LMNT Renewal Sheet Plan
Foundational steps (when starting from initial, unformatted tabs)
Use these steps when the sheet or tabs are not yet in the target format. They define format and setup; the rest of this plan defines content for each column and option.
Prerequisites (at execution)
- Sheet link: Use the Google Sheet URL below (or the copy you will work in). Ensure you have edit access.
- Tabs: Identify which tab is the initial tab to copy (e.g. the week-by-week source view). For LMNT, you will end up with three option tabs (A, B, C), each with the same column structure but different scope and timelines.
Step 1. Create a copy of the initial tab
- In the same Google Sheet, duplicate the initial tab (right-click tab → Duplicate, or copy entire sheet).
- Rename the copy to something clear (e.g.
Option A – 4-column,Option B – 4-column,Option C – 4-column). If you need all three options from scratch, duplicate the initial tab three times and rename each for Option A, B, and C.
Step 2. Restructure the copy to the target columns
- Replace or reorganize columns so the left-to-right order is:
- Column A: Week (week start date; weekly or monthly cadence OK)
- Column B: Ingestion (kickoff)
- Column C: In Modelling
- Column D: QA
- Column E: Dashboard / Reporting Unlocked
- Add a header row that clearly labels these five columns.
- Keep row structure that makes sense (e.g. one row per week or month). Empty cells are fine when nothing new happens that period.
Step 3. Translate content into the columns
- Use the Column structure, Specific content changes, Data sources, and How each option differs sections below to fill each option tab.
- Use the initial (and any reference) tab only as source material; map their content into the 4-column pipeline language (ingestion kickoff → modelling → QA → dashboard unlocked). Do not copy process jargon; use the terminology table and good/bad output language below.
Step 4. Add assumptions (Overview tab or first row)
- On the Overview tab and/or as a top row on each option tab, add the Assumptions block below (modeling 2–4 weeks, dashboard = certified/QA’d, inputs = kickoff, outputs = unlocked, pre-existing ingestion note). This sets reader expectations and avoids ambiguity.
What changed (from Shivani’s feedback — Mar 11 2026 transcript)
The original sheet was too detailed and hard to scan. Shivani made a copy to redesign it. The core issues:
- Inputs vs outputs were unclear. Inputs should be “when we start ingesting” — not “ingestion complete” or when something is processed. Outputs should be “when something is actually unlocked (usable)” — which is often 4–6 weeks after input, not the same week.
- Too much process jargon. “Revenue definitions confirmed”, “Emerson POS validated”, “one order taxonomy in place” — none of this means anything to Phil. Use business language.
- Phil’s goal: By September, OKRs populated from this data and no 5 hours/month manual wrangling. 2027 starts with clean, ingested data. Less focus on OKR planning dates, more on “when are we ingesting what, and when do we unlock reporting/dashboards?”
- Modeling is the bottleneck (2–4 weeks per piece) — note this in assumptions, not in every row on the sheet.
- Workstream column: Drop it. “When I see data foundation, I don’t even know what that means.” Everyone works across workstreams anyway.
- Outcome column: Keep it minimal and high-level — dashboard locked, time-saving win, reporting capability.
- Flow: Clear diagonal — ingestion kickoff → modeling → output unlocked. Sheet reads as a pipeline, not a Gantt.
Column structure (updated)
| Col | Header | Definition |
|---|---|---|
| A | Week | Week start date (weekly or monthly cadence OK) |
| B | Ingestion (kickoff) | When we START ingesting a source (not when complete). List net-new sources only. Wording: “[Source] ingestion kickoff”, “[Source] API access”. Do NOT say “ingestion complete”. |
| C | In Modelling | What’s actively being built/modeled that week. Simple names: “Shopify V1 table modeling”, “Amazon table modeling”, “E-commerce combined modeling”. |
| D | QA | What was in Modelling the prior week, now being validated. Mirrors In Modelling with a 1–2 week lag. |
| E | Dashboard / Reporting Unlocked | What’s actually usable by Phil, Shivani, Laura, Bess. Business language only: “Revenue reporting for retail, wholesale, Shopify, Amazon”, “Finalized wholesale sales, refunds — saves Finance time at month end”. This is the checklist column — Phil checks: did we unlock this? |
Assumptions (to add on Overview tab / first row of each tab)
Add these to the Overview tab and/or as a top row on each option tab:
- Modeling takes 2–4 weeks per source/asset. Modeling is the bottleneck (not ingestion).
