Brainforge Knowledge

walmart_emerson_snowflake

15 min read

Walmart 3P data source: data architecture and profiling memo

Audience: LMNT Leadership, Analytics, Engineering

Source: Emerson/Walmart 3P Dataset (LMNT_425 schema)

Date: 2025-12-15

1. Executive summary

What this is

Emerson Healthcare serves as LMNT’s data provider for Walmart retail performance. They aggregate Walmart’s 3P (third-party seller) data and make it available through a Snowflake Private Share, giving LMNT access to:

  • 26 LMNT product SKUs in Walmart’s catalog
  • 4,577 stores with active POS sales
  • $56M in combined revenue (POS + Digital) from March-December 2025
  • 3.35M+ units sold across brick-and-mortar and omnichannel
  • Distribution center logistics, store traits, and merchandising attributes

This is one of LMNT’s most comprehensive retail datasets, offering store-level sales visibility and omnichannel performance tracking that isn’t typically available from retailers.

Why this matters now

1. Walmart is a major revenue channel

With $56M in nine months, Walmart represents a significant portion of LMNT’s retail business. This data enables analysis that’s impossible with aggregated reporting.

2. Expansion opportunity

LMNT is in 4,577 of Walmart’s 6,192 stores (74% penetration). That’s 1,600+ stores where LMNT could expand.

3. Emerson relationship is evolving

Based on recent discussions, LMNT may be transitioning administrative responsibilities away from Emerson. Understanding what data we get, how we get it, and what alternatives exist is critical for continuity.

4. Store-level insights are rare

Most retailers provide aggregated reporting. This dataset enables:

  • Geographic performance analysis (which states/stores drive revenue)
  • Store segmentation (format, traits, demographics)
  • Inventory and distribution diagnostics
  • Digital vs. in-store channel analysis

Key findings

Strengths:

  • Strong distribution: LMNT is in 74% of Walmart stores nationwide
  • Balanced channels: POS (26.6M) are nearly equal
  • Recent launch: All items created Dec 2024-Jul 2025 (launch phase data)

Opportunities:

  • 9 non-selling SKUs: 26 items in catalog, only 17 actively selling
  • State concentration: Top five states account for majority of revenue
  • Data dependency: Currently reliant on Emerson’s Snowflake share

What you’ll find in this memo

  • Section 2: Understanding the Emerson data relationship and ETL considerations
  • Section 3: Table-by-table profiling with business interpretation
  • Section 4: Cross-table insights (distribution, assortment, sales performance)
  • Section 5: Deep-dive business analysis (state performance, Target comparison)
  • Section 6: Recommendations and next steps

2. Understanding the Emerson data relationship

How LMNT gets Walmart data today

Emerson Healthcare acts as an intermediary between LMNT and Walmart:

  1. Walmart provides 3P seller data through their Luminate platform
  2. Emerson ingests this data into their Snowflake environment
  3. LMNT accesses the data via Snowflake Private Share (zero-ETL replication)

What is Snowflake Private Share?

This is a unique Snowflake feature where data is shared directly between Snowflake accounts without copying or ETL:

  • No data movement: Tables appear in LMNT’s Snowflake as if they’re native
  • No latency: Updates from Emerson are instant
  • No ETL cost: No rows are counted/charged for ingestion
  • Governance: Emerson controls access; LMNT has read-only permissions

Key implication: This is why we recommended Snowflake as a warehouse option in the ETL assessment. Private Share is a significant advantage for this data source.

What happens if LMNT moves away from Emerson?

Based on recent discussions, LMNT may transition administrative responsibilities. Here’s what we need to understand:

Questions to answer with Phil/team

  1. Will LMNT still have access to Walmart data through Emerson’s Snowflake share?
  2. Does LMNT have direct access to Walmart Luminate, or is Emerson required?
  3. If we lose Emerson access, what are the alternatives?
    • Direct Walmart Luminate API integration (via Fivetran/Polytomic)
    • Another 3P data aggregator
    • Build custom ingestion

Risk assessment

  • Low risk: Emerson continues data sharing even if administrative relationship changes
  • Medium risk: Need to build Walmart Luminate connector (Fivetran/Polytomic supported)
  • High risk: Lose historical data access (need to export/backup now)

Recommendation: We should export critical historical data to LMNT-owned storage as a backup, regardless of relationship continuity.

3. Table-level documentation and profiling

Each section includes:

  • Metrics summary table
  • Business description
  • Profiling interpretation

All results reflect actual queries executed against the LMNT_425 schema.

