BCG Growth-Share Matrix

intent

The BCG Growth-Share Matrix reveals which business units fund others, which consume cash without clear payoff, and which require an immediate keep-or-kill decision. use it when a firm operates 3+ distinct business units competing for shared capital, especially during annual strategy reviews, PE/VC portfolio health checks, or AI capital reallocation decisions (deciding which units to harvest to fund AI capex). the matrix forces prioritization by mapping each unit on market growth rate (y-axis) vs. relative market share (x-axis), producing four quadrants: Stars (invest aggressively), Cash Cows (extract surplus), Question Marks (binary decide), and Dogs (exit or hold minimally). rest it on two anchors: the experience curve (high share typically yields lowest cost) and industry lifecycle (high growth demands reinvestment; maturity throws off cash).

inputs

firm context:

• ≥3 distinct business units with separate P&Ls, identifiable customer groups, and named competitors
• access to 2-3 years of external market data (not internal revenue only)
• capital allocation debate or budget cycle that needs forcing function

data required per SBU:

• unit name, revenue, and profitability (last 12 months, baseline)
• market growth rate: external CAGR for the addressable segment (e.g., Gartner, IDC, category association data). threshold is 10% (raise to 20-30% for AI/clean-tech). never use own revenue growth as proxy
• own market share: % of defined addressable market
• largest competitor's market share: % of same addressable market
• 2-year trend: qualitative direction (gaining share, flat, losing share)

external connections:

• market research databases (Gartner, IDC, Nielsen, IRI, or industry association reports). env var: MARKET_DATA_SOURCE (defaults to public company filings + analyst reports if proprietary DBs unavailable)
• internal finance system (P&L access for unit revenue and profitability; no auth required if user has read access)
• competitor intelligence: public filings, analyst coverage, or win/loss data (optional; if unavailable, relative share estimated from proxy metrics per Portfolio Packs table)

user mode (adaptive):

• engine mode: user provides structured BU data, run the process directly
• coach mode: user unfamiliar with BCG, guide step by step with [WAIT] hard stops

procedure

step 1: validate fit and define SBUs

input: firm structure, user's business unit list.

action: confirm each unit has (a) identifiable customer group, (b) named competitor(s), (c) separable P&L. apply the "name the primary competitor" stop rule: if you cannot name the largest rival, the boundary is wrong.

output: approved list of 3+ SBUs with clear definitions (e.g., "North America Enterprise SaaS Division," "Consumer AI Chatbot Product," "Legacy ERP Maintenance Service").

edge case: if units are highly interdependent (divesting a Dog would destroy a Cash Cow's channel), flag this and consider redefining boundaries or note the constraint in strategy recommendations.

---

step 2: calculate market growth rate per SBU

input: external market data source(s), SBU definition.

action: retrieve 2-3 years of addressable market size (not own revenue). calculate CAGR. confirm source is external (analyst firm, category association, public market data), not internal revenue. set growth threshold: 10% baseline, 20-30% for AI/clean-tech/emerging segments.

output: growth rate (%) per SBU; threshold used; data source and date.

edge case: if external data unavailable or stale (>18 months old), note uncertainty; use analyst consensus or triangulate from public comps. if growth rate is volatile or cyclical (e.g., commodity), use 3-year CAGR and flag.

---

step 3: calculate relative market share per SBU

input: own market share (%), largest competitor's share (%).

action: divide own share by largest competitor's share. >1.0 = market leader; <1.0 = follower. record both numerator and denominator for transparency.

output: relative market share ratio (e.g., 1.8, 0.6) per SBU.

edge case: if market is fragmented (top competitor <20% share), log that relative share ratio is low across all units and contextualize results. if own share > largest competitor but both are <10% total market, note that dominance is relative and market is open.

---

step 4: plot quadrants and assign SBU positions

input: growth rates, relative share ratios, SBU revenue (for bubble size).

action: create 2x2 grid. x-axis: relative market share (log scale, left = low/follower, right = high/leader). y-axis: market growth rate (linear, bottom = low, top = high). dividing lines: x = 1.0 (relative share), y = growth threshold (default 10%). plot each SBU as bubble; bubble size = revenue or (if comparable) profit contribution.

assign quadrant:

• Stars (high growth, high share): upper right
• Cash Cows (low growth, high share): lower right
• Question Marks (high growth, low share): upper left
• Dogs (low growth, low share): lower left

output: visual 2x2 matrix or table with SBU, growth %, rel. share, quadrant, revenue, profitability.

---

step 5: plot 2-year trend arrows and assess momentum

input: historical growth, share trend (gained, flat, lost), competitive actions.

action: for each SBU, draw trend arrow showing 2-year trajectory. direction indicates whether unit is moving toward Star, Cow, Mark, or Dog. analyze reason: e.g., "Question Mark moving toward Star" (new product adoption accelerating, share gained); "Cash Cow moving toward Dog" (market growth slowing, losing share to insurgent). record the reason.

output: trend arrows on matrix, reason per SBU (1-2 sentences).

edge case: if data is <2 years old or unit is new, project trend using competitive and market signals, note assumption.

