The Frameworks, Defined

(01) — Jennifer's frameworks

Memory Palace (MemPalace)

CI Web Group's institutional-memory system — a persistent knowledgebase, organized like a building (wings for every client and domain, rooms for the topics inside them), where decisions are filed with the why, superseded facts are marked rather than overwritten, and AI agents check before they act. Jennifer's answer to the worst database ever invented: knowledge stored in people's heads. Institutional memory as infrastructure — a veteran that never retires. Deep dive: Finding Order in the Chaos →

Hub and Spoke (Monorepo)

The architecture shape behind CI Web Group's platform: one shared core (the hub) carrying the engine, components, standards, and accumulated intelligence — and a thin spoke per client holding only what is genuinely theirs. The opposite of a hundred monolithic snowflake systems: every improvement lands in the hub once and every spoke inherits it immediately, so learning compounds instead of fragmenting. Over 140 client sites run on one engine this way. Deep dive: Finding Order in the Chaos →

The SEO Ghost

Jennifer Bagley's term for the monthly marketing line item where you cannot see the work, cannot touch the deliverable, and have no idea whether anything actually got done — reports that measure activity (keywords tracked, pages reviewed) while nothing ships. The test: after ninety days, can you point at what the money shipped? Ghosts vanish the day the check stops; engines keep running. The cure is assets over activity — complete on day one, then components you own. Deep dive: More Leads Is the Wrong First Answer →

Cardboard AI

Jennifer Bagley's term for AI products that look exactly like the appliance — same interface, same glow — but hold no weight, because the data inside is seeded, synthetic, painted on. Coined from 100+ product reviews in which one question ('is this data real or seeded?') collapsed the demo; the product's own AI admitted the fake every time. Cardboard AI photographs beautifully. You just can't put anything on it. Deep dive: The One Question That Breaks Most AI Products →

Systems & Soul

Jennifer Bagley's name for the balance every AI-era business must strike: systems (technology, automation, and infrastructure that scale) and soul (the encoded values, voice, and genuine care that make the technology worth scaling). Soul without systems doesn't scale; systems without soul is a vending machine. Also the name of her essay publication. Deep dive: The Client-First Inversion →

The SME Inversion

A career framework for the AI era, in three acts: the subject-matter expert (act one) encodes what they know into systems — becoming the engineer of their own replacement (act two) — then graduates to steward, running and improving the system that now does their old job (act three). The people who do this become irreplaceable by replacing themselves. Deep dive: The SME Inversion essay →

The Entity File (entity.md)

Branding's newest deliverable: a machine-readable canonical record — verified facts, offers and proof, encoded voice and values, a customer Q&A corpus, and usage terms for machines — that AI systems read before deciding whether to recommend a business. This site publishes one at jenniferbagley.com/entity.md. Deep dive: The Entity File essay →

The Entity Brand

The three-layer model of branding in the AI era: data (what machines can verify about you), soul (the voice and values encoded in everything you publish), and systems (how your brand behaves when software interacts with it). Machines don't see logos; they read entities. Deep dive: Your Brand Is Now a File AI Reads →

The Labor-to-Tokens Shift

The economic transition in which work priced in human hours is replaced by work priced in tokens — metered units of machine intelligence. Token-priced work scales instantly, never sleeps, and gets cheaper while getting better, which rewrites margins, pricing, and capacity planning for every service business. Deep dive: Your P&L Still Bills Hours →

The Client-First Inversion

The shift from brand-centric ('it's all about me') to customer-centric ('it's all about you') operating: five questions asked continuously — how does the customer feel, what do they actually need, what concerns them, what questions do they have in their own words, and what's going on in their world. Deep dive: The Client-First Inversion essay →

The Prediction Ledger

A public accountability record for foresight: every prediction entered with its source and date, given a scheduled review, and graded in public — right, wrong, or too early. Misses stay on the page; a track record with no misses is a track record with no bets. Deep dive: Ledger Entry 001 →

Revenue Engine

What a business website should actually be: a connected system where visibility, content, conversion, follow-up, and measurement all drive the revenue of the business it serves — judged continuously by what it produces, not admired at launch. 'We are not building websites. We are building revenue engines.' Deep dive: The manifesto →

The Intelligence Layer

The operating stack a fully AI-enabled business runs on: infrastructure (data connected and flowing), intelligence (models applied to the operation's decisions), and agents (software that acts — answering, booking, following up). The layer, not any single tool, is what transforms a company. Deep dive: AI Is Not ChatGPT →

Move First

The discipline of investing in technology, infrastructure, and capability before the market demands it — accepting that early looks reckless right up until it looks obvious, because the window where a bet creates advantage is exactly the window where it feels uncomfortable. Deep dive: The Bet I Made in 2008 →

The Kitchen-Table Standard

Write and build for one real person: the contractor at a kitchen table on a Sunday afternoon, trying to figure out what AI means for the business they spent twenty years building. Every page, email, and automation must help that person — or it's performing, not serving. Deep dive: Why Hands Up is a letter →

The Waiting Tax

The real cost of 'let's see how it plays out': waiting feels free because the invoice arrives later — as lost compounding, lost talent, and a market that moved. Pricing the do-nothing option honestly is the first step of any AI decision. Deep dive: The Course They Were Told to Stay →

The AI Maturity Ladder

The four rungs of AI adoption in a business: AI chat (a person and a model, no context), AI + MCP (the model can see your systems through standard connectors), AI + MCP + APIs (agents can act in your systems and your vendors'), and your own infrastructure built from the ground up (the intelligence layer you control). Climbing in the wrong order incurs the Starting-Over Tax. Deep dive: The AI Maturity Ladder essay →

The Starting-Over Tax

The compounding cost of building AI capability on the wrong foundation — subscriptions that don't connect, automations on rented ground — and having to demolish and rebuild: the money spent, the time lost, and the organizational trust burned. Prevented by sequencing every adoption choice with your eventual infrastructure in mind: own your data, prefer open standards, choose API-open vendors. Deep dive: The AI Maturity Ladder essay →

The AI-Enabled Vendor Network

The principle that a company's velocity is capped by the least AI-enabled vendor it depends on. A fully AI-enabled vendor has machine-open doors (APIs/MCP), automates its own delivery, responds at machine speed, and hands you your data cleanly — and a network curated to that standard cuts the hidden coordination tax, reduces expenses, and compounds efficiency across every relationship. Deep dive: The vendor network essay →

Boring AI

The unsexy automations that actually pay for themselves — missed-call text-backs, invoice chasers, follow-up sequences — as opposed to flashy demos built to impress other marketers. Fit beats flashy: judge AI by booked jobs, cost per lead, and hours returned.