Market Sizing and Category Economics: Role-Based Playbook

The shared spine contains the funded job, eligible buyer, category inclusions and exclusions, geography, period, revenue unit, source register, observed values, inferences, hypothetical inputs, scenarios, and refresh date. Every role may add a view, but no role should silently alter this foundation. If finance models annual recurring revenue while product counts workflows and strategy quotes ecosystem spending, the meeting needs a reconciliation before a recommendation.

Create a decision matrix listing the question, accountable owner, required evidence, maximum acceptable uncertainty, and action available to each role. A chief executive may authorize exploration, a product leader may select a workflow, finance may set an investment envelope, and a market researcher may approve a source for use. The distinctions keep expertise and authority visible while allowing the group to challenge the same assumptions.

A research owner should preserve the original source, definition, date, method, and transformations. Role leaders interpret those records for their decisions but should not rewrite source posture. For example, sales conversations may suggest a buyer problem, yet they do not become a population count. Product telemetry may show retained use in one cohort, yet it does not establish total category adoption. The evidence remains bounded by how it was collected.

When roles disagree, identify whether the conflict concerns evidence, inference, risk tolerance, or objective. Evidence disputes require source review. Inference disputes require alternative formulas or scenarios. Risk disputes require the accountable owner to state the accepted exposure. Objective disputes require executive alignment. Labeling the type of disagreement is faster and more honest than asking a larger market estimate to resolve a governance problem. It also determines the record to preserve: a corrected source, competing scenario, risk acceptance, or revised company objective.

The chief executive should inspect the revenue layer, category structure, adoption gates, substitutes, concentration, supplier dependencies, and obtainable constraints. A large total may still be a poor strategic fit if the company lacks trusted distribution, must customize every deployment, or depends on a supplier that captures most of the economics. A smaller segment may be more useful when the buyer, workflow, evidence, and delivery path are clearer.

Strategy should articulate why the company could earn a durable position without claiming an unsupported advantage. The rationale may involve a specific workflow, operating integration, evidence requirement, channel, cost architecture, or learning loop. Each element remains a hypothesis until observed. The next step should test the most consequential uncertainty at bounded cost rather than commit the organization to the most favorable long-range scenario.

Ask what would make the opportunity unattractive even if category interest grows. Possibilities include weak paid conversion, high implementation effort, low retained use, supplier cost exposure, compressed price, or an incumbent substitute that absorbs the feature. Review downside, base, and upside conditions independently. If the recommendation survives only when every uncertain variable improves, it is not a robust allocation case.

The executive output should state a decision, investment boundary, owner, observation period, and stop or scale conditions. It should also state what is not being approved, such as a public market claim, production release, broad hiring plan, or revenue forecast. This precision allows a company to learn without turning an exploratory category thesis into a commitment that later teams feel compelled to defend.

Product should define the smallest workflow that represents the category thesis and identify its trigger, context, actions, authority, exceptions, acceptance criteria, and recovery. It should distinguish current capability from configuration, manual procedure, integration work, and future intent. A feature list cannot establish serviceable demand if the complete workflow depends on unresolved data rights or customer process change.

Use discovery to test the job and its constraints, not to collect affirmative quotes for a deck. Record how the current process works, who owns the outcome, what alternatives exist, what failure costs, and which evidence would justify adoption. Include people who decline or cannot proceed. Their reasons may reveal category boundaries, readiness barriers, or substitute strength that an interested sample systematically misses.

Go-to-market should identify the budget owner, evaluator, user, blocker, channel, buying event, proof requirement, and contract unit. Pipeline stages must correspond to evidence such as confirmed problem, authorized evaluation, approved budget, negotiated scope, and accepted deployment. Label curiosity, requested information, pilot activity, paid purchase, retained use, and expansion separately. Each state supports a different inference about demand.

Pricing research should test the offer and value mechanism without presenting an assumed price as accepted willingness to pay. Record the scope, conditions, alternatives, and objections associated with every signal. A high proposed price may reflect a custom service burden rather than repeatable software value. A low price may stimulate usage that worsens cost. Commercial evidence becomes useful when it connects to product scope and delivery economics.