3.1 WALMART_ITEM_ATTRIBUTES

Summary metrics

MetricValue
Total rows26
Distinct items (WM_ITEM_NBR)26
Distinct UPCs (WM_UPC_NBR)16
Distinct vendors (VENDOR_NBR)1
Item creation date range2024-12-16 to 2025-07-11

Business description

Master catalog of Walmart items sold by LMNT.

Contains detailed product attributes including vendor data, UPCs, packaging dimensions, logistics flags, merchandising text, and lifecycle dates.

Interpretation

  • Full catalog includes 26 LMNT items
  • Only 16 unique UPCs, indicating pack variations or cross-linked UPCs
  • Single vendor present (EMERSON HEALTHCARE LLC)
  • Items are newly launched between late 2024 and mid 2025

3.2 WALMART_STORESALES

Summary metrics

MetricValue
Total rows2,257,319
Distinct stores4,577
Distinct items sold17
Total units sold3,352,022
Total sales$29,435,812.51
Date range2025-03-05 to 2025-12-01

Business description

Store-level POS transactions capturing sales amount, units, item identifiers, store identifiers, and Walmart fiscal attributes.

Interpretation

  • LMNT sells in ~4,600 stores nationwide
  • 17 items actively selling in brick-and-mortar
  • Strong sales velocity with over 3.35M units and $29.4M in POS revenue
  • Dataset covers ~9 months of sales (aligned with program launch)

3.3 WALMART_OMNISALES

Summary metrics

MetricValue
Total rows8,251
Distinct items17
Distinct service channels5
Distinct order channels2
Total GMV$26,625,400.03
Total net sales$26,625,400.03
Date range2025-03-05 to 2025-12-01

Business description

Daily ecommerce and omnichannel sales, including GMV, net sales, units, and service/order channel breakdowns.

Interpretation

  • Digital sales nearly match POS with $26.6M GMV
  • Same 17 items active online and in-store (unified assortment strategy)
  • Coverage across five service channels indicates deep digital penetration
  • Time range matches POS, confirming consistent ingestion

3.4 WALMART_STOREITEM

Summary metrics

MetricValue
Total rows5,516,563
Distinct stores4,584
Distinct items16

Business description

Store-item mapping table showing item availability, stocking flags, and merchandising attributes for each store.

Interpretation

  • LMNT items appear in ~4,584 stores, matching POS distribution
  • 16 unique items in availability mapping (vs. 17 selling)
  • Used for availability, distribution voids, and replenishment diagnostics

3.5 WALMART_STORETRAIT

Summary metrics

MetricValue
Total trait records30,954,234
Distinct stores4,956

Business description

Store trait assignments describing operational, geographic, service, and format characteristics for Walmart stores.

Interpretation

  • Coverage exceeds LMNT’s selling footprint (~4,580 stores)
  • Useful for segmentation, modeling, and identifying expansion opportunities
  • High record volume reflects many-to-many trait assignments

3.6 WALMART_STOREDIMENSIONS

Summary metrics

MetricValue
Total stores6,192
Distinct STORE_NBR6,192

Business description

Walmart’s complete store master with location, format, region, and operational attributes.

Interpretation

  • LMNT distributes to ~4,580 of Walmart’s 6,192 stores
  • Remaining ~1,600 stores represent expansion potential
  • Essential for accurate store segmentation and geographic rollups

3.7 WALMART_DCITEMS

Summary metrics

MetricValue
Total DC-item records219,253
Distinct WM items26
Distinct DCs47

Business description

Distribution center inventory dataset detailing vendor pack quantities, wholesale pack quantities, costs, shipping/receiving activity, and DC-level coverage.

Interpretation

  • All 26 LMNT catalog items appear in the DC network
  • 47 distribution centers provide strong national supply chain coverage
  • Supports DC operations, replenishment modeling, and inventory analytics

3.8 WALMART_CALENDAR

Summary metrics

MetricValue
Total calendar records21,553
Date range1990-12-29 to 2049-12-31

Business description

Walmart’s full fiscal, retail, and Gregorian calendar mapping, including LY/YOY comparison fields.

Interpretation

  • Covers 60 years of dates, enabling long-term YOY analysis
  • Required for aligning sales to Walmart fiscal periods
  • Supports forecasting models and future-year dashboards

4. Cross-table insights and strategic observations

This section synthesizes findings across all tables to answer key business questions.

4.1 Distribution footprint analysis

TableStoresCoverage
Total Walmart fleet (STOREDIMENSIONS)6,192100%
Stores with traits (STORETRAIT)4,95680%
Stores with store-item mapping4,58474%
Stores with POS sales4,57774%

What this tells us

LMNT has achieved strong national distribution. Present in nearly three out of four Walmart stores within nine months of launch.

