Market Share Analyzer — Competitive Intelligence from Sold Data

Convert MarketCheck sold transaction data into real-time market share analytics. Track brand and model-level share, segment conquest patterns, dealer group performance, EV adoption curves, and regional demand distribution — all without waiting 60-90 days for traditional syndicated reports.

Dealer Profile (Load First — Optional Context)

→ Full procedure: read _references/profile-loading.md

Parse marketcheck-profile.md if it exists → extract: state, franchise_brands, dealer_type, country. This skill works fine without a profile.

US-only: Requires get_sold_summary. If country == UK: inform "Market share analysis requires US sold transaction data and is not available for the UK market." → Full limitations: _references/country-routing.md

Confirm: "Using profile context: [state], [franchise_brands]"

User Context

Before running any workflow, collect the following (auto-filled from dealer profile where available):

• Role: OEM/manufacturer analyst, dealer group strategist, or market researcher
• Geographic scope: From profile location.state if user says "my market", otherwise ask. National (omit state), single state (2-letter code), or multi-state (run each separately)
• Time period: Specific month(s) for analysis. Always use first-of-month to last-of-month format (e.g., 2026-01-01 to 2026-01-31). If the user asks for "quarterly" data, run three consecutive months and aggregate.
• Comparison period: Prior month, prior quarter, or year-over-year month (for share change calculation)
• Brand focus (optional): From profile dealer.franchise_brands if available, otherwise ask
• Segment focus (optional): body_type (SUV, Sedan, Pickup, Hatchback, Coupe, Convertible, Van/Minivan, Wagon) or fuel_type_category (EV, Hybrid, ICE)
• Inventory type: New, Used, or Both (default Both)

If the user asks for "market share" without specifying a geographic scope, default to national and confirm.

Gotchas

• US-only — this skill requires get_sold_summary which is not available for UK. Inform UK users immediately.
• Share change must be in basis points (bps) — a move from 14.2% to 14.5% is "+30 bps", not "+0.3%". Flag makes gaining or losing > 50 bps.
• Quarterly data requires 3 separate monthly calls aggregated — get_sold_summary works on monthly date ranges. For "Q1" analysis, call January, February, and March separately, then sum volumes and recalculate shares.
• National vs state scope — if no state is specified, omit the state param for national data. Don't pass state=US or similar.
• top_n limits ranking results — set appropriately per workflow. For brand share use top_n=20, for segment conquest use top_n=15, for dealer group benchmarking use top_n=20.
• Always show both volume AND share % — raw counts without context are meaningless; percentages without counts lack scale.

Workflow: Brand Market Share

Calculate market share by make for a given period and compare against a prior period to identify gainers and losers.

1. Call mcp__marketcheck__get_sold_summary for the current period:

   • date_from: first of target month (e.g. 2026-01-01)
   • date_to: last of target month (e.g. 2026-01-31)
   • state: user's state filter (omit for national)
   • inventory_type: as specified (or omit for both)
   • ranking_dimensions: make
   • ranking_measure: sold_count
   • ranking_order: desc
   • top_n: 20

2. Call mcp__marketcheck__get_sold_summary for the prior period with identical filters but adjusted date_from / date_to (e.g., prior month or same month prior year).

3. Calculate for each make:

   • Current Share % = Make Sold Count / Total Sold Count x 100
   • Prior Share % = same calculation for prior period
   • Share Change (bps) = (Current Share % - Prior Share %) x 100 basis points
   • Volume Change % = (Current Sold - Prior Sold) / Prior Sold x 100

4. Present as a ranked table:

   • Columns: Rank, Make, Current Sold, Current Share %, Prior Sold, Prior Share %, Share Change (bps), Volume Change %
   • Sort by Current Share % descending
   • Highlight the user's brand of interest in bold
   • Flag makes gaining > 50 bps as "GAINING" and losing > 50 bps as "LOSING"

5. Add a summary paragraph: "The top 3 share gainers this period were [X], [Y], [Z], collectively picking up [N] basis points. The biggest losers were [A], [B], [C]. [User's brand] moved from #X to #Y position with a [+/-N] bps shift."

Workflow: Segment Conquest Analysis

Determine which brands are winning within specific vehicle segments (body types) and identify conquest opportunities.

1. Call mcp__marketcheck__get_sold_summary with:

   • date_from / date_to: target period
   • state: user's state filter (omit for national)
   • body_type: user's target segment (e.g. SUV)
   • ranking_dimensions: make,model
   • ranking_measure: sold_count
   • ranking_order: desc
   • top_n: 15

2. Repeat for the comparison period to calculate share change within the segment.

3. If the user wants multi-segment comparison, repeat step 1 for each body_type: SUV, Sedan, Pickup, Hatchback, Coupe, Van/Minivan.

4. For each segment, calculate:

   • Segment leader and their share %
   • User's brand rank within the segment
   • Gap to leader in units and share points
   • Fastest gaining model in the segment (largest positive bps change)

5. Present per-segment tables:

   • Columns: Rank, Make, Model, Sold Count, Segment Share %, Prior Share %, Change (bps)
   • Also present a Segment Summary table: Segment, Leader, Leader Share %, User Brand Rank, User Brand Share %, Gap to Leader

6. Conquest insight: "In the SUV segment, [Brand A] gained 120 bps primarily through [Model X] (+3,200 units). [User's brand] lost share to [Brand A] and [Brand B]. To recapture, focus on [Model Y] which competes directly and currently has lower DOM."

Workflow: Dealer Group Benchmarking

Rank dealer groups by sales volume and operational efficiency to identify top performers and laggards.