- “Dashboard” = certified, QA’d, and usable. Not a draft, not internal-only.
- Inputs = when ingestion starts (kickoff), not when it’s complete or when the connector is built.
- Outputs = when the deliverable is actually unlocked, often 4–6 weeks after the ingestion kickoff.
- Shopify, wholesale, retail/Emerson, Facebook, Google Ads, GA4 already ingested before Mar 10 (pre-existing; don’t list as new kickoffs).
Specific content changes to make
1. Inputs column — what to list and when
Only list net-new sources. Wording = “ingestion kickoff” or simple source name. Do NOT say “ingestion complete”.
| Source | When (approximate) | Notes |
|---|---|---|
| Amazon Seller Central | Mar 10 | Connector kicked off this week via Polytomic |
| Walmart.com (D2C) | Mar 17 | Separate from Walmart 3P; connector expected this week |
| Spins | Apr 1–Jun 1 window | ”Spins API access” — access available any time Apr 1–Jun 1 (per Spins email); kickoff when access granted |
| DSD from Encompass | ~May 5 | Dependency: Jeff Warren completing Dimensions cutover; replaces QuickBooks |
| Confido | ~May 19–Jun 2 | Unlocks retail revenue (net of chargebacks + trade spend); without Confido, only retail sales |
| Non-Emerson retail (Vitamin Shoppe, Costco FTP) | ~Jul 7 | Option A+; broadens retail coverage |
| ✚ Supply chain (Ware2Go/Geodis) | Jun 9 (B) / May 5 (C) | Option B starts Jun 9, Option C starts May 5 (1.5 months earlier) |
Note on Walmart: Split “Walmart 3P” (third-party fulfillment) and “Walmart.com” (D2C channel). Both should appear as separate ingestion items.
2. Outputs column — what to use
Good output language (use these or variants):
- “Shopify tables V1 (orders, customers, transactions)”
- “Finalized wholesale sales, refunds — saves Finance time at month end”
- “Revenue reporting for retail, wholesale, Shopify, Amazon” (with target date, e.g. ~Apr 14)
- “Retail VP dashboard live (Walmart vs Target, product/category)”
- “VP-level e-commerce dashboard drafted”
- “AI querying available in Omni (pilot)”
- “E-commerce combined view — Shopify + Amazon + Walmart.com”
- “Revenue reporting including Confido net revenue (once Confido in)” ← dependency note
Revenue dependency note: Until Confido is ingested, LMNT has retail sales but not full retail revenue (net of chargebacks, trade spend). Show this on the sheet as: “Full retail revenue reporting (dependency: Confido ingested)”.
Bad output language (avoid):
- “Amazon settlements modeling” — this is In Modelling, not an output
- “Retail Comparison Model QA” — process jargon
- “Revenue definitions confirmed” — too internal
- “E-commerce point-of-sale validated” — word salad
- “Wholesale partner view available” — vague
- “One order taxonomy in place” — process detail, not a win
3. Structure rules
- Drop workstream column entirely. Don’t add it back.
- Checklist orientation: Every row in the Dashboard column = a thing Phil can check off. “Did we unlock this by this date? If not, why?”
- Diagonal flow: Each row’s QA column should contain what was in Modelling the prior week. Each row’s Dashboard column should contain outputs that trace back 4–6 weeks to an Ingestion kickoff.
- Monthly vs weekly: Fine to use weekly rows; not every cell needs to be filled. Empty cells = nothing new that week.
4. Terminology
| Use | Avoid |
|---|---|
| mart, table, model, dashboard | ”mart” is fine; avoid “fact table”, “staging”, “raw” for Phil-facing language |
| ”ingestion kickoff" | "ingestion complete”, “data available”, “connector ready" |
| "Revenue reporting for [channels]" | "revenue definitions confirmed”, “net revenue definition finalized" |
| "Omni self-service live" | "Omni topics built”, “governed dashboards in Omni" |
| "saves Finance time at month end" | "Bess has reconciliation report” (who is Bess? Phil doesn’t know) |
Data sources to list (P0–P2)
All options must show all P0–P2 sources in the Ingestion column on the correct week. If a source was already ingested before the engagement start, note it in the assumptions row.
