# Demo dataset blueprint — synthetic seed records Concrete, **fully synthetic** seed data for the Acme Catering catering CRM demo in the Acme Catering demo org (`` / CX26). Nothing here is a real Acme Catering customer, contact, or figure. **Frame:** in this demo org, **Acme Catering is the company using D365 Sales**. The Accounts are **Acme Catering's catering customers**. Users are Acme Catering CSMs. **Before creating anything:** describe the target tables. Logical names, option-set values, and required fields must come from the live org, never from this file. This file specifies *what* to create; the org tells you *how*. --- ## Geography Acme Catering is headquartered in Chicago. Anchor the dataset in **Chicago metro**, with a secondary cluster for national-account and multi-location scenarios. Keep addresses plausible but generic. --- ## Demo users (CSMs) Map to real enabled users in the demo org — do **not** create licensed users. Read the org's `systemuser` list first and pick 3–4, then assign these personas to them. | Persona | Role | Used in | |---|---|---| | **CSM A** | Catering Sales Manager — Chicago North territory | S1, S2, S3, S5 | | **CSM B** | Catering Sales Manager — Chicago South / Suburbs | S3, S4, S5 | | **CSM C** | Catering Sales Rep — National Accounts | S4, S5 | | **Leader** | Director of Catering — manager of the above | S3 approvals, S5 | Record the actual GUID↔persona mapping in the run output so the presenter knows who to log in as. --- ## Scenario 1 — Lead-to-Customer Lifecycle ### Leads | Name | Company | Source | Owner | Purpose | |---|---|---|---|---| | Marcy Ellison | **Harborline Logistics** | Web form | CSM A (auto-assigned) | The hero lead — worked live | | Dennis Prakash | **Harborline Logistics Inc.** | Phone | CSM A | **Near-duplicate** — trips detection on create | | Renee Watts | **Lakeshore Dental Partners** | Web form | CSM B | Second lead so the queue isn't a single row | | Tomas Kirby | **Vantage Point Advisors** | Referral | CSM A | Qualified, ready to convert | - Include a **held-back lead** to create *live* during the demo (manual creation proof point). - Populate lead fields: est. headcount, event date, delivery vs pickup, budget. ### Opportunity / Quote path (from Tomas Kirby / Vantage Point Advisors) - Opportunity: **"Vantage Point Advisors — Quarterly Partner Lunches"**, est. value ~$18,500 - Quote with **line items** drawn from catering products (below), showing a **volume discount** - Quote revision #2 to prove revision handling ### Activities (spread across the above) - 2 completed phone calls, 1 completed email, 1 completed appointment - 2 open tasks — one **overdue**, one **due today** - Enable audit history on lead/account/opportunity --- ## Scenario 2 — Olo Integration and Ownership Preservation ### The worked lead - Lead: **Priya Raman — Northline Health Group (Downtown Clinic)** - Owner: **CSM A**. Source: Web form. Created ~14 days before demo day - History: 3 completed activities, 1 open follow-up task — proves it was genuinely worked ### The inbound Olo order - Simulated Olo order matching on **email + phone** - Order value ~$1,240, delivery, ~40 headcount - Must carry an **external order ID** and **source = Olo** so attribution is visible post-conversion ### The multi-match exception (customer explicitly asks about this) Create **two** plausible matches for a second inbound order: - **Contact:** Priya Raman @ northlinehealth.example — Downtown Clinic - **Contact:** P. Raman @ northlinehealth.example — Northline Health Group (parent) Same domain, similar name, different account. This is the "what happens if multiple records match" moment. Do not resolve it in the data — the ambiguity *is* the demo. ### Post-conversion state - Account created from lead retains **CSM A as owner** - **Source attribution preserved** (original web-form source, not overwritten by Olo) - Order history visible on the account timeline --- ## Scenario 3 — Ownership Change and Sales Credit Governance ### The contested account - **Account: Cardinal Ridge Corporate Center** — a multi-tenant office campus, high catering volume - Current owner: **CSM A** - Claimant: **CSM B** ### Make the dispute genuinely arguable | Actor | Activity to seed | |---|---| | CSM A | 4 activities over the prior 90 days — original relationship, first order | | CSM B | 5 activities in the last 30 days — including the meetings that led to the recent sale | ### The recent sale (the crux) - Closed-won opportunity: **"Cardinal Ridge — Q3 All-Hands Catering"**, ~$22,400 - **Close date: 6 days before the ownership-change request** — the "order placed before approval" case ### The request - Ownership change request record: CSM B requesting transfer, with justification text - Approval chain: **Level 1 = Director of Catering**, **Level 2 = VP** — status **Pending** - Audit/approval history populated with prior decisions on other accounts, so the history view isn't empty ### Supporting - 2 previously-resolved ownership changes on other accounts (one approved, one rejected) to make the ownership/dispute report meaningful --- ## Scenario 4 — Hierarchy, Duplicates, Data