Course Specification — Tier 2: Supplier Data Foundations

Programme: ZAYAZ Supplier Programme · Tier: 2 — Data Gatherer Brand: EcoWorld Academy (co-branded: "Brought to you by [CLIENT NAME] via EcoWorld Academy") Course ID: SP-T2 Status: Spec v1.0 — for production Upstream dependency: ESG Data Snapshot (SP-T1 m005) — schema v1.0. This spec fulfills the content debt flagged there: locate-it hints and minimal-setup guidance for all 15 Snapshot items (Appendix A).

---

1. Purpose & positioning

Tier 2 is where the programme changes gear: from understanding to doing. Tier 1 was watched; Tier 2 is performed. The course exists to get one thing to happen: the supplier's first real, structured dataset live on ZAYAZ. Everything in the course serves that outcome.

Three jobs:

1. Convert the Snapshot inventory into entered, published data — fast-lane items first, so momentum builds before the harder items appear.
2. Teach data literacy at working level — why the reporting fields are structured as they are, what a defensible number looks like, how to estimate honestly when exact figures don't exist.
3. Make the score move visibly — the supplier should see their score respond to their own data entry during the course. That moment, not any argument, is what makes the programme feel worth continuing.

The automation fence: everything in this course is achievable manually on the free tier. ZAYAZ automated collection (subscription) is shown once, late, as "how this feels when it runs itself" — demonstrated, never required, never blocking.

2. Learner profile

• Same person as Tier 1, now warmed up: passed the knowledge check, ideally completed the Snapshot and the client assessment.
• New constraint that shapes everything: this course competes with their day job. Tier 1 asked for 35 minutes of attention; Tier 2 asks them to go find their energy bills. Expect the course to be consumed in fragments over days or weeks — every module must be a self-contained work session with a clear "done" state.
• Snapshot tells us their starting position. Design for all three: suppliers with most sources at hand (fast path through), suppliers with many UNKNOWNs (locate-first path), suppliers with many DONT_HAVEs (setup-first path, longest journey — protect their morale explicitly).

3. Learning objectives

By completion, the learner can:

1. State the four qualities of a defensible data point (real source, defined period, consistent unit, repeatable method) and apply them when entering data.
2. Locate the source documents for each item in their Snapshot fast lane and enter the data correctly (right field, right unit, right period).
3. Track down or set up sources for UNKNOWN/DONT_HAVE items using the per-item guidance.
4. Produce an honest estimate where exact data doesn't exist, and flag it as an estimate on the platform.
5. Publish their dataset and explain what changed on their score and on the client's view.
6. Describe a simple routine for keeping the data current (and name what the automated alternative does).

4. Format & runtime

• Delivery: On-demand Synthesia modules, but structured as work sessions: each data module ends with a do-it-now task and a checklist state, not just a recap. "Pause and go do it" is the intended usage pattern, said out loud in m001.
• Video runtime: ~42 minutes across 9 modules (ceiling 6 min/module). Total effort honestly stated to the learner: 2–4 hours including data gathering, spread over as many sessions as they need. Understating this would cost trust; Tier 1's caveat ("real effort, paid back more than once") set this up.
• Presenter: Same as SP-T1 (mini-me avatar pending test; stock fallback). Scripts presenter-agnostic.
• Fallback texts: m001–m009.md → R2 fallback_text/SP-T2/.
• Completion gate: FIRST DATASET LIVE — precise definition in §6. No quiz. Optional 2-question self-checks may close modules m002–m007 (non-gating, instant feedback) — decide at scripting.

5. Module breakdown

#	Module	Runtime	Type
m001	Welcome, Data Gatherer — how this course works	4 min	Orientation
m002	What makes a number defensible	5 min	Concept
m003	Energy, fuel & heating — your first live data	6 min	Work session (E)
m004	Waste, water & travel	5 min	Work session (E)
m005	People data — headcount, safety, training	5 min	Work session (S)
m006	Governance — policies, certificates, your own suppliers	5 min	Work session (G)
m007	When you don't have the number — honest estimation	5 min	Concept + rescue
m008	Going live — publish, and watch your score move	4 min	Completion moment
m009	Keeping it alive — routines, automation, and Tier 3	3 min	Wrap + bridge

m001 — Welcome, Data Gatherer

• Congratulate the tier: badge visibly changed — first proof the system responds to them.
• Set the usage pattern explicitly: "This is not a course you watch in one sitting. Each module ends with a task. Do the task, come back. Your checklist remembers everything."
• Show the personalized checklist built from their Snapshot: fast lane (HAVE) first, then UNKNOWN, then DONT_HAVE. If no Snapshot exists: generic checklist + 2-minute detour prompt to complete it (per schema).
• Honest effort statement: 2–4 hours total including gathering, at their own pace. Frame: "at the end, something real exists that didn't before — and everyone who matters can see it."

