Hal Wilkerson

@boringlocalseo LOCALRANK

had a 2-hour breakfast in austin last week with a guy running 14 local service businesses across 6 states. combined revenue ~$47M/year. won't let me name him or any of the brands. he has ONE marketing employee. one. not a team. not an agency. one person. i asked the obvious. he said "we stopped doing marketing in 2024." i assumed he meant "switched to referrals" or "kept the same google ads." no. they LITERALLY stopped. zero paid ads. no SEO agency. no social media team. no yelp. no homeadvisor. nothing. what the one employee does: writes ONE piece of content per week per business. 14 pieces total per week. each one is a pricing page, a comparison page, or a "how to choose between X and Y" guide. that's the entire marketing operation. and the businesses are growing. one — commercial HVAC in texas — did $7M in 2023, $11M in 2024, on track for $17M in 2025. ad spend went DOWN, not up. went to zero. he showed me citation tracking. all 14 brands cited as #1 or #2 across ChatGPT, Perplexity, Gemini, and Google AI Overview for their niche+city. average #1 citation rate: 68%. that is insane. the one employee was a freelance writer before. he pays her $95k/year + benefits. jokes she's the highest-ROI hire in business history. the kicker: most of his competitors don't know AI citations are a thing. they're running 2019 playbooks. google ads, yelp ads, occasional agency engagement. he expects to acquire 3-4 of them at fire-sale prices in the next 18 months when their CAC 3x's and his stays at zero. this is the structural shift nobody is talking about. the marketing-agency-industrial-complex spent a decade convincing local owners that customer acquisition required armies of consultants and dashboards. it doesn't anymore. it requires one writer and a measurement tool. we built localrank.so to be the measurement tool. you bring the writer. or you become the writer. that's literally the only decision left. the agencies still pitching "12-month local SEO retainers" in 2026 are selling vinyl in the streaming era. acquisition is coming for them whether they see it or not. Comment "LOCALRANK" and I'll DM you The Local AI Citation Bible — 47 pages, the exact playbook 200+ local businesses use to get cited by ChatGPT, Perplexity, Gemini, and Google AI Overview. (Must be following)

Trying out @roam and gotta say... they are going to give @SlackHQ a run for their money!

@n_wittensleger @ZabihullahAtal We are building ours at PMOrchestrate.ai

@ZabihullahAtal Already seeing this in my pipeline. Founders aren't asking for "AI engineers" anymore — they want operators who can wire agents into actual workflows and own the outcome. Hybrid skill set: PM brain + light eng + ops mindset. Hardest archetype to source right now.

🚨 BREAKING: A new role is quietly emerging and it’s about to dominate the next 5 years. It’s not “AI engineer.” It’s not “prompt engineer.” It’s the Agent Operator. And it will sit inside almost every organization. Most people are still thinking about AI as a tool. That framing is already outdated. What’s actually happening is a shift from: humans using software to humans managing autonomous agents that execute work This is a fundamental redesign of how work gets done. So what is an Agent Operator? An Agent Operator is the person who: • Designs how agents interact with real workflows • Connects tools, data, and systems into agent pipelines • Translates business problems into executable agent behavior • Monitors, corrects, and improves agent performance over time They don’t just “use AI.” They orchestrate outcomes. and this matter because Every function marketing, legal, finance, biotech is becoming “agent-compatible.” Not because companies want it. Because they won’t have a choice. Agents can: • Run research loops • Execute multi-step workflows • Integrate across tools without APIs breaking the flow • Operate 24/7 at near-zero marginal cost The bottleneck is no longer capability. It’s implementation inside real-world systems. Required skills for AI Agent Operator role: → MCPs (Model Context Protocols) Understanding how agents access tools, memory, and structured context. → CLIs (Command Line Interfaces) Because serious agent workflows won’t live in GUIs—they’ll run in programmable environments. → Writing skills (the file kind) Clear specs, instructions, and structured documents. Agents run on precision, not vibes. → agents dot md fluency The ability to define agent roles, constraints, memory, and tool usage in persistent formats. → Business acumen Knowing what actually matters: Where automation creates leverage, not noise. What happens next Enterprises will begin to redesign workflows: Not around employees using dashboards… But around agents executing tasks. That means: • SOPs → Agent playbooks • Teams → Human + agent hybrids • Tools → Composable agent systems When that shift happens, companies won’t just need engineers. They’ll need operators who understand both the system and the business. The leverage is asymmetric One strong Agent Operator can: • Replace fragmented SaaS workflows • Multiply team output without adding headcount • Turn ideas into execution systems in days This is not incremental productivity. It’s operational transformation.

