Skip to main content

Business Model Analysis

Updated Jun 24, 2026 at 2:35pm

Research Draft High 1,169 words

Business model analysis is the qualitative work of figuring out how a company actually makes money — what it sells, who buys it, how it reaches them, how it prices, what it costs to deliver, and what the business depends on to keep running. It sits upstream of the numbers: the financial statements tell you what happened, the business model tells you why and whether it can repeat. The CFA Institute curriculum places it as the first analytical step in company analysis, ahead of industry and financial-statement work, because everything you later read in the financials is a consequence of the model. The core tension is that a business model is descriptive, not predictive — a well-described model still tells you nothing about valuation, and a beautiful model attached to an overpriced stock is still a bad investment.

The frameworks

Two frameworks dominate, and they are complementary rather than competing.

CFA Institute — five elements. The Level I equity curriculum decomposes a business model into five questions:

1. What it sells — the goods or services, their features, benefits, and differentiation. 2. Who it sells to — primary customers and segments (which drives scale and pricing). 3. How it reaches and delivers — sales channels and distribution (direct, retail, online, wholesale). 4. How it prices and gets paid — the pricing model and payment terms. 5. What it depends on — suppliers and key external relationships (dependency and bargaining-power risk).

A useful related distinction is between the business model (the whole system of value creation and capture) and the narrower revenue model — just the mechanism by which cash comes in (e.g., subscription, transaction fee, advertising, licensing, razor-and-blades). The pricing element above is where the revenue model is captured; the broader business model adds the cost, resource, and dependency context around it.

Osterwalder & Pigneur — the Business Model Canvas. From Business Model Generation (2010), the canvas maps nine building blocks on one page: Customer Segments, Value Propositions, Channels, Customer Relationships, Revenue Streams, Key Resources, Key Activities, Key Partnerships, and Cost Structure. The right side (customers, value, channels, relationships, revenue) describes value capture; the left side (resources, activities, partners, costs) describes value creation. It is more granular than the CFA's five elements and is the standard tool in strategy and venture circles.

Both reduce to the same triad popular among investors: how the company creates value, delivers it, and captures it.

How it's used in practice

For an investor, framework completion is the start, not the end. The practitioner workload is:

  • Map the revenue model precisely. Is it one-time transactional, recurring subscription, usage/consumption-based, advertising, licensing, or razor-and-blade? Recurring and usage models produce more visible, higher-quality revenue; transactional models reset to zero each period.
  • Decompose unit economics. Reduce the model to a single repeatable unit (a customer, a store, a subscription) and ask whether that unit is profitable. For subscription businesses the standard lenses are customer lifetime value (LTV) vs customer acquisition cost (CAC), the LTV:CAC ratio, and the CAC payback period (CAC ÷ monthly gross-profit-per-customer). A commonly cited venture benchmark is LTV:CAC of roughly 3:1 as a minimum and 5:1+ as strong — though these are private-company rules of thumb, not audited public metrics, and depend entirely on how LTV is estimated.
  • Judge incremental economics. A high gross margin (a software business near ~80% is the textbook example) means each new dollar of revenue is cheap to serve, so growth and profitability can rise together. A low-margin model must be far more efficient everywhere else to reach the same returns.
  • Assess capital intensity. Asset-light models (software, marketplaces, franchisors) convert growth to free cash flow with little reinvestment; asset-heavy models (airlines, utilities, manufacturers) must keep feeding capex to grow, which caps returns on capital.
  • Trace dependencies and bargaining power. Single-supplier reliance, customer concentration, platform dependence (e.g., reliance on one app store or ad network), and regulatory exposure are where models break.
  • Connect the model to the durability of returns — i.e., whether the value capture is defensible (the competitive-moat question handled by the sibling node).

The output should be a thesis: this is how the company earns a dollar, this is why that dollar is durable (or not), and this is what would have to be true to keep it coming.

Strengths & limitations

Its strength is that it forces an analyst to understand the asset before pricing it, and it surfaces structural risks (customer concentration, platform dependence, capital intensity) that ratios alone hide. It is also where the difference between a good business and a good investment gets clarified.

The limitations are real and under-acknowledged:

  • It is descriptive, not valuation. A canvas can be filled in perfectly for a company that is a terrible buy at today's price. The single most common misuse is treating "great business model" as a buy signal — narrative-driven overpayment is precisely how quality-business stories become losing trades.
  • It is qualitative and judgment-laden. Two analysts can map the same company differently. There is no agreed scoring system and no measured base rate for "model quality predicts returns."
  • Unit-economics figures are soft. LTV especially relies on assumed churn and discount rates; the 3:1 benchmark is folklore from venture practice, not an empirical threshold validated on public equities. Treat any single number as an estimate.
  • Models are not static. Disruption, regulation, and changing customer behavior re-write models (newspapers, retail, linear TV). A correct map of today can be obsolete in three years.

Sources

Disputes flagged: the LTV:CAC 3:1 / 5:1 benchmarks and the ~80% "high gross margin" reference are widely cited venture/SaaS rules of thumb, not empirically validated thresholds for public-equity returns; LTV itself is estimate-dependent.