m002 — What makes a number defensible

• The four qualities, each with a good/bad example: real source (a bill, not a memory), defined period (which 12 months?), consistent unit (kWh vs MWh — pick once), repeatable method (could you produce the same number next year?).
• Why it matters commercially, not academically: defensible data survives client questions; fragile data creates the follow-up emails this whole system exists to kill.
• Introduce the estimate flag concept early (fully taught in m007): an estimated number, flagged, is defensible; a guessed number, unflagged, is not.
• This module is the course's conceptual core — everything after applies it.

m003 — Energy, fuel & heating (first work session)

• Structure used by all four work-session modules (m003–m006): (1) what these fields mean and why clients care — 1 min; (2) document anatomy on screen: where on a real bill the needed number lives — 2–3 min; (3) entry walkthrough in ZAYAZ basic reporting: field, unit, period, source note — 1–2 min; (4) do-it-now task + checklist state.
• Energy first deliberately: nearly every company has energy bills (most common HAVE), the number is unambiguous, and success here sets the pattern. First data point live by end of m003 is the design goal.
• Covers Snapshot items: e_energy_bills, e_fuel_records, e_heating.
• Screen-capture scenes isolated (kit convention) — bills shown as anonymized fictional examples (reuse Brenna Components A/S fictional-company convention for all sample documents).

m004 — Waste, water & travel

• Same structure. Covers: e_waste_invoices, e_water_bills, e_travel_records.
• Waste anatomy: contractor invoices often show container volume, not weight — teach the volume-to-weight conversion honestly as a flagged estimate (first live use of the m002/m007 concept).
• Travel: expense-system exports count as a source; perfect trip-level data not required at this tier.

m005 — People data

• Covers: s_headcount, s_safety_log, s_training_records, s_working_conditions_policy.
• Sensitivity note in-script: aggregate numbers only — the platform never asks for personal data about individual employees. Say it explicitly; it defuses a common quiet fear at SMEs.
• "Zero is data": an accident log showing zero incidents is a reportable, score-feeding fact — but only if a log exists. No log ≠ zero accidents; that distinction is the module's teaching moment.

m006 — Governance

• Covers: g_org_responsibilities, g_code_of_conduct, g_supplier_overview, g_certificates, g_insurance_docs.
• Reframe from Tier 1 carried forward: governance = writing down what the owner already knows. The one-page responsibility note is a legitimate document — show a Brenna Components example.
• g_supplier_overview gets a forward-looking line: "keep this list warm — at Tier 5, it becomes the start of your own supplier ecosystem." (Seed, one sentence, no more.)
• Certificates/insurance: upload-or-reference walkthrough per platform capability (platform check needed: file upload vs metadata reference at free tier — resolve before scripting).

m007 — When you don't have the number

• The rescue module — placed after the work sessions so it meets learners exactly where they're stuck, but signposted from m002 ("if you hit a wall in any module, module seven is your way through").
• Honest estimation method: base (what you know) → assumption (stated) → calculation (simple) → flag as estimate → replace when real data arrives.
• Two worked examples: heating estimated from building size when no separate meter exists; travel estimated from expense totals.
• Hard line, stated plainly: an estimate is a bridge, not a destination. Flagged estimates score better than empty fields; real sources score better than estimates. The ladder points one way.
• DONT_HAVE morale passage: "some things can't be estimated — only started. Setting up a simple log today means next year, you're in the fast lane. That's not falling behind; that's how every documented company began."

m008 — Going live (the completion moment)

• Publish walkthrough: review entries → mark dataset live → watch the score update on screen, live. This is the course's emotional payoff; give the score movement its own scene, unhurried.
• Then the client view (same demo-dashboard framing as SP-T1 m005, framing rule applies): tier badge now "Tier 2 — Data Gatherer," score moved, data status visible. State only what is visible; no client-motive language.
• Completion mechanics stated: dataset live = tier complete = Tier 3 unlocked immediately. Congratulation email fires (asset, §8).

m009 — Keeping it alive

• The quarterly 30-minute routine: new bills in → numbers updated → estimates replaced where real data now exists. Small, named, calendar-able.
• The automation glimpse (the fence, shown once): ~90 seconds of ZAYAZ automated collection — connected sources, auto-updated fields, estimate replacement suggestions. Framing: "everything you just did by hand, running by itself." One sentence on where to learn more. No pricing talk in-course, no urgency, no repeat.
• Tier 3 bridge: "You have data. Data becomes powerful when it points somewhere. Next: setting a target an SME can actually own — and can defend."