Distribution and sales align tightly. 4,584 stores have LMNT on shelves, and 4,577 are actively selling. This suggests minimal “distribution voids” (stores with product but zero sales).

Expansion opportunity: ~1,600 stores. LMNT could potentially expand into the remaining 26% of Walmart’s footprint. These stores likely fall into:

  • Smaller format stores (Neighborhood Markets)
  • Geographic markets with lower CPG velocity
  • Stores not yet approved for LMNT’s category

Recommendation: Analyze store traits of the 1,600 non-selling stores to identify which are viable expansion targets vs. poor-fit formats.

4.2 Assortment and SKU performance

MetricValue
Total catalog items26
Items with store sales17
Items with digital sales17
Items in DC supply chain26

What this tells us

Unified omnichannel assortment. The same 17 SKUs sell in-store and online, indicating a consistent customer experience.

Full supply chain readiness. All 26 catalog items are set up in Walmart’s DC network, even if only 17 are actively selling.

9 SKUs are “zombie inventory.” These items exist in the catalog and DC system but have zero sales (see detailed table in Section 5.1). Reasons could include:

  • Items marked DELETE status (six of nine were deleted in May 2025)
  • Never launched to stores
  • Failed product tests
  • Seasonal/promotional items not yet activated

Recommendation: Review the nine non-selling SKUs with merchandising team. If not part of a future launch plan, remove from catalog to clean up reporting.

4.3 Channel mix: POS vs. digital

ChannelRevenue% of Total
POS (in-store)$29.4M53%
Digital (omni)$26.6M47%
Combined$56M100%

What this tells us

Walmart is a major revenue driver. 75M annualized from a single retailer.

Digital is nearly as important as in-store. Most CPG brands see 10-20% digital mix; LMNT’s 47% is exceptional and likely driven by:

  • Health-conscious customers shopping online
  • Subscription/repeat purchase behavior
  • Walmart+ membership penetration

Omnichannel strategy is working. Customers can discover LMNT in-store and reorder online (or vice versa).

Implication for analytics: Any Walmart reporting must include both POS and digital to accurately measure total performance. Analyzing only one channel would miss nearly half the picture.

5. Deep-dive business analysis

This section provides actionable insights based on profiling results.

5.1 Non-selling SKUs analysis

We identified nine items in the catalog for which no sales transactions exist:

Item nameStatusUPCObsolete dateObservations
LMNT ZS 4 GRF S 16OZDELETE810183490382025-05-13Grapefruit 4-pack can deleted May 2025
LMNT ZS ORNG ST 6CTDELETE810183490272025-05-13Orange stick 6-pack deleted May 2025
LMNT ZS RASBY ST 6CTDELETE810183490282025-05-13Raspberry stick 6-pack deleted May 2025
LMNT ZS 1 CTR S 16OZDELETE850055993102025-05-13Citrus single can deleted May 2025
LMNT ZS 4 BC LM 16OZDELETE810183490352025-05-13Black cherry lime 4-pack deleted May 2025
LMNT ZS 4 CTR S 16OZDELETE810183490362025-05-13Citrus 4-pack can deleted May 2025
LMNT ZS 4 WTR S 16OZACTIVE810183490372500-01-01Watermelon 4-pack (ONLY ACTIVE NON-SELLER)
LMNT ZS CTRS ST 6CTDELETE810183490252025-05-13Citrus stick 6-pack deleted May 2025
LMNT ZS GRPFT ST 6CTDELETE810183490292025-05-13Grapefruit stick 6-pack deleted May 2025

Key observations

1. Eight of nine were deleted in May 2025. These appear to be a product rationalization or test that didn’t perform.

2. One ACTIVE item with zero sales. LMNT ZS 4 WTR S 16OZ (Watermelon 4-pack) is still marked ACTIVE but has never sold. This could mean:

  • Item was just added to the catalog (late 2024) but not yet distributed
  • Test SKU waiting for launch
  • Data lag or missing transactions

3. Mix of sticks and cans. Both format types appear in non-sellers, suggesting the issue isn’t format-specific.

Action items

  • Confirm with merchandising: Is Watermelon 4-pack (674256756) planned for launch?
  • If not, mark as DELETE to clean up reporting
  • Remove deleted SKUs from future catalog loads to reduce noise in analytics

5.2 State-level sales performance

Revenue and units by month (last three months, pivoted):

