Working notes on AI-augmented marketing and getting your analytics to tell the truth.
Meta says 412. Google says 380. GA4 says 481. None reconcile. Here is the server-side attribution stack I build for PMC clients - UTMs, click IDs, Meta CAPI, Google Enhanced Conversions, and the weekly reconciliation pass that turns four platform reports into one defensible number.
Most of the confident advice about which index each assistant reads is not documented by the vendors. Here is what is actually measurable: what gets cited, how much traffic it is worth, and the fundamentals that raise you across all of them at once.
Every agency is pitching the same venture-backed SaaS startup while a middle market generating more than $10 trillion in annual revenue sits with broken sites, spreadsheet CRMs, and no marketing function. Here is why that gap is the entire thesis of PMC - and why unsexy B2B clients have better unit economics than the glamorous ones.
Five gates you can clear in a minute on your phone, and five that take longer and are the ones that actually cost money. With the thresholds corrected to what Google publishes, rather than the numbers everyone repeats.
Ninety-one percent of buyers say they would use their agent again. Only 18 percent of repeat buyers actually do. Repeat clients and past-client referrals made up 41 percent of the typical agent's business in NAR's 2025 Member Profile, and most agents lose that channel by month seven. Here is the Day 0 through Day 365 cadence that keeps it open.
In 2026 a guaranteed number one ranking is not a promise, it is a tell. Here is what "#1 on Google" actually leaves out, and what to demand from a marketer instead.
When a stylist leaves, the clients follow the channel, not the chair. A non-solicitation clause is not the protection owners think it is. Here is the shop-owned CRM, SMS line, email list and booking setup that changes what a departure costs.
The overlap tax is the monthly cost of paying multiple SaaS vendors to do the same job, plus the operational cost of your customer existing as a different record in each one. Here is the four-axis audit I run on every front office, with 2026 prices and the consolidation order that actually works.
In the first 90 days building PMC I stood up my own internal tooling, built a three-client retainer book, and ran the entire operation on the Claude stack. Here is the exact system, the math, and the parts most operators leave out of the pitch deck.
AI has made marketing accessible to everyone, which flooded the market with people who know just enough to spend your budget without earning a return. Here is the honest state of the industry, why it mirrors what is happening to software development, and how to tell the difference between someone who looks capable and someone who is.
After auditing a stack of local restaurants this year, the same picture keeps repeating. AI Overviews now appear on 68 percent of local searches against 39 percent for the map pack, and Gemini recommends locations at ten times ChatGPT's rate. Here is the phased roadmap that follows from that.
Most restaurants collect almost no first-party customer data, then pay the platforms to reach the customers they already served. Here are the practical mechanics of capturing diners at the table, online, and through the POS - without gating anything.
Most restaurant websites are not broken in obvious ways. They leak through small, fixable issues - PDF menus, dead-end forms, neglect signals - that cost covers without anyone noticing. Here are the five most common, and the fixes.
Your business listings are no longer just directory entries. They are the source data that Google, Apple Maps, and the AI assistants pull from. Here is how to audit and fix the listings layer for the AI era.
For a local restaurant, the Google Business Profile drives more discovery than the website. Most are claimed and then neglected. Here is the 2026 playbook for the levers that actually move local rankings and AI answers.
Early-stage sales runs out of the founders' heads and a few spreadsheets. That works until it doesn't. Here is where it breaks as you scale, and what a real source of truth for accounts and reorders looks like.
Technical SEO guides are written for enterprise sites with a million URLs. A small business does not need most of it. Here is the short list that actually moves rankings for a site with a few dozen pages, in 2026 terms.
ASO is not the keyword-stuffing game it was a few years ago. Screenshot captions are indexed, Apple auto-generates tags, and retention is a ranking signal. Here is what actually moves installs now, in order.
Most messaging frameworks are a tidy doc that dies the moment a real prospect asks a real question. Here is how to build one that holds up under pressure, mapped to how a buyer actually decides.
Agencies sell you a team and deliver a pod where nobody owns the result. A solo operator with AI leverage now ships the same volume with one accountable person. Here is the honest case, including where an agency still wins.
Companies are now choosing between another headcount and a bigger AI budget, and most of that spend is premium rates paid for commodity work. A look at what the model-cost myths get wrong, with the 2026 numbers.
Most audits that cost you nothing are a checklist someone runs to book a sales call. A real one orders the findings by consequence and tells you which it measured and which it judged. Here is the difference, and the five leaks worth looking for.
The published research on software waste is almost all produced by companies selling the cure, and some of it does not survive being read carefully. Here is what holds up, what does not, and why none of it describes a business your size.
GA4 removed four attribution models in 2023, including the ones most guides still tell you to pick. What is actually left, and the input work that decides whether any of it is trustworthy.
"AI-augmented but human-led" gets thrown around a lot. Here is the working definition behind every PMC engagement, and where AI helps versus where a human has to own the call.