6. Completion gate — precise definition

Tier 2 completes when the supplier's dataset is LIVE, defined as ALL of:

1. Reporting period set (a defined 12-month window).
2. At least 6 data points entered across all three sections — minimum 3 E, 2 S, 1 G — each with source note or estimate flag.
3. Dataset status set to published (visible to inviting client(s)).

Rationale: low enough that a small supplier with thin sources can genuinely reach it (energy + fuel + waste, headcount + safety, one policy = 6), high enough that the score visibly moves and every section is touched. Estimates count — flagged estimates are legitimate data at this tier (consistent with m007; the gate must not punish honesty). No video-watching requirement: a supplier who publishes a qualifying dataset without watching anything has completed the tier — the courses serve the outcome, not the other way around.

Platform events on completion: tier badge update (client dashboard) → Tier 3 unlock → congratulation email with Tier 3 CTA.

7. Branding & localization

• Co-brand lockup and EcoWorld kit standards as SP-T1 (Arial, #2D6A4F, #EAF4EE/#B7DFCA, [CLIENT NAME] variable).
• All sample documents (bills, logs, policies) use the fictional Brenna Components A/S — never real supplier data, never real client names.
• Languages: same launch set as SP-T1. Document-anatomy scenes use EN sample documents with localized VO at launch; localized sample documents are a fast-follow (bill layouts differ by country — note for the localization backlog, priority by supplier volume per market).

8. Production asset checklist

Asset	Notes
Architecture MDX	From this spec
Video scripts × 9	Work-session structure m003–m006; screen scenes isolated
Fallback texts m001–m009.md	R2 fallback_text/SP-T2/
Appendix A content → platform	15-item locate-it hints + setup guidance (checklist UI strings)
Sample document pack	Brenna Components A/S: energy bill, fuel statement, waste invoice, water bill, safety log page, one-page responsibility note
Offline collection sheet (XLSX)	One tab per section, columns matching platform fields + unit + period + source — for the person who gathers numbers from colleagues before entering; produced on the standard openpyxl pipeline
Self-checks (2 Q × m002, m007)	Non-gating reinforcement; concept modules only
Completion email copy	Congratulation + Tier 3 CTA, per language
Estimation examples sheet	The two m007 worked examples as a takeaway page

9. Success metrics

• Activation: % of Tier 1 completers entering ≥1 data point within 14 days of unlock.
• First-data speed: median time from Tier 2 start to first live data point (m003 design goal: same session).
• Completion: % reaching the live-dataset gate within 60 days (expect materially lower than Tier 1 — this gate costs hours, not minutes; benchmark before judging).
• Estimate usage: % of published datasets using ≥1 flagged estimate (validates m007; near-zero suggests learners stall instead of estimating).
• Path analysis: completion rate by Snapshot profile (mostly-HAVE vs mostly-UNKNOWN vs mostly-DONT_HAVE) — reveals where guidance is failing.
• Fence signal: % of Tier 2 completers viewing automation info after m009 (interest, not conversion, at this tier).

10. Open items before scripting

1. Platform check RESOLVED: free tier gets full certificate/insurance handling (upload). Wider principle adopted: free-tier suppliers unlock more platform features as they climb the tier ladder — progressive capability is part of the programme's value design. m006 scripts as a full upload walkthrough.
2. Score-movement latency RESOLVED: direct updates (score on publish) will be live; only batch/cron-style processes are not. m008's live score scene scripts as designed — and becomes a platform requirement: publish → synchronous score update.
3. Optional self-checks RESOLVED: in for the two concept modules (m002, m007) only — 2 questions each, non-gating. Work sessions m003–m006 get none; the task is the check.
4. 3E/2S/1G minimum RESOLVED — now a platform requirement: the scoring model must be tuned so that a dataset meeting exactly the minimum (6 points: 3E/2S/1G) produces a clearly visible score movement on publish. The m008 payoff scene depends on it.