  • m0 = current month-to-date (December 2025, partial)
  • m1 = last full month (November 2025)
  • m2 = month before last (October 2025)
StateRevenue M0Units M0Revenue M1Units M1Revenue M2Units M2
Texas$108,49314,067$382,85350,627$469,63061,645
Florida$68,6297,697$225,42725,365$260,48229,189
California$43,0554,832$138,69715,272$158,65217,222
North Carolina$35,0463,946$117,24613,084$134,97115,133
Georgia$34,0313,810$114,70212,860$131,35514,701
Oklahoma$33,6723,759$116,98713,069$144,29316,223
Tennessee$33,0903,697$112,59712,564$130,90614,657
Missouri$30,7243,428$106,46911,911$122,69813,805
Arizona$30,6883,447$106,58711,914$121,37813,650
Arkansas$30,3463,384$103,03811,496$125,22914,004
(remaining 40 states omitted for brevity)………………

Top 10 states account for $1.9M per month

From the data above (using m1/November as baseline):

  • Top five states: TX, FL, CA, NC, GA = $1.13M (46% of November total)
  • Top 10 states: + OK, TN, MO, AZ, AR = $1.87M (76% of November total)

Geographic insights

State concentration

Top five states (TX, FL, CA, NC, GA) account for nearly half of total Walmart revenue. Smaller states show consistent but low velocity, suggesting distribution exists but velocity per store is low.

Implication: Growth upside likely comes more from deepening velocity in top states than expanding footprint into new ones.

Under-penetrated opportunity states

States with decent size but lower relative revenue:

  • New York: $60K/month (population rank #4 nationally)
  • Washington: $65K/month (population rank #13)
  • Massachusetts: $31K/month (population rank #16)
  • New Jersey: $27K/month (population rank #11)

These states may have:

  • Lower Walmart density (stronger regional competitors like Wegmans, H-E-B)
  • Demographic mismatch with LMNT’s target customer
  • Distribution gaps (not in enough stores yet)

Recommendation: Cross-reference store count by state with revenue per store to identify:

  1. States with good distribution but low velocity (need marketing/demos)
  2. States with low distribution but high velocity per store (expand presence)

5.3 Comparative analysis: Walmart vs. Target

We also profiled Target’s data (separate data source). Here’s how Walmart compares:

Target monthly sales (2025)

MonthRevenueUnits
Oct 2025$4.58M397K
Nov 2025$4.07M355K
Dec 2025 (MTD)$1.16M101K

Walmart monthly sales (estimated from 9-month total)

MetricWalmartTarget
Avg monthly revenue$6.2M$4.3M
Avg monthly units650K380K

Key takeaway: Walmart is outperforming Target by approximately 45% in revenue and 70% in unit volume.

Why the difference?

  1. Store count: Walmart has more stores where LMNT is distributed
  2. Customer base: Walmart’s broader demographic may align better with LMNT
  3. Digital mix: Walmart’s 47% digital vs. Target’s 20% digital (approximate)

Target channel split

October 2025:

  • Store (STR): $3.67M | 323K units (~80% of revenue)
  • Online/fulfilled (FF): $0.88M | 72K units (~19%)
  • FC: ~$28K | ~2K units (<1%)

November 2025:

  • Store (STR): $3.27M | 291K units (~80%)
  • FF: $0.77M | 61K units (~19%)
  • FC: ~$28K | ~2K units (<1%)

Observation: Target’s digital mix (~20%) is less than half of Walmart’s (~47%). This suggests:

  • Walmart+ is driving more digital adoption
  • LMNT may benefit from Walmart’s stronger e-commerce platform
  • Target customers prefer in-store discovery

6. Recommendations and next steps

Based on this profiling analysis, here are our recommended actions:

6.1 Immediate actions (this week)

1. Clarify Emerson relationship

Owner: Phil/LMNT Leadership

Action: Confirm future state of Emerson data access

  • Will Snowflake share continue?
  • Do we need to prepare for alternative ingestion?

Why it matters: Losing access mid-project would derail analytics roadmap.

2. Export historical data as backup

Owner: Brainforge Data Team

Action: Export all nine months of Walmart data to LMNT-owned storage (S3/Snowflake)

Why it matters: Even if Emerson continues sharing, we should own our historical data.

3. Clean up non-selling SKUs

Owner: LMNT Merchandising + Brainforge

Action:

  • Confirm Watermelon 4-pack (674256756) launch plan
  • Remove DELETE-status SKUs from future reporting models

Why it matters: Reduces noise in dashboards and prevents confusion about assortment.