Governance ### The healthcare hierarchy (3 levels — answers "how many levels are supported") ``` Northline Health Group (L1 · parent · owner: CSM C) ├── Northline Medical Center – Chicago (L2 · owner: CSM C) │ ├── Northline MC – Cardiology Pavilion (L3 · owner: CSM C) │ └── Northline MC – Outpatient Surgery (L3 · owner: CSM A) ← different owner ├── Northline Health – Oak Park Clinic (L2 · owner: CSM B) ← different owner └── Northline Health – Naperville Campus (L2 · owner: CSM C) ``` This directly proves **"can ownership differ by location" = yes**, and shows **3 levels**. ### Contacts (spread across facilities) - Dr. Alan Whitfield — Chief of Staff, Northline Medical Center - Gina Alvarez — Practice Administrator, Oak Park Clinic - Marcus Bell — Facilities Manager, Naperville Campus - Priya Raman — Downtown Clinic (links back to Scenario 2) ### Duplicate pairs — each created via a **different channel**, with conflicting values **Pair 1 — channel conflict** | | Record A | Record B | |---|---|---| | Name | Northline Health – Oak Park Clinic | Northline Oak Park Clinic | | Source | Web form | ezCater | | Created | 14 months ago | 3 weeks ago | | Phone | 312-555-0142 | *(blank)* | | Address | Full address | Partial | | History | 6 orders | 1 order | → Older record has history, newer has cleaner naming. **Survivor rules visibly matter.** **Pair 2 — Olo-generated** | | Record A | Record B | |---|---|---| | Name | Naperville Campus – Northline | Northline Health – Naperville Campus | | Source | Olo | Manual | | Email | shared domain | shared domain | | Owner | CSM C | CSM B | → Merging changes ownership. Forces the ownership-on-merge conversation. **Pair 3 — contact-level duplicate** - Two contact records for **Gina Alvarez** with different emails (work vs personal) and different loyalty IDs — exercises the **email / phone / device ID / loyalty ID** identity-resolution keys. ### Validation case - One account (**"Beacon Street Medical"**) missing required fields — industry, billing address, primary contact — to demo required-data validation. --- ## Scenario 5 — CSM Productivity and Leadership Visibility ### CSM A's day (the "morning of" view) - **8 open tasks**: 3 overdue, 3 due today, 2 due this week - 2 appointments on demo day - 3 Outlook-captured email activities (proves activity capture) - 1 manually logged phone call ### Lapsing accounts (must actually be lapsed) | Account | Last activity | Prior value | |---|---|---| | **Greenfield Academy** | 118 days ago | $9,200 | | **Halstead & Boyle LLP** | 147 days ago | $14,600 | | **Riverbend Manufacturing** | 201 days ago | $31,000 | High prior value + long silence = a compelling win-back alert. ### Pipeline breadth (for credible dashboards) - **18–25 opportunities** spread across all stages - Distributed across **CSM A / B / C** so ownership reporting has shape - Close dates spread across the current and next two quarters - Mixed values: several $2–8K, several $10–25K, 2–3 at $40K+ (national accounts) ### Closed-won history (for revenue trend charts) - **6+ months** of closed-won opportunities, **8–15 per month**, varied values - Include a visible **seasonal lift** — holiday catering — so the trend line tells a story - Mix **first-party vs ezCater-sourced** so the first-party conversion narrative is chartable ### Leadership view needs - Opportunities tagged by **source** (web, Olo, ezCater, referral, campaign) - Accounts tagged by **segment** (corporate, national, healthcare, education, SMB) - Enough closed-lost with **loss reasons** for win/loss reporting --- ## Catering products (for quote and order line items) Build a small, realistic catering catalogue. Per-person pricing where natural. | Product | Unit | Price | |---|---|---| | Original Sandwich Box Lunch | per person | $14.50 | | Skinny Sandwich Box Lunch | per person | $12.75 | | Sandwich Party Platter (serves 10) | each | $89.00 | | Salad Party Platter (serves 10) | each | $79.00 | | Soup Bar (serves 10) | each | $65.00 | | Cookie Tray (24 ct) | each | $32.00 | | Beverage Service — Bottled | per person | $2.75 | | Boxed Lunch — Gluten-Friendly | per person | $15.25 | | Delivery & Setup Fee | each | $25.00 | Add a **volume discount tier** (e.g. 10% at 50+ people) so quote pricing rules are demonstrable. --- ## Data hygiene rules 1. **All dates relative to demo day.** Recompute if the demo moves. "Overdue" must be truly overdue. 2. **Synthetic contact details only** — use `.example` domains and `555-01xx` phone numbers. 3. **No real Acme Catering customers, contacts, or financials.** Not from the RFP, not from email, not from the working doc. 4. **Realistic but round-ish values.** Avoid figures that look like real contract values. 5. **Name records for findability** — the presenter has 15 minutes per scenario. 6. **Leave the intended mess in.** Duplicates and ambiguity are the demo in Scenarios 1, 2, and 4. 7. **Reuse before creating.** Query the org first; the CX26 org ships with sample accounts (Fabrikam, Trey Research, Northwind Traders, Coffee Lab APJ) that are *not* catering-relevant — prefer purpose-built records over bending those, but do not delete them.