---

Appendix A — Per-item guidance (fulfills SP-T1 schema content debt)

Serves two surfaces: (a) Tier 2 checklist UI strings (UNKNOWN → "Find out", DONT_HAVE → "Set up"), (b) course scripts m003–m006. EN master; localize with course batch.

E items

e_energy_bills · Find out: Check with whoever pays invoices — energy bills arrive monthly or quarterly from your electricity supplier; most supplier portals let you download 12 months in one go. If your landlord pays energy, ask them for your share or the building figure. · Set up: Register for your energy supplier's online portal (free, ~10 minutes) — history is usually available immediately.

e_fuel_records · Find out: Fuel card statements (from the card provider), or search your accounting system for fuel purchases. Leasing companies often report fuel per vehicle. · Set up: A fuel card (even for one vehicle) turns every fill-up into automatic records; alternatively a simple per-vehicle log — date, litres, odometer.

e_heating · Find out: Separate gas/oil/district-heating invoices; if heating is electric, it's already inside your energy bill — mark it that way, don't double-count. If the landlord provides heat, ask for the annual figure. · Set up: No separate meter and no landlord figure? Module 7 shows the building-size estimation method — flagged estimate now, meter or landlord figure later.

e_waste_invoices · Find out: Your waste contractor's invoices — often filed with facilities or accounting. Container size and emptying frequency are usually printed on the invoice. · Set up: Ask your waste contractor for an annual summary (standard request, usually free); if collection is municipal, note container size and pickup schedule — module 7 covers conversion.

e_water_bills · Find out: Annual or quarterly bill from the water utility; in rented premises, ask the landlord for your share. · Set up: Water utilities issue statements to the account holder on request — one email; if truly unavailable, a flagged headcount-based estimate bridges the gap.

e_travel_records · Find out: Expense system exports (search "travel"), travel-agency annual statements, or accounting entries for flights/hotels/mileage. · Set up: From today: one expense category or spreadsheet for travel — totals are enough at this tier; trip-level detail is a later refinement.

S items

s_headcount · Find out: Payroll knows this to the person — ask for headcount and contract-type split (full/part-time, temporary) as of a chosen date. · Set up: Fewer than ~20 people? A dated one-page list is a legitimate record — write it today, update on change.

s_safety_log · Find out: Ask whoever handles HR or safety; check any legally required incident register for your industry — it may exist without being called a "safety log." · Set up: Start a log today — date, what happened, action taken. An empty log is a valid record ("zero is data — if the log exists"). One page or one spreadsheet is enough.

s_training_records · Find out: Certificates from courses, safety briefing sign-off sheets, onboarding checklists — usually scattered; one folder makes them a record. · Set up: One sheet — name, training, date. Add historical entries people remember, then maintain forward.

s_working_conditions_policy · Find out: An employee handbook, contract annex, or collective agreement reference counts — check with HR or whoever writes contracts. · Set up: One page covering hours, leave, conduct, and complaint route beats no page; write down current practice — that IS the policy, documented.

G items

g_org_responsibilities · Find out: Company registry extract shows ownership; any org chart (even in a drawer or an old pitch deck) shows responsibility. · Set up: One page: who owns the company, who decides what, who is responsible for safety, finances, quality. Fifteen minutes with the owner produces it.

g_code_of_conduct · Find out: Check documents signed for large customers — many suppliers have accepted a customer's code of conduct without keeping a copy filed; that document counts. Note: a code of conduct is also part of ZAYAZ platform onboarding, stored on the platform — this Snapshot item is about your company's own code, if one exists beyond that. · Set up: Guidance defers to the onboarding requirements design (CoC + minimum ESG data, stored in the code-of-conduct table) — align this item's setup path with that work before scripting m006. No standalone template asset in this course.

g_supplier_overview · Find out: Accounting system → vendor list export, sorted by spend. Top 10–20 by spend is the useful core. · Set up: List your main suppliers from memory and invoices — name, what they supply, country. Keep it warm: it becomes the seed of your own ecosystem at Tier 5.

g_certificates · Find out: ISO, industry, and product certificates often live framed on walls or in a sales folder — collect digital copies in one place; check expiry dates while at it. · Set up: No certificates is a valid answer at this tier — record "none" honestly; if customers keep asking for a specific one, that's market signal for later, not a Tier 2 requirement.

g_insurance_docs · Find out: Your broker or insurer can resend current policy documents same-day — one email. · Set up: If cover is genuinely missing (not just the paperwork), that's a business risk beyond ESG — flag it to management; for the platform, record what exists.