6.2 Short-term initiatives (next two weeks)

4. Build store-level performance dashboard

Owner: Brainforge Data Team

Deliverables:

  • State-level revenue trends (m/m, YoY)
  • Top 100 stores by revenue
  • Distribution void analysis (stores with product but zero sales)

Why it matters: Enables merchandising and sales teams to identify underperforming markets.

5. Analyze expansion opportunity

Owner: Brainforge + LMNT Merchandising

Action:

  • Profile the 1,600 stores where LMNT is not present
  • Use WALMART_STORETRAIT to segment by format, region, demographics
  • Identify top 200 “best fit” stores for expansion

Why it matters: Provides data-driven roadmap for growing Walmart distribution from 74% to 85%+.

6. Digital vs. in-store attribution analysis

Owner: Brainforge Data Team

Action:

  • Break down OMNISALES by service channel (five channels identified)
  • Identify which digital channels drive most revenue
  • Analyze customer behavior (first purchase in-store, reorder online?)

Why it matters: LMNT’s 47% digital mix is exceptional. Understanding what’s driving it can inform strategy at other retailers.

6.3 Medium-term initiatives (next 30 days)

7. Compare Walmart vs. Target performance

Owner: Brainforge Data Team

Deliverables:

  • Unified dashboard showing both retailers side-by-side
  • SKU-level comparison (which products sell better where?)
  • Channel mix comparison (digital vs. in-store)

Why it matters: Informs which retailer to prioritize for future launches and promotional investment.

8. Build promotion/event tracking

Owner: Brainforge + LMNT Marketing

Action:

  • Tag promotional periods in calendar
  • Measure lift from in-store demos, ads, price reductions
  • Create “promotion effectiveness dashboard”

Why it matters: CPG brands live and die by promotional performance. We need to measure what works.

9. State-level velocity analysis

Owner: Brainforge Data Team

Action:

  • Calculate revenue per store by state
  • Identify “high velocity” states (e.g., Texas) vs. “low velocity” states (e.g., New York)
  • Recommend either:
    • Marketing investment (to boost velocity in existing stores)
    • Distribution expansion (to add more stores in high-velocity states)

Why it matters: Provides clear priorities for growth strategy.

6.4 Data quality and governance

10. Define data refresh cadence

Owner: Brainforge + Emerson

Action:

  • Document how often Emerson updates the Snowflake share
  • Set expectations for data freshness (daily, weekly?)
  • Create alerting for data delays

Why it matters: Stakeholders need to know when data is “current” vs. “stale.”

11. Document data lineage

Owner: Brainforge Data Team

Action:

  • Create data dictionary for all Walmart tables
  • Document field definitions, grain, update frequency
  • Share with analytics team for self-service

Why it matters: Enables LMNT team to explore data independently without Brainforge bottleneck.

7. Appendix: technical details

7.1 Snowflake Private Share details

  • Share name: LMNT_425 (Emerson-provided schema)
  • Access method: Snowflake Private Share (zero-ETL)
  • Refresh cadence: TBD (needs confirmation with Emerson)
  • Data retention: Full history since March 2025 launch

7.2 Table relationships

WALMART_ITEM_ATTRIBUTES (26 items)
    -> WALMART_STORESALES (4,577 stores, 17 items)
    -> WALMART_OMNISALES (8,251 daily records, 17 items)
    -> WALMART_STOREITEM (5.5M store-item records)
    -> WALMART_DCITEMS (219K DC-item records)

WALMART_STOREDIMENSIONS (6,192 stores)
    -> WALMART_STORETRAIT (31M trait assignments)
    -> WALMART_STORESALES (join on STORE_NBR)

WALMART_CALENDAR (21K date records)
    -> WALMART_STORESALES (join on fiscal dates)

7.3 Data volume estimates

TableRowsApprox sizeMonthly growth
WALMART_STORESALES2.3M~500 MB+250K rows/month
WALMART_OMNISALES8K~2 MB+800 rows/month
WALMART_STOREITEM5.5M~1 GBRelatively static
WALMART_STORETRAIT31M~2 GBStatic (reference)
WALMART_STOREDIMENSIONS6K<1 MBStatic (reference)

Total dataset: ~3.5 GB (including all reference tables)

8. Questions and next steps

If you have questions about this memo, please reach out:

Brainforge Data Team:

Key decisions needed:

  1. Emerson relationship clarification (Phil)
  2. Watermelon 4-pack launch plan (Merchandising)
  3. Data refresh expectations (Emerson + Phil)

Next deliverable: Store-level performance dashboard (Week of 2025-12-16)

Document version: 1.0 Last updated: 2025-12-15 Authors: Brainforge Data Team (Uttam Kumaran, Awaish Kumar)

Graph View