The Agency Hour

Agency Mavericks

The Agency Hour is for digital agency owners like you who want to start, grow and scale their businesses. It’s like hanging out with friends while getting a masterclass in how to tackle the most significant obstacles you’re facing as an entrepreneur.

  1. Oct 22 ·  Video

    The Biggest Mistake Agencies Make When Hiring

    Send us Fan Mail Somewhere out there, a business is paying a full salary to a person whose job, two weeks of every month, is copying and pasting between spreadsheets. Command C, Command V, lunch. One of our Headless Agency members told us about it on a call last week. We all lost it. Then I remembered how most agency owners spend their Tuesday, and I stopped laughing. This week I give you one question to ask before you hire anyone, and it tells you whether you're about to pay a salary for computer work that an AI employee should be doing. This is a solo episode. Three stories from the Feed (AI search traffic, Anthropic's $16.8 billion quarter and what HubSpot's blog team learned from a year of going AI first), what's happening inside real agencies on our client calls, then the big idea: typing on a keyboard is basically free now, and your clients pay you for knowing what to do. In this episode: The D or S test: write every task in a role, mark D if it needs a decision and S if someone could follow it off a checklistWhy Marie Haynes's data on 145 million sessions shows ChatGPT sending 0.62% as many visits as Google organic, and why those visitors still convert at 13.8% against about 3%Anthropic's July to September revenue of $16.8 billion, up from $11.5 billion the quarter before, and why you should sell AI to your clients before they buy it from the labWhat HubSpot's blog team found after a year of going AI first, and the MIT Technology Review finding that only 30% of managers and owners rate judging AI output as an absolute mustA Sydney lead gen agency that moved more than 1,000 pages from Thrive Theme to Divi with an AI employee, at about $200, then $60, then $10 per siteHow a New Zealand content writer rebuilt a sales dashboard in a day, and the $78,000 roadmap it helped close at 1.45% of the client's target revenuePeter lifting his discovery sessions from $500 to $1,500, plus Jeremy, Candy, Daniel, Bashir and James and what an AI employee did for each of themWhy more revenue does not need more hands on keyboards, and how the fear of AI comes from 45 years of identity built on typingThis week's homework: open last week's calendar, mark every hour spent on steps someone else could follow, write the biggest block as 10 plain-English lines and hand it to an AI employeeChapters: 0:00 Cold open: the copy and paste salary 0:47 Intro 1:08 The plan and fan mail 1:39 The Feed: ChatGPT traffic against Google organic 3:52 The Feed: Anthropic's $16.8B quarter 5:03 The Feed: HubSpot's year of going AI first 7:01 Inside real agencies: client calls 10:41 The big idea: typing is basically free now 15:49 The one question before you hire anyone 16:48 Homework and fan mail 18:38 Outro Mentioned in this episode: Search Engine Land (Marie Haynes's AI search traffic data): https://searchengineland.comViktor: https://mvrx.link/get-viktor (affiliate link)Tuesday's video, the three things that get an agency to a million with AI: https://youtu.be/9WeTF557N8gLeave us fan mail: https://mvrx.link/fanmailWork with us: https://theagencyhour.com/headlessEvery episode of The Agency Hour: https://theagencyhour.com

  2. Oct 15 ·  Video

    How a Video Agency Went From $1M to $2.5M (with Darren Rayner)

