The Four-Stage Maturation Loop
From Artisan to Orchestrated — placing any FDE team on the maturity map and recommending the single next investment that moves them forward.
V1 · Pedagogy
✅ PASS (5/5)
Sweller, Wiggins & McTighe, revised Bloom's, retrieval practice, Vygotsky's ZPD — confirmed and reused from prior lessons.
V2 · Framework lineage
✅ PASS (3/3)
Four-stage model anchored to CMMI v3.0 (Carnegie Mellon SEI, 2023). Logistics 2026 context from Logistics Viewpoints, Inbound Logistics, and Logistics Management.
V3 · Quantitative claims
✅ PASS (3/3)
Gartner forecast: 40% of enterprise applications will feature task-specific AI agents by end of 2026 (up from <5% in 2025). Reported supply-chain AI outcomes: 25% faster disruption response, 30% fewer manual interventions.
V4 · Logical consistency
✅ PASS (4/4)
Section times sum to 75 min; four stages defined non-overlappingly; Analyze-peak outcome assessed by an authentic memo; Bloom progression Remember → Analyze → Evaluate (no >1-tier leap).
Meridian Freight Systems is a 12-year-old freight brokerage in Memphis. Eighteen months ago a senior customer asked them to "figure out the AI thing." They hired their first Forward Deployed Engineer. She delivered. The customer renewed at a 40% expansion.
The head of the FDE practice has three FDEs, a shared Slack channel, no written playbook, and the uneasy feeling that the model that delivered the first seven engagements will not deliver the next twenty. He cannot articulate why.
By the end of this lesson you will have a four-stage map for FDE-team maturity, the ability to place any team on it from a written brief, and the discipline to recommend the single next investment that moves the team to the next stage without skipping the one in between.
By the end of this lesson, you will be able to:
The §7 authentic memo is the assessment. The four-stage model in §4 is the lens; the leading-indicator step is what separates an FDE who can describe the future from one who can manufacture it.
Before this lesson zooms out to the team, recall the individual. These are Lesson 1.2's §11 spaced-review stems.
The model that follows is the team-level analog of the Trinity: an individual FDE has three personas in one head; a team has stages that govern how those individual personas combine into a system that can serve many customers.
Forward Deployed Engineering teams mature through four stages. The model below adapts and compresses Carnegie Mellon's CMMI lineage (Initial → Managed → Defined → Quantitatively Managed → Optimizing, v3.0 2023) into four FDE-shaped stages, and adds a Partner dimension that pure CMMI does not include because FDE work is, by definition, embedded in customer environments.
Click any stage to see its detail, dominant evidence, decisive question, and failure mode:
Artisan
"Ask Asha."
Repeatable Capability
Playbook written and used.
Pod-Led
Rituals that change things.
Orchestrated
Field lessons reach product.
Stage 1 · Artisan
Individual brilliance. Bus factor of one.
Why a loop, not a ladder
CMMI is drawn as a ladder; the FDE Maturation Loop is a cycle. Every stage produces lessons that have to feed back — into the playbook (Stage 2), the rituals (Stage 3), or the product roadmap (Stage 4). A team can mature to Stage 4 and slide back to Stage 2 if the orchestration layer is built but the playbook isn't refreshed. Maturity is not a state. It is a loop.
Read Bayou Cargo's three-year history below — a competitor of Meridian's, a few years ahead. Five snapshots, each with the senior FDE's classification and the single decisive piece of evidence.
What the history shows
Three things to carry forward. First, the transition from Artisan to Repeatable Capability did not happen when the playbook was written. It happened when the playbook was used to onboard. Second, Pod-Led arrived only after a ritual produced a visible change — running a retro is not Pod-Led; producing a change from a retro is. Third, Orchestrated did not begin when the platform team was hired; it began when a partner co-build operated on a written playbook. Each transition is identified by an action, not by a structure.
Vignette A · Coastline Hauliers (annotated)
Vignette B · Inlet Logistics (your turn)
Click the stage you'd assign:
Evaluate move — your turn (scaffold drop)
For Inlet Logistics: what is the single next investment that moves them toward Pod-Led, and what is the leading indicator that would prove it is working within 60–90 days?
Leading indicator (60–90 days): three of the last twelve weekly retros produced a change visible in the next week's work. Lagging indicator: time-to-first-useful-shipped for a new FDE drops from week 4–5 to week 2–3.
Brief
The head of Meridian's FDE practice has asked you, a senior FDE on retainer, to read his two-page situation memo and produce a one-page recommendation. He wants a single classification, three pieces of evidence, one next investment, one leading indicator. The memo will go to his CEO on Monday.
