus decision, keeping the human in the loop * 16:42 Where the line between human judgment and AI sits today * 18:08 Who Truffle fits and who it does not * 19:17 The real cost of screening resumes yourself * 22:29 One developer delivering a full feature from a brief * 25:33 What is next for Truffle * 28:08 Where to find Sean ## Links * Truffle: https://hiretruffle.com (free trial) * Sean: sean@hiretruffle.com Want your own business running without you in every decision? https://crm.frontrangesystems.com/discovery

  10. 28 Jul

    Why 95% of AI Pilots Fail and How to Build a Business You Can Sell, with Ted Wolf

    Ted Wolf started at IBM in the 1970s, built an IT services company to 650 people with his brother, and sold it to Iron Mountain in 2018. Now he runs GuideWise, helping founders put agentic AI into the machinery of their business without the hype. He also wrote a book, The Intelligent Business Equation, a formula for turning a people dependent company into one you can actually sell. In this conversation Ted and Matt get into what it really feels like to scale to 650 people (and scale back down through four recessions), why most AI pilots fail, and the five moves that make a business transferable at a high multiple. ## In this episode - What scaling to 650 people and then reverse scaling through four recessions actually felt like - Why AI has to be grounded in a business objective that hits the bottom line, not writing better emails or prettier slides - The reason roughly 95% of AI pilots fail, and what the winners do differently - How AI makes a business simpler instead of more complex by killing the daily fires - The five closes that turn a people dependent business into one you can sell, refinance, or pass on - Where AI does not belong, and why character and guardrails matter more as it spreads ## Timestamps - 00:00 Ted's background: IBM in the 70s, 650 people, sold to Iron Mountain - 01:15 What scaling to 650 people actually felt like - 03:41 The three things to learn: technology, how a business runs, how you run as a person - 04:59 Rolling out new tech and getting real adoption - 07:23 Start small, lower the risk, land the beach - 11:17 Why 95% of AI pilots fail - 13:06 How AI simplifies a business and kills the firefighting - 16:02 The 90 day AI boot camp and the flywheel of learning - 20:01 The book: The Intelligent Business Equation - 20:55 The five closes that make your business sellable - 25:01 Where AI should stay out, character and guardrails - 26:32 Critical thinking and human connection in an AI world - 28:17 What is next for GuideWise: building a community - 29:38 High tech, high touch ## Ted Wolf - Book: The Intelligent Business Equation - Website: guidewise.ai - Email: ted@guidewise.ai - LinkedIn: Ted Wolf If you want to build these systems into your own business, book a call: https://crm.frontrangesystems.com/discovery

  11. 21 Jul

    How to Build a Business That Runs Without You, with David Hori

    David Hori is the founder of Top Line Operators. He buys businesses that already work and makes them work better, and along the way he has become obsessed with one question: does this business need its owner to survive? We get into what he looks for before he buys, how he untangles an owner running five companies under one roof, and why AI multiplies whatever is already there instead of fixing what is broken. ## In this episode - The criteria David uses before he will buy a business, and why he never touches turnarounds - Why "you are the business" is the most expensive sentence an owner can hear - The durability diagnostic: four pillars that decide whether your business is worth buying - Owner dependency, concentration risk, and the tribal knowledge locked in one person's head - Why AI is an amplifier, not a magic pill, and where he actually uses it - The fear that stops owners from hiring people smarter than themselves ## Timestamps - 00:41 Meet David Hori and Top Line Operators - 02:14 What makes a business worth buying - 03:17 The process once he is inside - 04:01 Staying small by keeping things complicated - 10:14 AI is an amplifier, not a magic pill - 14:20 The durability diagnostic and its four pillars - 16:10 Tribal knowledge and the owner who is the bottleneck - 20:11 What a business operating system does for an owner - 21:30 David's free offer for the first ten owners - 26:41 Where to find David ## Find David - Instagram: @thedavidhori - Web: toplineops.com If your business still runs through you, that is the thing worth fixing first. Book a call: https://crm.frontrangesystems.com/discovery

About

Everything your business knows walks out at 5pm. You rent it back every morning. And every time someone leaves, you lose a piece of how the place runs. That's the problem The Operating System Show exists to solve. I'm Matt Dixon, fractional CTO. I talk to the founders, COOs, and executives who are figuring out how to get their companies to run without every answer living in one person's head. Every week: what broke, what they rebuilt, and where the real leverage was. New episodes every week. Subscribe so you don't miss one.