What this part delivers, and why
Foundations exists to answer one question before any diagnostic begins: does this client's leadership genuinely understand what content engineering is, why it matters now, and that they want us to assess them? Skip it and the Assess engagement lands on people who think they hired you to "do AI" or "fix the blog" — and the findings get argued with rather than acted on. We build shared language, show the art of the possible, tie it to their real frustrations, and walk out with a signed mandate.
- 1 · Frame the problem — exec briefing on why content engineering, why now, and the cost of inaction.
- 2 · Establish shared language — a workshop on the strategy / engineering / operations triad, and the glossary.
- 3 · Show the art of the possible — the "Can You Tell?" test, AI Share-of-Voice, and structured-vs-blob content, made tangible.
- 4 · Connect their pain to the model — lightweight discovery linking real frustrations to the discipline.
- 5 · Secure the mandate — agree the sponsor, the scope of Assess, and what success looks like.
Frame the problem
Get the leadership team to understand why content engineering, why now, and what it costs to keep ignoring it — using the AI forcing function as the hook, not the headline. The goal of this session is interest, not a decision.
- Exec briefing deck (template below)
- A sense of the client's AI ambition & pressure
- Shared framing of the problem
- Exec appetite to continue
Time-poor execs do not care about content models. They care about the gap between the AI demos they've been sold and the results they're getting. Open with that gap — "you've spent on AI, here's why it isn't paying off yet" — and you'll have the room. Lead with taxonomy and you'll lose them inside ninety seconds.
Do not drown the first meeting in jargon. DITA, MCP, RAG, SKOS, GraphRAG — every one of those is a reason for a CMO to glaze over and conclude this is an IT problem. Keep the whole briefing in plain business language; the acronyms can come out to play in the Assess engagement, with the practitioners.
Establish shared language
Give everyone in the room the same words. Most disagreement about "content" is really people meaning three different things at once — strategy, engineering, operations. The triad fixes that, and the glossary keeps it fixed after we leave.
- The one definition — content engineering structures content as reusable, machine-readable data (10m)
- The triad — Strategy (the "CEO of content") / Engineering (the "CTO of content") / Operations (the engine room) (25m)
- Map their own world onto the triad — who owns what here today? (25m)
- Spot the gap — which leg of the triad is thin or missing (20m)
- Hand out and walk the glossary; agree the terms we'll all use (10m)
Illustrative, not measured — Assess replaces these with the client's real scores. The shape, though, is almost always the same: engineering trails.
- Triad explainer one-pager (template below)
- Playbook glossary (from Part 6)
- Shared vocabulary across stakeholders
- A first, rough read on which triad leg is weakest
When you ask "who owns content engineering here?" the usual answer is a long pause, then someone points at the web team and the web team points back. That silence is the finding — it tells you the engineering leg of the triad doesn't exist as a role yet, which is exactly what Assess will confirm and quantify.
Show the art of the possible
Make the abstract concrete. People remember what they can see and touch — so we demo the "Can You Tell?" test, sketch AI Share-of-Voice, and put structured content next to a "blob" side by side. This is the session that turns intellectual agreement into wanting it.
- Structured vs blob — show the same content as a freeform page, then as typed fields; ask which an AI can reuse, assemble, and cite (15m)
- The "Can You Tell?" test — run the swipe tool live; let them try to spot human vs machine and watch the ~50% guess rate land (15m)
- AI Share of Voice — sketch the idea: are you cited when your category is asked about, or is your competitor? (10m)
- So what — none of this works without the structure underneath; that's the whole point (5m)
Content Science, Oct 2025 — the organisations that scale tend to be those with mature content operations. Maturity is the multiplier, not a stick.
- "Can You Tell?" swipe tool (in the hub)
- A structured-vs-blob worked example
- One illustrative SOV sketch
- Visceral understanding of the payoff
- Demand to know "where do we stand?"
The "Can You Tell?" demo is persuasive precisely because it's honest — so keep it honest. Use a real, mixed set; don't rig it. If someone in the room reliably spots the machine-written ones, that's a useful conversation, not a failure. The point is the bar, not a magic trick.
Connect their pain to the model
Lightweight discovery — not the Assess interviews, just enough to tie the discipline to the frustrations they already live with. Translate a vent ("we re-do everything for every channel") into the engineering language ("you have no single-source reuse"). The "Is this us?" checklist is the spine of this step.
- Short conversations with the sponsor and 2–3 content/marketing leads
- Walk the "Is this us?" symptom checklist together — which signs do they recognise?
- Name each symptom in triad terms — strategy gap, engineering gap, or ops gap
- Use O'Keefe's one-question heuristic — "are you doing a lot of copy-and-paste?" — as the fastest tell
- "Is this us?" symptom checklist (below)
- The triad framing from Step 2
- A shortlist of recognised symptoms
- Their pain, mapped to the discipline
- The case for Assess, in their own words
Secure the mandate
Convert understanding into a decision. Name the executive sponsor, agree the scope and shape of the Assess engagement, and write down what success looks like — so Assess starts with a mandate, not a vague hope.