The two-page case · click any sentence to flag it as evidence
Where we are
We have three FDEs. We have run seven customer engagements; six are live, one churned, one expanded to a larger contract. Our wins this year have been concentrated in three customer segments — port operations (2), intermodal rail (2), and retail distribution (3) — but only because each of our three FDEs happens to be strong in one of those segments. The fourth FDE we are about to hire will likely be strong in a fourth segment; we have not yet decided which.
What we have built
Each FDE keeps her own Companion OS folder in a personal Notion workspace. There is a shared Slack channel where engagements are discussed informally; the channel is searchable but not organized. There is no written playbook. Onboarding our second FDE took eight weeks; onboarding our third took six. We have not yet onboarded anyone using anything other than apprenticeship under our most senior FDE.
How we work today
A 20-minute daily standup. No weekly demo day. No retros. The senior FDE reviews PRs ad hoc when she has time. The two newer FDEs ask her for advice over DMs roughly four times a day each. Our most senior customer says she expects to call our senior FDE personally when something breaks.
What the CEO wants
He has told us he wants twenty active customer engagements next year — roughly three times our current count — and he expects revenue per engagement to hold. He has approved the budget to hire four more FDEs in the next six months.
What I think is happening
I am 70% sure my team will not scale to twenty engagements on the model that delivered the first seven. I cannot point at the single thing I would change first. I am hoping you can.
1 · Classify Meridian's current stage
2 · Pick the strongest leading indicator (60–90 days)
Multiple indicators are reasonable. Pick the one that most directly proves the next investment is working.
3 · Write the memo
Open with the answer (Pyramid Principle). Include three pieces of evidence (cite specifics from the case). Name the single next investment as a verb-first 30-day action. Close with the leading indicator as a number with a target.
Meridian is in Stage 1 (Artisan), on the cusp of Stage 2.
Three pieces of evidence. (a) Each FDE keeps her own Companion OS in a personal Notion workspace — knowledge has not been institutionalized. (b) There is no written playbook; onboarding takes 6–8 weeks via apprenticeship — the Stage-2 test (a new FDE shipping in week 3 using only the playbook) is not yet possible. (c) The senior customer expects to call the senior FDE personally when something breaks — the bus factor of one is still the operating model.
Next investment (30 days). By June 16, the senior FDE writes the first version of the Meridian playbook (≤ 12 pages, four named patterns drawn from the seven completed engagements) and migrates personal Companion OS folders into a single shared template. The next FDE hired is onboarded using only the playbook.
Leading indicator (60–90 days). The fourth FDE ships her first useful customer artifact by the end of her third week, measured by a delivered PR or Companion OS entry her customer sees. Target: week 3, not week 6–8.
Anti-recommendation. Do not hire four FDEs in six months. Hire two, write the playbook, then hire the remaining two against it. Hiring four into Stage 1 produces four Artisans, not a Stage-2 team.
Equating maturity with headcount
"We have eleven FDEs, so we must be Orchestrated." Meridian could double in size and stay Artisan.
Correction
Maturity is identified by actions: did the playbook onboard the new FDE; did the retro produce a change; did a field lesson reach the product. Count actions, not bodies.
Skipping Repeatable Capability
"Let's hire a platform team and a partner pod." Without the written playbook, the platform team has nothing to standardize.
Correction
Every skipped stage has to be reclaimed later, usually at 3–4× the original cost. Do Stage 2 fully before any Stage-3 ritual is introduced.
Tools as a substitute for process
"We bought Linear and Notion templates and an LMS." Tools accelerate a process that exists.
Correction
Tools do not create a process. Write the playbook in a Google Doc first; only then choose a tool to host it. The playbook is the four named patterns drawn from your actual completed engagements.
Bring to mind a team you've personally worked with — an FDE team, consulting team, product team, research team.
The Maturation Loop is the framework you'll use in Capstone Milestone 6 to argue your engagement evidence is institutionalizable — not a one-off Artisan-stage win.
What is the difference between the Artisan and Repeatable Capability stages?
Success criterion: Artisan = institutional knowledge in named individuals; Repeatable Capability = knowledge in a written playbook used to onboard to shipping-velocity in ~week 3. The transition is identified by the action of onboarding from the playbook, not by the playbook's existence.
Next lesson
Diagnose → Prototype → Evaluate → Adopt. Together with the Trinity (1.2) and Maturation (1.3), it forms the three-axis map every later course references.
Capstone hook
You'll use the Maturation Loop to argue that your engagement evidence is institutionalizable. Today's §7 is the rehearsal.
Optional reading
The canonical capability-maturity lineage today's four-stage model adapts. Useful for triangulating against a 5-level framework.
Companion OS
File your §7 Meridian memo here. You'll return to it for Capstone Milestone 6.