Playbook Broken by Marc Sirkin Podcast

Playbook Broken by Marc Sirkin

Playbook Broken serves executives navigating the moment when traditional strategy isn't just failing but becoming dangerous. Hosted by Marc Sirkin, Chief Growth Officer, Walk West and former CEO, Third Door Media. marcsirkin.substack.com

Episodes

  1. Aug 24

    The playbook always sucked

    “The playbook always sucked” is a direct quote from Nick Zeckets, Chief Firestarter at Smoke Signals AI, on Playbook Broken. His argument: the B2B go-to-market playbook was mediocre well before AI showed up, and stacking AI on top of it doesn’t fix that. It only speeds up the wreck. We get into what actually replaces it: signal most teams never think to pull, and why ABM’s real failure was the tooling underneath it, not the framework. Chapters: 00:00 The Playbook Always Sucked 04:00 Frankenstein AI and the SaaSpocalypse Myth 13:00 GitHub Repos, OSHA Filings, and Alpha Signal 17:00 Who Owns Signal Inside the Org 19:00 Faster Playbook or Different Playbook This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit marcsirkin.substack.com

  2. Jul 14

    The AI Didn't Remove the Operational Tax. It Moved It.

    Thirty days. Four teams. The readout happened. The main thing we learned: we have a data problem. And you probably have the same problem as we do! Not the headline anyone wanted. The honest one. Four teams, multiple prototypes, multiple different problem spaces, common binding constraints. The AI didn’t remove the operational tax. It moved it. From doing the work manually to preparing the data so AI can actually do the work. Every team hit the same wall from a different direction. Files named differently. Fields filled out differently. Source of truth in dispute. In this episode: what each team built, what they found, and a four-stage maturity model for what’s actually happening at Walk West and at most companies running these sprints right now. Three things from the readout: * Don’t turn a finding into a program. Pick one source of truth for one workflow. Make it canonical. Build on top of that. Then do it again. * The teams that win the next decade aren’t the ones who installed AI fastest. They’re the ones who got to stage three first. Stage three runs on clean data and shared process, not prompts. * AI didn’t change the work. It just made it impossible to keep ignoring. Thanks for following along. This was worth doing in public. More Playbook Broken soon! Chapters 00:00 The finding nobody wanted 00:45 Four teams, four prototypes: the recap 02:00 One binding constraint across all four teams 02:30 AI moved the operational tax, it didn't remove it 03:00 The maturity model: four stages 03:30 Stage 1: Individual AI literacy (why most rollouts stop here) 03:45 Stage 2: Individual agentic workflows 04:00 Stage 3: Departmental agentic workflows (where the readout broke down) 04:30 Stage 4: AI native 05:00 Where does the truth live? 06:00 Don't turn a finding into a program 07:00 Honest callback: ranking vs. energy 07:30 What the teams that win are actually building This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit marcsirkin.substack.com

    The AI Didn't Remove the Operational Tax. It Moved It.

  3. Jul 7

    Every AI Prototype Is a Data Audit in Disguise

    Three weeks in. Four teams. Thirty days. Here’s what nobody told us going in: you think you’re running an AI sprint to learn about AI. What you actually learn is who’s on the boat. The builders aren’t engineers. They’re producers, PMs, designers. They didn’t wait for training. They just started. The people who waited, they’re a signal too. Not a moral judgment. A behavioral one. And AI surfaces it faster than any other kind of change work I’ve seen. The other thing that showed up this week: every prototype hit a data wall. Walk West has few formal, structured, clean data sets about itself. Asana, EverHour, Google Drive, but not consistently used from project to project. So every AI tool the team tried to build became a data audit in disguise. One team built a tool to predict whether a project would hit margin before it got sold. It didn’t make the final cut. But in building it, she found that margins vary wildly from project to project and nobody’s really looking at them. That’s not a productivity tool. That’s a different category of finding. Three things from week three: * Waiting for permission is itself the signal. * AI didn’t create the data and process problems. It just made them impossible to ignore. * The sprint didn’t make choices. It made the existing choices visible. Episode 4 and the grand finale is up next… stay tuned! Chapters 00:00 What you actually learn in an AI sprint 01:00 Builders vs. waiters: a behavioral signal, not a moral one 01:30 Infotech CEO: unlimited access and the restructure model 02:00 Walk West team gains momentum 03:00 Cross-team convergence: hitting the same walls 04:00 Every prototype is a data audit in disguise 05:00 The margin tool: a different category of finding 06:00 What leadership does with the visibility This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit marcsirkin.substack.com

    Every AI Prototype Is a Data Audit in Disguise

  4. Jun 30

    Things on Screens Have Gravity

    Week two of the Walk West sprint. The team built a few things. A rough client portal prototype. Some data real, some fake. Some buttons worked, some didn’t. That’s exactly the right fidelity for that stage. The minute it hit the screen, the conversation got sharper. Questions shifted from “should we build a portal?” to “where does a client approval actually go and who can prove it happened?” Those are operating questions. The prototype made them possible. But that same prototype almost became the project. Because things on screens have gravity. And AI makes that problem worse, because it makes unfocused work just as fast as useful work. Three things from week two: * A prototype is evidence of possibility, not of priority. Those two need to stay in tension. * AI doesn’t just make useful work faster. It makes unfocused work faster too. The discipline is knowing the difference. * Transformation has a translation layer. Insider language makes people feel stupid. That’s the mentor’s fault, not the team’s. Episode three is up next… stay tuned! Chapters 00:00 When build shows up before spot is done 01:00 Ranking the ideas: what rose and what didn't 02:00 A prototype is evidence of possibility, not priority 03:00 The client portal: what the prototype made visible 05:00 When approval history becomes the real product 06:00 The jagged edge of AI inside an organization 08:00 The translation layer: PRD vs. six plain questions 09:00 Fast convergence vs. divergence across teams 10:30 Cross-team collaboration has to be made explicit 11:30 Permission to work on more than one idea 13:00 AI as a behavior diagnostic This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit marcsirkin.substack.com

    Things on Screens Have Gravity

  5. Jun 25

    AI Without Leverage Is Just a More Expensive Version of the Same Firm

    Walk West is a 20-person marketing agency in Raleigh. I joined a few months ago as chief growth officer. We ran a 30 day challenge with four cross-functional teams running a sprint to figure out where the firm goes next. This is episode one of four and kicks off the series. In episode one: why we haven’t picked a project yet, what a team of producers surfaced in 40 minutes that nobody planned for, and the real lesson from day one. AI is a process diagnostic before it’s a process replacement. Three things I’d steal from that first week: * Energy is data. When someone lights up about something not on the agenda, that’s a signal worth following. * The producer usually has the prototype. Leadership has the framing. You need both. * Spot is a phase, not a meeting. Skip it to feel like you’re moving and you’re just running on vibes. More as the work unfolds. Chapters 00:00 The first question most companies get wrong 01:00 What Walk West is and why I took the role 02:05 The sprint structure: four teams, 30 days, no roadmap 02:55 Spot, Build, Test 03:35 Team 4's first meeting: 14 ideas, none chosen yet 04:30 The producer who already had her own AI workflow 05:00 Nick Saban, Lionel Rosen, and process thinking 05:50 What the room actually revealed 07:00 Three takeaways from Day 1 This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit marcsirkin.substack.com

    AI Without Leverage Is Just a More Expensive Version of the Same Firm
  • 5 Episodes

About

Playbook Broken serves executives navigating the moment when traditional strategy isn't just failing but becoming dangerous. Hosted by Marc Sirkin, Chief Growth Officer, Walk West and former CEO, Third Door Media. marcsirkin.substack.com

Information