Already ingested (pre-Mar 10): Shopify orders/customers/transactions, wholesale (Emerson), retail/Emerson POS, Facebook Ads, Google Ads, GA4.
To be ingested during engagement (mark in Ingestion column):
- Amazon Seller Central (Mar 10)
- Walmart.com / D2C (Mar 17)
- Walmart 3P (already confirmed via Polytomic — note alongside Walmart.com)
- Spins (Apr 1–Jun 1 access window)
- Marketing costs: Google Ads, Facebook Ads, LinkedIn — some already done; note remaining
- DSD from Encompass (~May 5)
- Confido (~May 19–Jun 2)
- Non-Emerson retail — Vitamin Shoppe, Costco FTP (~Jul 7, Options A+)
- ✚ Supply chain — Ware2Go/Geodis (Jun 9 for B; May 5 for C)
How each option differs
| Option A ($68K) | Option B ($92.6K) | Option C ($107.4K) | |
|---|---|---|---|
| Supply chain | Scope/assessment only (no ingestion) | Ingestion kickoff Jun 9 | Ingestion kickoff May 5 (1.5 mo earlier) |
| Dashboards live | Apr 21 (Retail VP) | Mar 31 (Retail + Wholesale) | Mar 31 (Retail + Wholesale, parallel) |
| Revenue reporting | ~Apr 14 | ~Mar 31 | ~Mar 24 |
| Full e-comm view | ~Jun 2 | ~May 26 | ~May 19 |
| 2027 planning ready | Sep 1 | Aug 25 | Jul 28 (1 mo earlier) |
| Sep weeks | Active delivery | Active delivery | Buffer / refinement |
How the sheet should be used
Checklist mode (weekly or monthly check-in):
- “Did we start ingesting this? If not, why — was it delayed, access not granted, blocked?”
- “Did we unlock this output by this date? If not, what moved?”
- Weekly: note what moved, what was added, what was delayed (e.g. “Spins took another week — no access yet”)
This is NOT a Gantt chart. It’s a pipeline view: ingestion → modeling → output unlocked. Phil sees whether the supply chain of data is moving.
First page / Overview tab content
Add to the Overview tab (or as Row 1 on each option tab):
Assumptions:
- Modeling takes 2–4 weeks per source. Modeling is the bottleneck — not ingestion.
- "Dashboard" = certified, QA'd, and usable. Not a draft.
- Inputs = when ingestion kicks off (starts), not when it's complete.
- Outputs = when something is actually unlocked and usable (typically 4–6 weeks after ingestion kickoff).
- Pre-existing ingestion (before Mar 10): Shopify, wholesale/Emerson, retail/Emerson POS, Facebook Ads, Google Ads, GA4.
Links
- Format and initial steps: The foundational steps above are aligned with the sheet 4-column translation plan (copy tab → create columns → translate content → assumptions). Use that plan for generic format/setup; this plan for LMNT-specific content.
- Sheet: https://docs.google.com/spreadsheets/d/1cZajTRo5h5tABRWG2C78eaJnoq86MfVsYb3jtQigYA8
- Option A: https://docs.google.com/spreadsheets/d/1cZajTRo5h5tABRWG2C78eaJnoq86MfVsYb3jtQigYA8/edit?gid=600118241
- Option B: https://docs.google.com/spreadsheets/d/1cZajTRo5h5tABRWG2C78eaJnoq86MfVsYb3jtQigYA8/edit?gid=392897337
- Option C: https://docs.google.com/spreadsheets/d/1cZajTRo5h5tABRWG2C78eaJnoq86MfVsYb3jtQigYA8/edit?gid=2041456063
- Renewal plan doc: https://docs.google.com/document/d/13WI5kigYYmJhRQ5pbivk8XCLPpR2aS2KYBeBLltBV8w/edit
- Transcript (Shivani feedback): store in vault or Drive (e.g.
knowledge/clients/magicspoon/transcripts/); do not commit local Downloads paths - Vault resources: knowledge/clients/magicspoon/resources/
Out of scope
- Changing the original initial or reference tabs
- Building automation; translation is manual or one-off edits
- Creating a separate doc; views live as three tabs in the existing sheet