- Confirm the executive sponsor — one named, accountable person
- Agree the scope of Assess — which domains, teams and properties are in
- Define success — what the Assess readout must answer for this sponsor
- Sign the mandate / charter one-pager; book the Assess kickoff
- Mandate / charter one-pager (below)
- Recognised symptoms from Step 4
- Named sponsor + signed mandate
- Agreed Assess scope & success criteria
- Assess kickoff booked
A mandate without a single named sponsor is not a mandate — it's a group of people who all assumed someone else was accountable. If nobody will put their name on the charter, you don't have buy-in yet; you have polite interest. Better to find that out now than three weeks into Assess.
RACI & effort summary
Who does what across the part. R Responsible · A Accountable · C Consulted · I Informed.
Hover a row to focus it; click a role header to spotlight that person's whole column.
| Activity | Exec / sponsor | Content lead | Marketing lead | Lead consultant |
|---|---|---|---|---|
| Frame the problem | A | I | C | R |
| Establish shared language | C | C | C | R |
| Art of the possible | I | C | C | R |
| Connect their pain | C | C | C | R |
| Secure the mandate | A | C | C | R |
| Week | Focus | Consultant days |
|---|---|---|
| Week 1 | Frame the problem, establish shared language, prep the demo | ~2 |
| Week 2 | Art of the possible, lightweight discovery, secure the mandate | ~2.5 |
The artifacts you use and leave behind
Four core templates are spelled out below; the full set produced in this part is indexed at the end.
Why content engineering, why now — the 8-slide spine
- 1 · The hook — "AI is the payoff of content engineering done well." One line, one slide.
- 2 · The gap — you've invested in AI; the returns aren't matching the ambition. Why?
- 3 · What content engineering is — structuring content as reusable, machine-readable data (plain language, no jargon).
- 4 · The triad — strategy / engineering / operations, and where the missing piece usually sits.
- 5 · The forcing function — RAG, AI search and agents all need structured content to work.
- 6 · The cost of inaction — without structure, AI amplifies the mess faster, and competitors get cited instead of you.
- 7 · The honest reality — adoption is near-universal, scaling is not; maturity is the multiplier (see note).
- 8 · What we'd do next — a short, low-risk diagnostic (Assess) to find your starting point.
Evidence to cite honestly on slide 7: a 2025 study of content operations found ~86% of enterprises use AI but only ~29% report scaling it well — and the organisations that do scale tend to be the ones with mature content operations (Content Science, Oct 2025). Frame maturity as the multiplier, not as a stick.
Strategy / Engineering / Operations — the three disciplines
| Discipline | Owns | Shorthand | What it produces |
|---|---|---|---|
| Content strategy | What content to make, and why — audience, journeys, editorial plan | "CEO of content" | The plan: what earns its place |
| Content engineering | How content is structured as reusable, machine-readable data — models, metadata, taxonomy, schema | "CTO of content" | The structure: what makes content AI-ready |
| Content operations | The people, workflow, tooling and governance that run the system day to day | The engine room | Published content, repeatably |
The one-line version: strategy decides what & why; engineering structures it as reusable, machine-readable data; operations runs the system. Engineering is the leg most teams are missing — and the one that makes content retrievable and citable by AI.
Signs you have a content-engineering problem
Tick the ones you recognise — the tally tells you whether it's worth diagnosing.
- You re-create the same content from scratch for every channel — lots of copy-and-paste
- Nobody can say with confidence how much content you have, or where it all lives
- "Content" means three different things depending on who's in the room
- You've bought AI tools, but the output still needs heavy rework to be usable or on-brand
- Your content isn't tagged consistently, or your taxonomy exists only in someone's head
- Translating or localising a campaign costs far more than it feels like it should
- When you ask an AI assistant about your category, it cites a competitor, not you
- No single person owns the structure of your content — it falls between strategy and IT
Recognise three or more and you have a content-engineering problem worth diagnosing. The fastest single tell (O'Keefe): "Are you doing a lot of copy-and-paste? If you are, your content back end can't support any channel — including AI."
Authority to run the Assess engagement
- Executive sponsor — one named, accountable person: __________
- Why now — the business pressure driving this (one or two sentences)
- Scope of Assess — which content domains, teams, sites and regions are in / out
- What success looks like — the questions the Assess readout must answer for the sponsor
- Stakeholders & access — who we'll need to talk to, and what data we'll need
- Timeline & budget — the indicative shape of the Assess engagement (≈3–4 weeks)
- Sign-off — sponsor signature & date; Assess kickoff booked for: __________
Keep it to one page. The point is not a contract — it's a single artifact that proves a named person has agreed to the diagnosis and what it's for.
Entry & exit gates
The quality bar that says this part is genuinely ready to start, and genuinely finished.
- A senior contact willing to convene the leadership team
- Enough access to gauge the client's AI ambition and pressure
- The exec briefing tailored to this client's context, not generic
- Stakeholders share the same definition and vocabulary
- Their real pain is mapped to the discipline, in their own words
- An executive sponsor is named and accountable
- The Assess scope & success criteria are agreed and the mandate signed
- The Assess kickoff is booked