    Send us Fan Mail Eighteen months ago, Darren Rayner's video agency was at about $1 million in revenue. Now it's at $2.5 million. Darren got there by giving his best ideas away on LinkedIn, putting out a free report for his ideal clients, and booking a pile of 15-minute calls with the people who hire video agencies. Those calls have converted something like $600,000 to $700,000 in the last 12 months, and he hasn't spent a cent on ads for them. He charges clients for discovery before he pitches, which I don't see many video agencies doing. Darren runs Magnafire, a Vancouver creative studio that makes social-first video and photo for outdoor and sport brands like Red Bull, Meta, lululemon, Arc'teryx and Brooks Running. Full transparency, he's a client of ours. In this episode we get into how paid discovery works, how he lands big brands, why audiences are losing trust in AI-generated content, and what he'd tell every brand to do for social in 2027. Before the interview, I answer a fan mail question from Jonathan Holborn: how do you decide when Viktor can do the job and when a tool is worth paying for? In this episode: Why a video agency charges for discovery, and how it weeds out pitches that were never going anywhereGoals, audience, KPIs: the three questions Darren asks every clientGiving value away on LinkedIn, and the free Outdoor and Sport Social Report 2027The journalist framework: report what's working, add your take, hand over the playbook15-minute calls with ideal clients that converted $600K to $700K in 12 monthsFrom $1M to $2.5M in about 18 months, and what Darren thinks drove itWhy entirely AI-generated content costs brands trust, and where Darren does use AIThe content pyramid: one hero film, eight shoulder assets, and why the shoulder assets reached about 20 times more peopleThe one social strategy for 2027: world-building and recurring formatsValue per second, and what Red Bull and That Pedal Show teach about building an audience firstFan mail: how I decide when Viktor can do the job and when a tool is worth paying forChapters: 0:00 Cold open: $1M to $2.5M 0:17 Intro 1:10 Fan mail: how do you vet AI tools? 4:42 Meet Darren Rayner and Magnafire 5:36 Why a video agency charges for discovery 7:10 What paid discovery is, and how it filters serious clients 10:14 How clients react to paying for a pitch 12:21 How discovery changes the shoot 13:44 Goals, audience, KPIs 17:41 Who Magnafire works with 18:23 How Darren lands Red Bull and lululemon 19:53 Giving value away on LinkedIn 22:02 The journalist framework 23:29 15-minute calls that converted $600K+ 25:18 100 conversations in 100 days 27:26 From $1M to $2.5M in 18 months 28:54 Why Darren didn't go it alone 32:50 AI content and audience trust 36:43 Where AI belongs in your workflow 40:12 One shoot, every platform: hero and shoulder assets 42:37 TikTok vs Instagram vs YouTube Shorts 46:01 The one social strategy for 2027 47:30 Think like a YouTuber: binge-worthy brands 49:25 Value per second: why brands still make ads 53:31 Why big brands are slow to change (and Red Bull isn't) 54:43 Build the audience first: That Pedal Show 56:17 Where to find Darren 56:45 Outro Connect with Darren: Magnafire: https://www.magnafiremedia.com/LinkedIn: https://www.linkedin.com/in/darren-rayner-8b90384/Mentioned in this episode: Viktor: https://mvrx.link/get-viktor (affiliate link)GoHighLevel: https://mvrx.link/ghl (affiliate link)Expert Secrets by Russell BrunsonThat Pedal Show on YouTubeLeave us fan mail: https://mvrx.link/fanmailWork with us: https://theagencyhour.com/headlessEvery episode of The Agency Hour: https://theagencyhour.com

  3. Oct 9 ·  Video

    Google Just Put It in Writing: A Human Has to Check AI Content.

    Send us Fan Mail Google updated its AI content guidance this month. It now says it is critical to manually fact check and review all AI-generated content before publishing. Most agencies will read that as bad news. I read it as the best news this industry has had this year. If AI does the typing, the human who signs off is the part your client pays for. This week's big idea is how to price that: sell the outcome and the sign-off. McKinsey already does, with about a quarter of its global fees tied to results. This week: the news, one big idea, some homework, and your questions. Your homework is to Google your top five clients' brand names, screenshot the AI overview, and send each client what you're fixing. In this episode: OpenAI's new always-on dots agents and Meta's Muse for Small Business, and why both ask for approval before they actAI Overviews now show on more than 80% of branded searches, and what to tell your clients about clicksAccenture's 22% jump in a day, and McKinsey and EY moving to outcome-based feesGoogle's new line: manually fact check and review all AI-generated content before you publishHubSpot's blog traffic collapse, and what it says about publishing AI content at volumeThe model that works: AI does the typing, the human does the judgmentTreat AI like a team of juniors: nothing ships without sign offHow to price it: sell the outcome and the judgment, and stop selling four blogs a monthThe human-in-the-loop checklist: facts and numbers, real authors, original stories, titles, meta and alt textHow to answer a client who says "I'll just use AI myself"Chapters: 0:00 Cold open 0:18 Intro: what's in this episode 1:35 The Feed: OpenAI Dots and Meta Muse 3:32 Google's AI overviews on branded searches 5:11 Big four consultancies move to outcome-based fees 6:42 The Deep Dive: Google's new rules for AI content 9:46 The model that works: AI types, humans judge 10:49 Nothing ships without sign off 11:51 How to price it: sell the outcome and the judgment 13:06 The human-in-the-loop checklist 14:53 Done Beats Perfect: Google your clients' brand names 15:45 Fan mail: send us your questions 16:57 When a client says they'll just use AI themselves 19:24 Outro and next week's guest Work with us: Free AI For Agencies community and AI Employee course: https://www.skool.com/agency-mavericksApply to work with Agency Mavericks: https://theagencyhour.com/headlessMentioned in this episode: OpenAI launches dots, always-on AI agents with their own cloud computers (The Next Web): https://thenextweb.com/news/openai-dots-always-on-ai-agents-cloud-computers-devdayMeta launches Muse for Small Business, connecting its AI agent to Shopify, Stripe, QuickBooks, Klaviyo and Meta ad accounts (Shopifreaks): https://www.shopifreaks.com/meta-launches-muse-for-small-business-connecting-its-ai-agent-to-shopify-stripe-quickbooks-klaviyo-and-meta-ad-accounts/Google AI Overviews jump from 26% to 80% of branded queries (Search Engine Land): https://searchengineland.com/google-ai-overviews-jump-branded-queries-september-492962AI Overviews on branded queries tripled in September (DemandSphere): https://www.demandsphere.com/blog/branded-ai-overviews-september-2026/Accenture jumps 22% as outlook eases AI disruption fears (Resultsense): https://www.resultsense.com/news/2026-10-02-accenture-forecast-it-services-rally/AI is reshaping how McKinsey makes money (Business Insider, via Yahoo Finance): https://finance.yahoo.com/news/ai-reshaping-mckinsey-makes-money-195132745.htmlGoogle Search's guidance on generative AI content: https://developers.google.com/search/docs/fundamentals/using-gen-ai-contentGoogle: Manually Factcheck & Review AI-Generated Content For Accuracy (Search Engine Roundtable): https://www.seroundtable.com/google-updates-ai-content-guidelines-factcheck-review-42217.htmlCreating helpful, reliable, people-first content (Google Search Central): https://developers.google.com/search/docs/fundamentals/creating-helpful-contentHubSpot's SEO collapse: what went wrong and why? (Search Engine Land): https://searchengineland.com/hubspot-seo-organic-traffic-drop-451096YouTube to cut reach of channels reposting barely changed clips (Search Engine Journal): https://www.searchenginejournal.com/youtube-original-shorts-reposted-clips-reach/591774/Viktor: https://mvrx.link/get-viktorNext week: Darren Rayner from Magnify Media in Canada. He's one of our clients and a video production agency that works with brands like Lululemon, Red Bull and Meta. He shares his process for getting more reach from video on social. Leave us fan mail: ask anything about growing your agency and I may answer it on the show: https://mvrx.link/fanmail Every episode of The Agency Hour: https://theagencyhour.com Got a question you want answered on the show? Email me: troy@agencymavericks.com. I read every one myself.

