Empire Teas CRM — Client Kick-off Question Set (Draft)
Date: 2026-07-22 Prepared by: Thilina (with Claude assisting) Source: open questions in
project-prd-signals.md(as of PR #1,feature/bulk-tea-pricing-research-prototype), cross-checked against the discovery corpus (.forge/discovery/research/) andbulk-tea-pricing.md. Status: Draft for delivery-head review — question ids reference the PRD signals file.
Of the 35 open questions in the signals file, 8 are internal team tasks (Pipedrive audit) and this draft clears or converts a further 13 — leaving 14 questions from the signals file, several of which are quick confirmations rather than open discovery. To these this draft adds 5 pricing-model supplements (P1–P5) that the signals trim left unlogged but that are needed to finalize the F26 bulk-pricing schema — 19 client questions total. Everything cleared is listed in the appendix with the reason, so nothing is silently dropped.
One assumption underpinning the cuts: the team gets admin access to Empire’s Pipedrive before or at kick-off. The migration (F4) requires it regardless; several questions below are then answered from data instead of meeting time.
Part 1 — Must ask: only Empire Teas can answer these
Architecture-blocking (ask first)
Q1. External AI / data privacy (OQ-15) Are there confidentiality or data-privacy constraints on sending buyer email content to an external AI service (the Claude API)?
- Why it matters: a “no external AI” answer switches the whole AI layer to a self-hosted model (accepted risk R-PRD-002; spike SP-ARCH-001 is scoped for it). Binary decision — both paths are architected, but we need the answer to build.
- What we expect: most likely permitted with contractual safeguards (no-training/retention terms); but this must come from Empire, not assumption.
Q2. ERP landscape — current and new (OQ-2) We understand you run an ERP today and a new ERP is on the way. Three parts:
- What is the current ERP, and does it expose an API (or an import format)?
- What is the new ERP, what’s its expected timeline, and will it expose an API?
- At CRM go-live, which ERP should the CRM hand off to — and should we integrate with the current ERP first and migrate the integration to the new one later, or hold at the manual handoff until the new ERP lands?
- Why it matters: decides whether the CRM→ERP handoff on Won is automated or stays the manual document path (accepted risk R-PRD-001). V1 ships the manual path either way; the answers set the integration sequencing after that — most likely integrate-with-current-first if it has an API and remains in place long enough to be worth it.
- Follow-up we’ll need later, not at kick-off: exact handoff payload and any status sync-back (OQ-19 — deliberately deferred until the target ERP is named), asked per ERP if we end up integrating twice.
Organisation & roles
Q3. Who uses the CRM (OQ-3) Which teams/user types will actually use it? Confirm the four assumed roles — Sales Rep, Sales Manager, R&D/tea taster, Admin — and whether QA/compliance, export-documentation coordinators, or management (read-only) also need access.
- What we expect: the four roles confirmed, plus possibly an exec read-only view. The taster assumption in particular needs validating.
Q4. AI output review ownership (OQ-16) Who reviews and approves AI outputs (triaged leads, drafted replies) — one designated person, or each rep for their own records?
- What we expect: per-rep with manager oversight, but this is an org decision.
Q5. Weekly bulk-price ownership (OQ-39) Who sets and publishes the weekly bulk price — one central pricing person/role, or each rep for their own deals?
- Why it matters: shapes the F26 weekly price-update workflow and who holds the pricing permission (possibly a new role).
- What we expect: research suggests a centralized weekly publish aligned to the Colombo auction cadence is the natural fit, but Empire’s actual practice decides.
Sales process & domain specifics
Q6. Bulk price validity window (OQ-33) When you set a bulk price, how long does it hold — one week / until the next Colombo auction, ~30 days, or negotiated per deal? And is a quote’s validity the same clock or a separate one?
- Candidate answer to confirm: research points to two separate clocks — catalog price valid ~1 week (auction cadence; the prototype demonstrates effective-date + 7 days) and quote validity ~30 days (industry standard). If Empire confirms this, the F26/F7 expiry model is settled and risk R-17 closes.
Pricing-model deep-dive (supplements — see note)
The next five questions have no OQ number yet — they revive the candidates the research (
bulk-tea-pricing.md §8) surfaced but the signals trim deliberately left unlogged. The recommended schema (effective-datedproduct_pricekeyed on product × market × currency × incoterm) flexes to any answer, but the F26 spec can’t be finalized without them. If the delivery head agrees they’re in scope, they should be logged as OQ-42+ inproject-prd-signals.md.
P1. The price key — what exactly carries a price? When you set “the price of a bulk tea,” what is it attached to — a specific SKU/grade (e.g. a Uva BOP1), a broader product family, or is it set ad-hoc per deal? And within one destination market, do two buyers ever get different list prices for the same tea, or is any buyer-level difference handled as a discount at quote time?
- Why it matters: fixes the price-table key — the single most load-bearing unknown in the model. (Research candidate “OQ-B”.)
P2. Currency Are all bulk prices quoted in USD, or do some markets get EUR / GBP / other? If multi-currency: is each a separately maintained price, or a conversion of the USD number?
- Why it matters: decides whether currency is a real price-key axis (separately maintained rows) or just a display conversion. Research says bulk is typically quoted FOB Colombo in USD — confirm.
P3. Incoterm basis of the published price Is the weekly price you set an FOB Colombo number, with freight/insurance added per deal on top (the additional-cost-lines model the prototype shows), or do you maintain separate CIF prices per destination?
- Why it matters: decides whether Incoterm is a price-key axis or a quote-time cost build-up. The prototype currently assumes FOB-plus-cost-lines.
P4. Volume tiering Does the per-kg price change with order size (e.g. trial pallet vs full container)? If yes — formal tiers, or rep discretion?