---

step 6: quantify cash flows and map dependencies

input: profitability per SBU, investment/capex requirements per SBU, strategic synergies (if any).

action: identify cash generators (Cash Cows, profitable Dogs) and cash absorbers (Stars, Question Marks). sum: cash surplus / deficit. confirm Cash Cow generation can fund Star and Question Mark investment needs. identify any hidden dependencies (e.g., Question Mark uses Cow's manufacturing, Sales channel, brand).

output: cash flow summary (e.g., "Cows generate $120M surplus; Stars + Marks require $95M; net surplus $25M"). dependency map if relevant.

edge case: if Cow generation < Star + Mark needs, flag capital constraint and signal that not all Stars can be funded or some must be exited.

---

step 7: assign strategy per SBU and set decision deadlines

input: quadrant position, trend, cash flow, competitive context, synergies.

action: for each SBU, assign strategy:

• Star: aggressive investment (capex, R&D, marketing to defend/extend leadership). monitor for deterioration; if growth slows and share erodes, prepare exit.
• Cash Cow: extract surplus; minimize capex; harvest. maintain quality and customer retention to avoid fall to Dog. refresh offerings if market shifts.
• Question Mark: binary decision. define criteria (e.g., "hit 20% share within 18 months or exit"). set decision deadline (named month/year). if criteria not met by deadline, divest or fold into another unit.
• Dog: exit plan (timeline, buyers, or shutdown). exception: retain only if (a) synergy is named and quantified (e.g., "provides sales channel for Cow, saves $5M annually"), or (b) strategic option value is explicit (e.g., "holds option on emerging market; kill if market does not inflect by 2026").

output: strategy table (SBU | quadrant | strategy | decision date | named rationale).

---

step 8: produce portfolio map and recommendations

input: all prior outputs (definitions, growth, share, quadrants, trends, cash flows, strategies).

action: synthesize findings into a single narrative. highlight:

• which unit is the primary cash generator and its health trajectory
• which units are burning cash and whether they have a path to self-sufficiency or exit date
• the key bottleneck or decision (e.g., "Question Mark requires $50M capex over 3 years; if not approved, must exit by Q2 2025")
• any matrix anomalies or red flags (e.g., Cow moving toward Dog, Mark with no exit criteria, Dog retained without quantified synergy)

output: portfolio map (visual or detailed table) + narrative summary per output contract below.

---

decision points

fit check (step 1):

• if firm has single product or <3 distinct units → redirect to Ansoff Matrix for growth direction or Porter's Five Forces for competitive depth. abort BCG.
• if units are highly interdependent (e.g., shared manufacturing, sales channel, brand equity) and divesting a Dog would destroy a Cow → note constraint in strategy and consider redefining unit boundaries or lowering exit probability for dependent Dogs.

data availability (step 2-3):

• if external market growth data unavailable → use analyst consensus, triangulate from public comps, or note uncertainty and use qualitative assessment (growing, flat, declining). flag as "estimated" in output.
• if competitor share data unavailable (private company, fragmented market) → use proxy metrics (Portfolio Packs table: SaaS units use ARR share vs. ICP rivals; AI products use monthly active API users vs. nearest rival; retail uses GMV share). note proxy used.

growth threshold (step 2):

• if industry is AI, clean-tech, or other high-growth segment → raise threshold to 20-30% to avoid inflating Question Marks with inflated baseline growth.
• if industry is mature or cyclical → use historical average or recession-adjusted rate; do not use peak year.

quadrant assignment (step 4):

• if SBU falls near dividing line (e.g., growth = 9-11%, relative share = 0.9-1.1) → note sensitivity; small data shifts change quadrant. do not over-interpret position; watch trend arrow.

trend analysis (step 5):

• if trend is flat or data is sparse → rely on competitive intelligence (e.g., competitor product launches, share wins, pricing changes) to infer direction. flag assumption.

cash flow constraint (step 6):

• if Cow generation < Star + Mark investment needs → decision forced: (a) exit or deprioritize some Stars/Marks, (b) fund Stars/Marks externally (debt, equity, divestiture), or (c) accept slower investment and longer payoff. escalate to CFO/board.

Question Mark strategy (step 7):

• if Question Mark has no exit criteria or decision deadline → assign one. default is exit by 18-24 months unless explicit criteria (e.g., "hit 25% share, <$30M annual burn, clear path to Cow") are met.
• if Question Mark is VC-backed or strategic option → quantify the option value (e.g., "holds emerging market position; retain if TAM inflects within 3 years"). do not call every uncertain unit "strategic."