  4. Oct 1 ·  Video

    Let AI Do the Typing. Get Paid for the Thinking.

    Send us Fan Mail "No matter how prominently we show the links, users won't click." An OpenAI engineer wrote that in February 2023. It surfaced this month in the New York Times' court brief against OpenAI and Microsoft, and it should change how you report to your clients. If your monthly report leads with traffic, that number is only going one way. Traffic was never what your clients cared about anyway. They care about what happens after the click. As AI takes over more of the delivery, your job is to sit with the client, understand what matters to their business, and get paid for the thinking. Let AI do the typing. This is the first episode of the new format: the week's news, one big idea, some homework, and your questions. This week's big idea: twenty years of SOPs and offshore teams are giving way to skill files and AI employees. Here's what to hand to AI, and the jobs you should never hand over. In this episode: The OpenAI engineer who admitted users won't click, and what it means for your traffic reportsClaude Sonnet 5.5: AI work keeps getting cheaper, so don't drop your fees to match your costs"I'll just do it myself in Claude": how to answer clients who say itWhy you can outsource marketing but not accountabilityFrom SOPs to skill files: the new way to hand off repeatable workThe jobs AI won't replace: sales, strategy, relationships and tasteKeyword research for a few dollars with DataForSEOHow Viktor runs a 90-day content strategy from keyword research to monthly reportThe litmus test: if the job is typing into a computer, AI should be doing itThis week's homework: design one meaningful client report, then hand it to AIChapters: 0:00 Cold open 0:16 Intro: the new format 3:40 The Feed: OpenAI says users won't click 7:08 Claude Sonnet 5.5 and the falling cost of AI work 10:37 You can't outsource accountability 13:04 The Deep Dive: from SOPs to skill files 16:22 Your three jobs AI won't replace 18:42 What AI is already doing: keyword research 19:50 How Viktor runs a content strategy 24:38 The litmus test: should AI do this job? 26:00 AI employees are vending machines, not chatbots 27:47 Done Beats Perfect: hand off your monthly client reports 30:18 Fan mail: send us your questions 31:24 Outro Work with us: Free AI For Agencies community and AI Employee course: https://www.skool.com/agency-mavericksApply to work with Agency Mavericks: https://theagencyhour.com/headlessMentioned in this episode: Court filing quotes OpenAI engineer: users "won't click" links (Search Engine Journal): https://www.searchenginejournal.com/openai-engineer-users-wont-click-links-court-filing/591303/Introducing Claude Sonnet 5.5 (Anthropic): https://www.anthropic.com/claude-sonnet-5-5You can outsource marketing. You can't outsource accountability. (Search Engine Land): https://searchengineland.com/outsource-marketing-not-accountability-491648Simple Numbers, Straight Talk, Big Profits! by Greg Crabtree: https://gregcrabtree.net/books/simple-numbers-straight-talk-big-profits/DataForSEO: https://dataforseo.comViktor: https://mvrx.link/get-viktorLeave us fan mail: ask anything about growing your agency and I may answer it on the show: https://mvrx.link/fanmail Every episode of The Agency Hour: https://theagencyhour.com Got a question you want answered on the show? Email me: troy@agencymavericks.com. I read every one myself.