1. Call mcp__marketcheck__get_sold_summary with:

   • date_from / date_to: target period
   • state: user's state filter (omit for national)
   • ranking_dimensions: dealership_group_name
   • ranking_measure: sold_count
   • ranking_order: desc
   • top_n: 20

2. Call mcp__marketcheck__get_sold_summary with same filters but:

   • ranking_measure: average_days_on_market
   • ranking_order: asc
   • top_n: 20

3. Call mcp__marketcheck__get_sold_summary with same filters but:

   • ranking_measure: average_sale_price
   • ranking_order: desc
   • top_n: 20

4. Merge the three result sets by dealership_group_name. Build a Dealer Group Leaderboard:

   • Columns: Rank (by volume), Dealer Group, Sold Count, Market Share %, Avg DOM, Avg Sale Price, Efficiency Score
   • Efficiency Score = Sold Count / Avg DOM (higher is better — moves more units faster)

5. If the user specifies a make, add a make filter to all calls to see dealer group performance within a single brand's network.

6. Provide analysis: "AutoNation leads in volume with X units (Y% share) but Lithia has the lowest average DOM at Z days, suggesting faster inventory turns. For [Brand] specifically, the top 3 performing groups are..."

Workflow: EV Adoption Tracking

Monitor electric and hybrid vehicle penetration rates over time against the total market.

1. Call mcp__marketcheck__get_sold_summary for EV sales:

   • date_from / date_to: target period
   • state: user's state filter (omit for national)
   • fuel_type_category: EV
   • ranking_dimensions: make,model
   • ranking_measure: sold_count
   • ranking_order: desc
   • top_n: 15

2. Call mcp__marketcheck__get_sold_summary for Hybrid sales:

   • Same filters but fuel_type_category: Hybrid

3. Call mcp__marketcheck__get_sold_summary for total market (no fuel_type_category filter):

   • ranking_dimensions: make
   • ranking_measure: sold_count
   • top_n: 1 (we just need the total count)

4. Repeat steps 1-3 for the prior period (prior month or same month last year) to calculate trend.

5. Calculate:

   • EV Penetration Rate = EV Sold / Total Sold x 100
   • Hybrid Penetration Rate = Hybrid Sold / Total Sold x 100
   • Combined Electrified Rate = (EV + Hybrid) / Total x 100
   • Period-over-period change for each rate

6. Present:

   • EV Penetration Summary: "EVs represented X.X% of [state/national] sales in [month], [up/down] from Y.Y% in [comparison month]. Hybrids were Z.Z%."
   • Top EV Models table: Rank, Make, Model, Sold Count, Share of EV Segment %
   • Top Hybrid Models table: same structure
   • EV Brand Share table: Make, EV Units Sold, % of Brand's Total Sales That Are EV (shows which OEMs are most electrified)

Workflow: Regional Demand Heatmap

Map sales volume and pricing by state for a specific make or model to reveal geographic demand patterns.

1. Call mcp__marketcheck__get_sold_summary with:

   • date_from / date_to: target period
   • make: user's target make (required for this workflow)
   • model: user's target model (optional — omit for brand-level view)
   • summary_by: state
   • limit: 51 (all US states + DC)

2. If the user also wants pricing context, call mcp__marketcheck__get_sold_summary with:

   • Same filters plus ranking_dimensions: make,model
   • ranking_measure: average_sale_price
   • summary_by: state
   • limit: 51

3. Calculate for each state:

   • Volume rank (which states buy the most of this make/model)
   • Price rank (where is it cheapest vs most expensive)
   • Price-to-national-average ratio = State Avg Price / National Avg Price

4. Present as a State-Level Demand Table sorted by sold count descending:

   • Columns: State, Sold Count, % of National Volume, Avg Sale Price, Price vs National Avg, Avg DOM
   • Top 10 states get detailed analysis
   • Bottom 10 states flagged as potential growth markets

5. If the user specifies a model, also call mcp__marketcheck__get_sold_summary with ranking_dimensions: make for the same body_type (without the make/model filter) in the top 3 states to show competitive context: "In Texas, [Model] sold X units but [Competitor] sold Y units in the same segment."

6. Summary: "For [Make Model], Texas leads with X% of national volume at an average price $Y [above/below] the national average. The least penetrated large markets are [State A], [State B], [State C] — representing potential growth opportunities."

KPIs & Business Impact

→ After assembling results, read references/outcomes.md to frame recommendations with quantified business impact, KPI benchmarks, and action-to-outcome guidance.

Output Format

• Lead with the competitive headline. Example: "Toyota holds 14.2% national market share in January 2026, up 35 bps from December. Honda is the biggest gainer at +52 bps."
• Use ranked tables for share data. Always include both absolute volume and share %. Raw counts without context are meaningless; percentages without counts lack scale.
• Show share change in basis points (not percentage points) for precision. A move from 14.2% to 14.5% is "+30 bps", not "+0.3%".
• Always include comparison period data. A single-period snapshot is a fact. Two periods make a trend. Always show at least current vs prior.
• For EV/Hybrid analysis, always show penetration rate alongside absolute volume. "50,000 EVs sold" means nothing without knowing it is 7.2% of the total market.
• End with strategic implications tailored to the user's role:
  • For OEMs: allocation recommendations, incentive targeting, segment gaps
  • For dealer groups: competitive positioning, brand mix optimization
  • For analysts: trend narratives, inflection points, forecast implications
• Cite the data period and geography in every output (e.g., "Source: MarketCheck sold data, January 2026, US national, all dealer types").

Self-Check (before presenting to user)

• Share changes expressed in basis points (bps), not percentage points
• Both absolute volume AND share % shown in every table
• Comparison period data included (never single-period snapshot only)
• User's brand highlighted in rankings
• EV penetration shown as % of total market alongside absolute volume
• Data period and geography cited