@raleighfelton @teddy_riker @3uildxyz Incredible thinking and very true. (Save the white background which blinds me when I’m trying to read in the dark 😎) Kidding aside this is exactly what I’ve been saying for some time. Thanks for taking the time to put this into words! I’ve shared with my AEO/GEO team.

This is exactly why I built @3uildxyz . Most brands are in the same position as a bad MCP integration — technically accessible but the agent has to guess, and it guesses wrong. We produce the canonical record that sits on the brand's side of that chain. Signed, structured, published at a permanent URL. The agent reads the record instead of inventing one. Just published the founding argument: 3uild.xyz/v/001 Would love your thoughts on this.

This is the next big thing. Hiring AI Integrators or AI Orchestrators. Ton of opportunities in this space for the next few years. College grads, heck High School grads, pay attention.

How big are your dreams?

@Markpricern1 @realEstateTrent True. Would be nice though. But it’s kind of the same concept of term limits. Ain’t gonna happen.

@halwilkerson @realEstateTrent No, I’m quite aware that most of them get money from insider trading, books they write, deals they make outside of politics, and talk shows. There are many other sources, I would think. It’s just the principle of the matter.

Awesome that we all get to be pawns in this political battle. It’s a national embarrassment.

@Markpricern1 @realEstateTrent Do you think congressional salaries are their primary source of income? Think again. Nice thought though.

@realEstateTrent I think all these politicians shouldn’t get paid until they come to agreements on any legislation where others aren’t getting paid. That would probably create a quick turnaround on anything. Let’s start a petition.

Set this up in 30 minutes eh. Kind of hard to believe you wrote this post in less than 30 minutes. I like the concept though. The only challenge is Cowork is only for one person to access on one machine. We are doing something similar to this with a hybrid model with Cowork and OpenClaw so team can access through Slack. But yes the key is definitely the context.