  5. Sep 29 ·  Video

    The AI Proposal That Lost Us a Client (with Duncan Isaaksen-Loxton)

    Send us Fan Mail Duncan Isaaksen-Loxton used AI to write a 13-page proposal for an eight-website job. The research that used to take two days took 30 minutes. The proposal was done in a couple of hours. Then the client wrote back: "It sounds like you did this proposal with AI." The trust they'd built on the discovery call was gone, and so was the job. Duncan runs sixfive.io, a fractional CTO for small businesses. He looks after the tech so owners can get on with the work they're actually good at. In this episode we get into why business owners DIY their tech until it bites them, what happens when your team pastes client data into personal ChatGPT accounts, why AI without solid foundations just speeds up the mess, and why the human relationship still wins the work. In this episode: The "this will only take half an hour" myth, and what DIY tech really costs youStaff using personal ChatGPT accounts with business data (and whose problem that is when it leaks)A hospital that handed a government RFP to ChatGPTHow to sell backups and security without scaring peopleWhy you've shared an API key in Slack (yes, you have) and how to stopOnboarding, offboarding and why most businesses only need three toolsGoogle Workspace vs Microsoft, and the Teams rantAI is a power drill: it doesn't make you a tradieWhat to do when a client sends you a 30-page AI documentThe AI proposal that lost Duncan a client, and what he changedChapters: 0:00 Cold open 0:25 Intro 2:58 Who is Duncan and what sixfive.io does 7:11 From web developer to fractional CTO 9:33 Why business owners DIY their tech 12:45 Staff putting business data into personal ChatGPT 14:14 Data breaches and how AI amplifies bad decisions 17:00 Selling protection without the fear 21:11 Slack Connect, API keys and secrets hygiene 23:17 Onboarding, offboarding and the three-tool stack 26:09 Google Workspace vs Microsoft 29:31 Techies, builders and designers: where websites fit 31:40 AI content and the power-drill problem 35:09 The 30-page AI document policy 39:40 Human connection can't be automated 43:15 The AI proposal that lost a job 46:40 How to reach Duncan 49:08 Outro Connect with Duncan: sixfive.io: https://sixfive.io (book a 15-minute chat)LinkedIn: https://www.linkedin.com/in/loxton/Mentioned in this episode: Viktor: https://mvrx.link/get-viktorDoppler: https://doppler.comEvery episode of The Agency Hour: https://theagencyhour.com Got a question you want answered on the show? Email me: troy@agencymavericks.com. I read every one myself.

  6. Jun 5

    Headless: The Future of Work

    Send us Fan Mail If you want to increase capacity, improve margins, reduce operational drag, and build a business that scales without becoming dependent on more headcount, click the link below and apply to work with us.  https://mvrx.link/headless-podcast  In this episode, Troy Dean reads his new manifesto, Headless: The Future of Work, and explores one of the biggest shifts in business since the internet itself. For decades, we've built companies around a simple assumption: if you want more output, you hire more people. That assumption is breaking. AI agents, wearables, voice interfaces, and headless software are changing the way businesses operate. The question is no longer how many people work for you. The question is how much value each person can create. In this episode, Troy explores: Why headcount is becoming a misleading measure of success How AI is replacing execution while humans focus on strategy Why the office may become obsolete The collapse of the traditional agency growth model How "headless businesses" will outperform their competitors The future of work, leadership, productivity, and leverage Why value per human will become the most important business metric of the next decade This isn't a conversation about prompts, tools, or hacks. It's a conversation about what happens when the interface leaves the desk. If you're a founder, operator, executive, consultant, or agency owner trying to understand where work is heading, this episode is for you.

    Headless: The Future of Work
4.2
out of 5
5 Ratings

About

The Agency Hour is for digital agency owners like you who want to start, grow and scale their businesses. It’s like hanging out with friends while getting a masterclass in how to tackle the most significant obstacles you’re facing as an entrepreneur.

You Might Also Like