METAMON — AI ADOPTION METHOD
“People use it”
isn’t enough to renew the licenses.
Metamon measures why the generative AI you already deployed isn’t being used, implements exactly one fix, and re-measures two weeks later. It is Metamorphose’s own AI adoption method — the one we run our own company on.
PROOF — MEASURED ON OURSELVES
We run our own company this way
WHO IT’S FOR
If any of these sound familiar, Metamon is for you
- Company-wide licenses are in place, but you can’t report usage to management
- People say “we use it,” but nobody can say which tasks, or how much it helped
- Renewal is coming and there is no basis for deciding to continue or cut back
- Several PoCs ran, but no task made it into production
If even one applies, measuring why today’s AI isn’t used is cheaper than building the next one.
WHAT CHANGES
What gets solved
| Today | After Metamon |
|---|---|
| Usage is a “feeling” | Numbers at 3 layers: monthly, retained, high-frequency |
| Answer quality is “pretty good” | Judged by fully-wrong rate, unsupported-answer rate, human-correction rate |
| Nobody knows why it isn’t used | Identified in 4 classes (unclear rules / no templates / wrong task fit / distrust) |
| The fix is “another training session” | Top 3 fixes ranked by impact, one of them already implemented |
| Impact can’t be explained to management | A before/after scorecard to decide continue vs. stop |
A basis for renewal decisions
See in numbers how much license spend is idle. Continue, cut, or embed — in management language.
A repeatable in-house pattern
The 90-day adoption plan is written as remaining tasks, ready to roll out across teams even when staff change.
A go / no-go table
Which tasks are safe for production and which should stop — judged by criteria agreed before we start, so it never turns political.
WHAT’S DIFFERENT
How Metamon differs from typical AI adoption support
| Aspect | Typical AI consulting / assessment | AI tool vendors | Metamon |
|---|---|---|---|
| Starting point | Pre-rollout analysis and roadmap | Product rollout and training | Post-rollout measurement — starting from the teams that don’t use it |
| Deliverable | Proposal, priority list | Licenses, manuals | Before/after scorecard + one implemented fix |
| What is measured | Expected benefit (estimates) | Usage logs (dashboard) | Usage ×3, quality ×3, satisfaction, time saved — on one page |
| Pass criteria | Decided afterwards, per project | None | Agreed before we start (e.g. fully-wrong rate <5%, time saved ≥20%) |
| Neutrality | Often steers to own products | Own product assumed | Product-neutral — no recommendations, no referral fees |
| Who does it | A team of junior consultants | Vendor reps | A PMO who has rescued stalled projects makes the calls; an AI team does the aggregation |
| Practiced in-house | — | — | Metamorphose runs its own company the same way |
Others sell “putting AI in” or “building AI.”
Metamon sells “after it’s in — used, and paying off.“
THE METHOD
The Metamon method — 7 phases, 4 cross-cutting tracks
- Assess
- Discover
- Design
- Build
- Validate
- Adopt
- Scale
Managed across four tracks: Business, Technology, Governance, and Change & PMO.
The entry service, the Generative AI Usage Diagnostic, condenses Validate → Scale into six weeks.
- Answer quality ≥ 80 / 100
- Fully-wrong answer rate < 5% (stricter for finance / legal judgments)
- User satisfaction ≥ 70 / 100
- Task time saved ≥ 20%
HOW IT WORKS
How it works — 6 weeks, about 8.5 hours of your team’s time in total
| Week | What happens | Your time |
|---|---|---|
| 1–2 | Usage-log export requested on day one / 10-question survey (5 min) / interviews with the sponsor and managers | Sponsor 60 min · 5 managers × 30 min |
| 3 | Usage rates computed / short interviews with non-users and drop-offs | 6 people × 20 min |
| 4 | Quality review of 30 output samples / root-cause classification / top 3 fixes chosen, one implemented | Mid-review 45 min + 30 min for the affected team |
| 5–6 | Two weeks of operation → re-measurement / final scorecard / 90-day plan / report | Sponsor preview 45 min + final report 60 min |
How results are framed: as “remaining tasks in the rollout design,” never as “the team failed.” The AI sponsor sees the results before anyone in management does.
If usage logs can’t be exported: we substitute self-reported survey data plus manager interviews, and mark the scorecard accordingly.
PRICING
Plans & pricing
Fixed fee, including one implemented fix and re-measurement. No variation by hours worked.
- Adoption scorecard (before/after)
- Root causes of non-use + top 3 fixes
- One fix implemented + re-measured after 2 weeks
- 90-day adoption plan
For companies over 300 employees, we start with one lead department and size the work from there. Price is set with that scope. If you go on to the 90-day accompaniment (monthly), 50% of the diagnostic fee is credited. Usage logs and answer data are viewed inside your environment; only aggregates leave it. We complete vendor security checklists.
WHY THIS PRICE
Why ¥600,000 — what the alternatives cost
| Approach | Indicative cost | What you have after the diagnostic |
|---|---|---|
| Assign one employee full-time to measure it yourselves | ≈ ¥600k/month incl. social insurance; without a measurement framework, 3–6 months of trial and error = ¥1.8–3.6M | Numbers, but no pass line — management still can’t say “good or bad” |
| Hire a freelance PMO / consultant | ¥1.2–1.8M/month (PM/PMO market rate), at least 1.5 months = ¥1.8–2.7M | Depends on the person; no repeatable pattern stays behind |
| Large-firm AI assessment | ¥3–5M (4–8 weeks, team of 3–4) | Pre-rollout proposal and roadmap; post-rollout measurement out of scope |
| Metamon Compact | ¥600,000 (6 weeks, fixed) | Before/after scorecard / top 3 fixes (one implemented) / 90-day plan = a repeatable pattern |
Why we can do it at this price: the measurement framework — metrics, pass criteria, survey, classification — already exists. We don’t sell exploration time; an AI team does the aggregation and classification, and the PMO only judges and reports.
One more comparison: a 300-person company paying ¥4,500 per seat per month spends about ¥16.2M a year. At 30% usage, roughly ¥11.3M of that is idle. ¥600,000 is 5% of that — for the basis to continue, cut, or embed.
* Indicative figures from public sources as of September 2026 (AI consulting fee surveys, published PMO rates, Microsoft 365 Copilot list price).
WHO YOU’LL WORK WITH
Your consultant
Kazuo YutaniFounder, Metamorphose LLC
A PMO who has rescued stalled programs in large-scale SAP, financial services and pharma. He runs his own company on AI operations while working 274 client hours a month. Metamon is that operating model, written down as a method from rollout to adoption and impact measurement.
FAQ
Frequently asked questions
We haven’t rolled out generative AI yet. Does this apply?
The entry diagnostic is for organizations that already have AI in place. Pre-rollout task selection (Assess / Discover) is offered separately.
Does the product matter — Copilot, ChatGPT, Gemini?
No. We are product-neutral and never recommend switching products.
If usage turns out to be low, will the person who championed the rollout look bad?
Results are framed as remaining tasks in the rollout design, and the sponsor sees them before management does.
What do we have in hand after six weeks?
Three things: the before/after scorecard, the root-cause classification with three fixes (one implemented), and the 90-day adoption plan.
Where does the name Metamon come from?
From Metamorphose — the process of an organization changing shape together with AI. Metamon is a registered trademark of Metamorphose LLC.
CONTACT
Tell us the one thing that’s stuck.
Not a sales call — a quick look at where things are stuck.
Tell us “this is where our team stops.”