everyone buying Mac minis and setting up OpenClaw for AI agents is solving the wrong problem this is not an infrastructure problem this is a CONTEXT problem the model doesn't matter. the hardware doesn't matter. docker containers don't matter what matters is what the AI knows about your business before you type a single word i run my entire business from one terminal. $100/month. Claude Code CLI. took 30 minutes to set up but the 30 minutes isn't what makes it work what makes it work is what i built around it there's a file called CLAUDE . md that sits in the project root every single conversation the AI opens, this file loads automatically it's the AI's operating manual for my entire business it knows the current metrics, the pricing structure, how i position, who the key contacts are, the file loading strategy for every type of task, communication rules by channel, which tools connect to what when i open a conversation i don't explain anything it already knows but CLAUDE . md is just the index the actual intelligence sits in a knowledge base. modular markdown files organized by folder → core/ has my identity, background, skills, team → network/ has every key relationship with full conversation history. how i met them, what we've discussed, what the dynamic is, what deals are on the table → venture/ has every business and partnership. current status, contacts, strategy, revenue → system/ has client architectures, deployment details, every contact at every client → sales/ has methodology, buyer psychology extracted from 500+ recorded calls, pricing frameworks, objection handling, ICP profiles → platform/ has content voice files for each platform separately. linkedin voice is completely different from twitter voice is completely different from instagram voice. IG same principle applies (story sequences, scriptwriting) separate formatting rules, separate audience profiles, separate examples → planning/ has goals, pipeline, active opportunities → communication/ has style rules and templates broken down by channel → log/ has daily activity tracking, decision logs with full context, network changes every single file has a "Load for:" trigger at the very top it's basically keywords that tell the AI when to load that file → mention someone's name in conversation → their entire relationship file loads automatically with full history → mention a venture → that venture file loads with status, strategy, contacts, everything → mention a platform → that platform's voice file loads with formatting rules, audience, examples the AI only pulls what's relevant for the current task. nothing else. this is what prevents context bloat and keeps the outputs actually useful instead of drowning in irrelevant information then there's MCP Model Context Protocol connects your terminal to every single tool you use - Gmail — multiple accounts - Slack — full DM and channel access - Notion — pages and databases - Google Drive and Calendar - Fireflies — every call transcript and meeting summary - Plane — project management and task tracking - Airtable — CRM, dashboards, agent tracking - Apify — web scraping and data collection so i'm sitting in my terminal and i type - "check last 5 emails with this person" → pulls them instantly - "what did we discuss on our last call" → reads the full transcript and summarizes - "draft a follow-up in our usual tone" → writes it in the exact communication style i use with that specific person based on the relationship history - "sync the project" → gathers Slack messages + email threads + call transcripts → presents consolidated findings - "write a proposal for this deal" → pulls relationship context, previous proposals, pricing history, and generates something that actually sounds like me one terminal. zero tab switching. zero copy-pasting context between tools i built custom slash commands that switch the AI's operating mode entirely - /content activates the content engine with platform-specific voice files loaded - /advisor switches to strategic thinking mode. first principles, decision frameworks - /rotation handles scheduling and coordination there's a "finalize" system where the AI can discuss ideas freely, brainstorm, debate strategy, explore options but it writes NOTHING to any file until i explicitly say "finalize" this prevents half-baked ideas from getting documented prematurely. the AI explores with you but only commits when you're ready multi-agent mode is where it gets wild spawn multiple AI agents running in parallel on the same knowledge base one agent researching a deal. another drafting content. another analyzing call transcripts. another syncing project updates all sharing the same source of truth. all on one laptop connected a Telegram bot for mobile access. team members run the same knowledge base here's what nobody talks about with this setup it compounds → every conversation you have adds context to the system → every deal you close adds intelligence about that buyer type → every call transcript processed adds to the sales psychology database → every correction you make teaches the voice better → every decision you log becomes retrievable strategic context six months in, the AI knows more about how i operate than most people who work with me and that's the part everyone's missing the whole industry is solving for compute. better models, faster hardware, more GPUs compute is not the constraint context is the constraint the AI with the best model but zero business context produces generic garbage the AI with a decent model but deep business context produces things that actually move the needle this is why every ChatGPT wrapper and SaaS tool feels the same. they're all context-poor the moat was never the model. never the hardware. never the framework the moat is: does this thing understand your business at a cellular level before you type a single word mine does people spending $20K on Mac minis, Docker containers, OpenClaw setups, self-hosted models all of it solving for the wrong variable structured knowledge base + $100/month terminal subscription that's the whole setup focus on REVENUE GENERATING ACTIVITIES not installing systems cuz you feel FOMO after going through your X feed

@BradKirschbaum 😂 big agent energy right there.

You really think your agents can beat my agents? 😆 good luck

@BradKirschbaum @BuySellSMB Lol earned a follow! Thank you for your service. And the chuckles.

@BuySellSMB No. Do it. Live on the edge of bankruptcy. Hate yourself for a few years. Cling to delusions of grandeur. Builds character.

If you work at a private equity fund and you're thinking about buying small businesses... Don't do it. Save yourself. Run for the hills. If your boss asks why, just remind them: - No management in place - The owner refuses to figure out Teams - Most of them are not in LA, NYC, MIA - The websites are bad - Someone has to actually run the thing - Some don't have websites! - The CPA is always on vacation and refuses to send financials - The inventory is... uh... not tracked - Some owners have boats on the books! - Each owner will refuse to learn what working capital is

@realEstateTrent It's spelled A-L-G-O-R-I-T-H-M 😜

👀

@GuntherEagleman Won't pass. The same reason term limits will never pass. Great idea though.

It needs to happen.

@recap_david AI Ads

I know a guy running AI street interviews as paid ads... 4.5x ROAS. $0 spent on creators. $0 spent on production. Here's 11 AI video tactics brands are using to create viral ads for nearly $0: 1. AI street interviews (you can't tell they're fake) 2. Podcast-style clips that look 100% real 3. Viral food videos with zero kitchen 4. Anatomical video animations 5. Localization at scale — clone ads into any language instantly 6. Try-on videos without a single model 7. Shock hooks getting 47% higher completion rates 8. Product-style commercials 9. POV adventure content for travel brands 10. Real estate transformation videos 11. AI unboxing videos Most marketers have no idea these exist yet. I put together a free PDF with the exact prompts for all 11 along with the tool I'm using. Comment "AI ADS" and I'll send you everything (must be following so I can dm you!).

Powerful.

@mikefutia SEO

I just vibe-coded a complete SEO Agency In a Box with Claude Code. One keyword in → full research, competitor analysis, SEO-optimized article, metadata, and AI graphics out. All on auto-pilot. Perfect for SEO agencies and DTC brands who are tired of spending an entire day producing a single piece of SEO content. Creating quality SEO content is a huge pain. You're reading competitor articles one by one, trying to find content gaps, writing drafts from scratch, then going back to optimize everything before you even think about graphics or metadata. This vibe-coded tool solves it: → Set up your brand once (value prop, tone of voice, writing samples) → Enter a topic and keywords → SERP API + Firecrawl scrape and analyze the top-ranking competitors → Gemini generates a full research plan with intent, gaps, and strategic positioning → Claude writes the article based on your brand guidelines and the research → Nano Banana Pro generates thumbnail and in-blog images → SEO metadata (title, description, slug) created automatically What this outputs: - Full competitive research and content brief - SEO-optimized articles with internal links and proper formatting - AI-generated images and thumbnail - Meta title, description, and slugs - Batch mode for multiple articles at once - Content rewrite mode for existing URLs that aren't ranking - LinkedIn and X repurposing built in Built 100% in Claude Code. (I'm giving away the prompts) I recorded a full walkthrough showing exactly how every part of this works, and I'm including ALL of the prompts I used for Claude Code. Want access for free? > Like this post > Comment "SEO" And I'll send it over (must be following so I can DM)

Totally get it. We started in the vacation rental space and then had to pivot due to Hawaii regulations and now I love the corporate rental space. I've actually been shocked at the demand, and owners love that higher rev. We do take a margin on repairs though but I can see your justification which makes me wonder if that would work on our side.

@halwilkerson No, we don't do any short or medium term Too much hassle and complexity... better to try to be the best at what we do and let someone else be the best at that

The reason our property management business doesn't charge mark-up on repair costs: When we call them with the bad news that there's something wrong with their building, we never want an owner to wonder if we're motivated by anything other than doing what's best to protect their investment.

@moseskagan Makes sense. Are you in any other space other than conventional long-term? We do hybrid corporate rentals and long-term with a smattering of vacation rentals. We have found it's a great way to increase RPU but it does take a lot more touch points than traditional LTR

@halwilkerson Rev / unit much higher

@moseskagan Interesting- in what ways?

@halwilkerson We are starting to have more single fam and condo owners... in some ways, it's a better business

@moseskagan Do you have much in the way of single-family owners that just have one or is most of your stuff multi-family?

@halwilkerson 1200 doors