Finance should separate recurring platform revenue, usage, implementation, support, and other services by period. It should map direct model, data, tool, infrastructure, review, support, and implementation costs to the same buyer or workload unit. Actual, allocated, estimated, accrued, and missing costs should remain distinct. This structure reveals whether growth in activity improves contribution or merely increases unpriced consumption and exception work.

Finance should also test cash timing, contract obligations, refunds or credits where applicable, supplier invoice grain, and the cost of failed or refused work. Accounting treatment depends on the business, agreement, jurisdiction, and applicable standards, so the market model is not an accounting conclusion. Its role is to expose assumptions and provide a reconciled operating view for qualified owners to interpret.

An investor or capital committee should examine source quality, category boundary, adoption evidence, competition, substitutes, cost structure, concentration, delivery repeatability, and management's learning cadence. A growing category does not guarantee that a particular company can reach buyers, maintain differentiation, price above cost, or execute safely. Conversely, a company may find a defensible narrow segment without needing every broad category forecast to be correct.

The investment question should specify expected learning per unit of capital and the next decision enabled by that learning. Scenario probabilities, if used, are modeled judgments rather than observed frequencies unless a suitable dataset supports them. Material investment decisions require appropriate professional analysis. The playbook helps organize evidence and challenge, but it does not recommend a security, guarantee a return, or certify a company's prospects.

Hypothetical model: a team identifies 3,000 organizations, assumes 30 percent have a qualifying workflow, 60 percent meet readiness criteria, and 20 percent may buy within the period. The model yields 108 potential purchases. At an illustrative 45-unit annual contract, addressable annual revenue is 4,860 units. The source owner labels the population as observed only if a real cited source exists; all percentages and price remain assumptions.

Product reports that the current offer can serve only half the workflows without custom integration. Go-to-market can reach 120 eligible organizations. Finance estimates 17 units of direct and implementation cost per contract. Operations can onboard 16 customers. Strategy therefore cannot use 4,860 as the company plan. Each role applies its constraint to the same model rather than producing a more convenient independent total.

The chief executive authorizes a research and design phase, not a broad launch. Product tests whether one connector pattern can increase serviceability. Go-to-market validates budget ownership and offer scope among a balanced sample. Finance monitors cost by accepted workflow. Research replaces the readiness assumption with evidence. Operations defines onboarding quality and stop conditions. These are illustrative role actions, not a claim that the modeled opportunity is attractive.

At the next review, the group compares predictions with observed qualification, evaluation, cost, acceptance, and refusal. A lower buyer count may be offset by more repeatable delivery, or stronger interest may be offset by unacceptable review burden. The model changes only when evidence or a declared strategic assumption changes. The decision log records who accepted the new interpretation and which downstream commitment, if any, followed.

Research delivers sources and confidence. Strategy delivers the category thesis and alternatives. Product delivers the serviceability and workflow map. Go-to-market delivers reach and commercial state evidence. Finance delivers revenue and cost reconciliation. Operations delivers capacity and quality constraints. Executives deliver the bounded decision. Reviewers deliver challenge on claims and material specialist risks. Every output names assumptions, unresolved gaps, and a refresh condition.

The playbook cannot establish guaranteed market growth, buyer adoption, customer value, price, margin, share, or company performance. It should not be used to imply product availability, certification, regulatory acceptance, or investment suitability. Public claims and material legal, financial, accounting, security, privacy, competition, and investment decisions require qualified review. Evidence gathered for one role remains limited to its source population and method.

Within a verified configuration, OmegaOS can connect a shared market-intelligence record to role assignments, decision rights, governed work, economic observations, evidence, and Mnemosyne learning. That coordination can reduce repeated reconstruction and preserve why a role accepted or rejected an assumption. It does not remove the separation of duties or grant a model authority over capital, contracts, public communication, product release, or regulated decisions.

Use a narrow workspace with the minimum permitted context, one model version, named owners, and explicit review gates. Verify connectors, entitlements, source rights, and current runtime behavior for the intended use. Preserve disagreement and held states. The proportionate bridge is a common operating record around the role decisions; it is not a claim that OmegaOS has measured or will capture the agentic-company market.