- Why it matters: formal tiers would add a quantity-band dimension to the price table; discretion stays as the existing quote-time discount field. Research says volume is a real but secondary axis — confirm which mechanism.
P5. How weekly prices are maintained today When this week’s auction results come in, how does the current price actually get recorded and communicated — an Excel sheet from the broker reports, an email, someone’s head?
- Why it matters: whatever artifact exists today is both the migration seed for price history and the validation of the weekly “publish this week’s price” workflow the prototype demonstrates.
Q7. Certifications mapping (OQ-9) Which certifications apply to which products/SKUs — and how is each buyer’s required certification set tracked today (spreadsheet / CRM field / memory)?
- Why it matters: the F9 certification-mismatch check cannot be built without both sides of this mapping; the answer also identifies the migration source.
Q8. Payment terms on quotes (OQ-10) What payment terms are standard per market (LC, open account, DA, DP)? Should payment terms appear on the quote as a field, or are they handled entirely in the ERP?
- Candidate answer to confirm: a simple payment-terms field on the quote, with detailed LC administration staying in ERP/export-team scope.
Q9. Sample rounds (OQ-11) How many rounds of samples does a typical private-label deal go through before approval?
- Note: the data model supports unlimited rounds either way — the answer calibrates UX expectations and alerting, so this is a low-stakes, quick question.
Q10. Own-brand sales scope (OQ-20) Do Empire’s own-branded finished goods (Thurson / Hyson / Akina / Ceylon Wild Harvest) sell through this B2B pipeline, or only bulk + private-label?
- Why it matters: yes/no decides whether products/deals need a brand attribute. The prototype models no brands.
Channels & data completeness
Q11. Where inquiries arrive (OQ-14) Where do inquiry emails land — shared mailbox, individual rep inboxes, website form, WhatsApp?
- Why it matters: determines the trigger for automated AI inquiry extraction (deferred until known).
Q12. Mailbox setup (OQ-24, the half the audit can’t answer) Do reps send from individual mailboxes or a shared team mailbox, and on which providers (Gmail / Outlook / Exchange / IMAP)?
- Note: whether Pipedrive email sync is on and holds historical mail, we check ourselves in the audit. Exchange-on-prem/IMAP answers trigger scoped spike SP-ARCH-002.
Q13. Data tracked outside the CRM (OQ-4) What sales data still lives outside Pipedrive — Excel, WhatsApp, personal email, physical files?
- Why it matters: invisible to any audit by definition; determines whether the sample/quote/comms scope is complete or missing surfaces.
The open-ended one
Q14. How the team really uses Pipedrive (OQ-41) Walk us through a normal day in Pipedrive — and which capabilities would the team most miss?
- Why it matters: the catch-all behind the “lose nothing on migration” promise; surfaces relied-upon features the itemized audit wouldn’t think to check. Best run as a short interview/screen-share, not a questionnaire line.
Part 2 — Confirm-only: we propose, they nod (fold into the same meeting)
C1. Migration approach (OQ-21) — Proposed default: migrate everything, preserve original Pipedrive IDs for audit, and let AI triage flag junk after migration rather than dropping records before it. Confirm.
C2. Destination-market list (OQ-38) — We derive the candidate market list from Pipedrive account countries during the audit, then ask only the half that’s genuinely theirs: “which of these markets do you price differently for the same tea?” (Working list from the prototype: EU–Northern, EU–Western, Middle East, East Asia, North America, Eastern Europe.)
C3. Pricing approval authority (OQ-40) — Proposed default: quote pricing/margin approval sits with Sales Manager and above, mapped onto the existing four roles. Confirm, or name who actually holds it.
Part 3 — Appendix: questions we are NOT asking the client, and why
Answered internally via the Pipedrive audit (reclassify owner to Team)
| OQ | Topic | How we answer it instead |
|---|---|---|
| OQ-1 | PL vs bulk client mix | Count orgs/deals per pipeline in Pipedrive — more accurate than an estimate |
| OQ-5 | Leads vs deals usage | Directly observable: Leads-inbox usage, conversion patterns |
| OQ-7 | Tea grades in use | Grades visible in products/deal titles; standard Ceylon taxonomy already researched (sri-lanka-tea-industry.md §5) |
| OQ-8 | Typical PL volumes / pack formats | Read from actual deal records |
| OQ-12 | PL deal cycle length | Compute from historical deal timestamps (stale-alert thresholds stay admin-configurable) |
| OQ-13 | Bulk deal cycle length | Same as above |
| OQ-17 | Inquiry languages | Sample real inbound emails stored in Pipedrive (contingent on sync being enabled — checked in audit) |
| OQ-28 | Weighted forecast / rotting usage | Visible in Pipedrive admin settings; overlaps the OQ-26/30 audit |
Deferred — asking now would be premature
| OQ | Topic | Why deferred |
|---|---|---|
| OQ-18 | Where docs/SOPs are stored | Only gates the deferred RAG assistant — ask when that feature is scheduled |
| OQ-19 | ERP handoff payload | Blocked by OQ-2 (which ERP is the handoff target — current or new); V1 manual handoff content is already defined in the PRD |
Already internal team tasks in the signals file (unchanged)
OQ-6 (migration volume), OQ-22 (user/license count), OQ-23 (Pipedrive add-ons), OQ-25 (templates/sequences), OQ-26 (audit inventory), OQ-27 (SKU list source), OQ-29 (Goals usage), OQ-30 (pipeline/stage setup) — all resolved by the Pipedrive audit the team runs with admin access.
Next step once the delivery head signs off: reconcile this cut with Dinil’s in-progress kick-off question set (Rev 2 of the PRD notes he was finalizing one on 2026-07-21), then apply the owner/status reclassifications from Part 3 to project-prd-signals.md in a follow-up commit on his branch or a fresh one.