Dog retention (step 7):

• if Dog is retained, synergy must be named and quantified (e.g., "provides legacy service revenue for 200 enterprise Cow customers, saves $5M/year in churn; keep until Cow modernized"). no unquantified synergies.
• if Dog is being divested → identify acquirer, timeline, and transition plan (customers, contracts, employees).

single-product startup or early-stage firm:

• if user is a startup with one product or product line → do NOT use BCG. redirect to Ansoff Matrix (which growth direction to pursue) or Five Forces (competitive analysis).

technology-transition market:

• if market is in flux (e.g., traditional enterprise shifting to cloud, mobile disrupting desktop) → relative share becomes less predictive. use BCG but weight trend arrows and leading indicators (customer satisfaction, win/loss, product roadmap momentum) heavily. consider running Five Forces or disruption assessment in parallel.

output contract

portfolio map (required):

format: visual 2x2 grid or markdown table with columns: SBU | market growth (%) | rel. market share | quadrant | revenue ($M) | profitable? (Y/N/partial).

example:

BCG Portfolio Map: Acme Corp | Threshold: 10% | Date: 2025-01-15

SBU | Growth | Rel.Share | Quadrant | Revenue | Profitable?
Enterprise SaaS | 22% | 1.8 | Star | $450M | Y (15% margin)
Legacy Licensing | 2% | 2.1 | Cow | $280M | Y (35% margin)
AI Chatbot (new) | 85% | 0.4 | Mark | $15M | N (-$8M EBITDA)
Maintenance Services | -1% | 0.7 | Dog | $50M | Y (20% margin)

trend arrows (required):

format: narrative or visual arrows showing 2-year trajectory per SBU.

example:

Trend: Enterprise SaaS moving Star → Star (share + 300bps YoY, market +22%); momentum sustained.
Trend: AI Chatbot moving Mark → (path to Star if <$20M CAC, or exit if >$25M); inflection point Q3 2025.
Trend: Legacy Licensing Cow → Dog (growth -3% YoY, losing share to cloud; margin compression begins); watch renewal rates.

cash flow summary (required):

format: cash generators, absorbers, net balance.

example:

cash generators: Legacy Licensing ($98M EBITDA), Enterprise SaaS ($68M EBITDA) = $166M.
cash absorbers: AI Chatbot ($8M loss), Star capex ($20M/year) = $28M annual need.
balance: $138M surplus available for reinvestment, dividends, or M&A.

strategy assignments (required):

format: table with SBU, quadrant, strategy, decision date, rationale.

example:

SBU | Quadrant | Strategy | Decision Date | Rationale
Enterprise SaaS | Star | aggressive invest ($30M capex 2025, target 25%+ share by 2027) | ,  | defend leadership, fund from Cow surplus
AI Chatbot | Mark | invest aggressively OR exit by Q3 2025 | Q3 2025 | criteria: <$20M CAC to fund, >50% YoY user growth to sustain. if neither met, divest or fold into SaaS product.
Legacy Licensing | Cow | harvest (maintain current capex ~$5M/year, extract max margin, prepare sunsetting roadmap) | ,  | milk cash; modernize features only to retain top-100 customers; plan migration to SaaS over 5 years
Maintenance Services | Dog | exit or hold minimally (no capex; support declining customer base; divest if buyer emerges) | 2026 Q2 | no quantified synergy; if customer base shrinks below 50 accounts, consider sunsetting.

key decision (required):

one-line forcing function produced by the matrix.

example: "AI Chatbot unit has 18 months to hit unit economics inflection (CAC payback <12 months) or will be exited; this requires $8M incremental marketing spend Q1-Q2 2025. board must approve or greenlight exit planning."

red flags / anomalies (if present):

call out any items that break the standard logic (e.g., "Cow does not cover Star investment needs; capital constraint present"; "Dog retained without quantified synergy , sync with strategy owner"; "Question Mark has no exit criteria").

refresh cadence (required):

state when matrix will be recalculated (e.g., "refresh Q4 2025 when market growth forecasts update; sooner if acquisition or major product launch changes unit positioning").

outcome signal

the user knows the skill worked when:

• the portfolio map clearly assigns each business unit to a quadrant with growth rate, market share, and revenue visible. the user can see at a glance which units fund others.
• trend arrows show 2-year momentum, making it obvious which units are strengthening, stagnating, or deteriorating. the user can predict which Cows are moving toward Dogs.
• cash flow summary quantifies the surplus or deficit, forcing a capital decision: either Stars are fully fundable, or some must be deprioritized or exited.
• each Question Mark has a named decision deadline and binary criteria (invest OR exit by X date). no Question Mark is left in limbo.
• each Dog has either (a) an exit plan with a timeline, or (b) a named, quantified synergy (e.g., "saves $X annually via channel reuse"). no Dogs retained on vague "strategic" reasoning.
• the "key decision" section names the single biggest bottleneck or forcing function emerging from the analysis (e.g., "Question Mark's go/no-go decision by Q3 2025; requires $X capex approval from board by Q1 2025").
• if used in a meeting or board setting, the matrix stops hand-waving about which units to fund. decisions become binary and time-bound.
• the user schedules a refresh cadence (stated date for next recalculation), signaling the matrix is not a one-time output but a living planning tool.

---

coaching novices (adaptive front door)

engine mode: user has structured BU data (growth %, market share %, financials). run the process directly (steps 1-8). output portfolio map and recommendations in full.

coach mode: user unfamiliar with BCG or hesitant about unit definitions. step through at user pace. use [WAIT] as a hard stop. do not advance until user acknowledges or provides data.

coach mode flow:

"BCG shows which businesses fund others, which burn cash, and which need a hard keep-or-kill call. it uses two numbers: market growth rate and your share relative to your biggest rival.

first, a quick fit check: