Positioning

How do you build a position? A six-step framework#

There is no honest formula that outputs the right position. Positioning is a hypothesis built from linked evidence. These six steps make the reasoning auditable.

1. Define the buying situation#

Start with a situation, not a demographic label. Document the job or outcome, the event that makes action urgent, the people who use / champion / approve / pay / block, the cost of delay, current budget and workflow ownership, and the success measure the customer will accept.

"Mid-market software companies" is not a buying situation. "A 50–500 employee B2B software company has received a 250-question security review that is blocking a six-figure enterprise deal" is. This step draws directly on Jobs to Be Done, Customer Use Cases, and the Ideal Customer Profile.

2. Identify the real competitive alternatives#

Ask recent buyers and lost prospects one question: if this product were unavailable, what would you do? Capture behavior, not just named vendors — alternatives usually include a direct competitor, an adjacent suite, a manual process, internal software, a consultant, reassigning the job to another employee, accepting the risk, or doing nothing.

Then quantify frequency:

Alternative share = qualified decisions in which an alternative was seriously considered ÷ all qualified decisions observed

Build the denominator from win/loss interviews, CRM notes, proposals, call recordings, onboarding surveys, churn interviews, and discount approvals — the same evidence categories antitrust regulators use to establish substitution. Do not infer it from website traffic or keyword volume.

3. Isolate differentiated capabilities#

For each important alternative, list capabilities the startup has that the alternative lacks, stated as facts a product or operations team can verify.

A capability does not have to be unique in the world — only meaningfully different from the alternatives the target customer is actually considering. It can be product architecture, workflow design, service model, distribution, data access, compliance status, implementation speed, or commercial terms.

4. Translate capability into value and proof#

Features are causal inputs; value is the outcome. Use a value chain:

capability → changed customer behavior → measurable outcome → economic or risk value → proof

For example: source-cited answer generation → fewer manual evidence lookups → shorter questionnaire cycle → faster enterprise deal progression → timestamped workflow data and customer case evidence

For each claim, record six things: the mechanism (why the capability should cause the outcome), the metric and time window, the baseline performance of the alternative, the proof (product data, controlled test, audit, customer record), the boundary conditions under which the claim holds, and the owner who maintains the evidence.

This is where marketing discipline becomes legal exposure. Under FTC policy, an objective claim implicitly represents that you already hold a reasonable basis for it, and the required level of support scales with the type of claim, the product, and the consequences of being wrong. Treat proof as a precondition to publishing a quantified claim, not a follow-up task.

5. Select the best-fit customer#

The best-fit customer is the segment for which the differentiated value is important, frequent, urgent, reachable, and economically serviceable. Useful markers: the trigger occurs often enough to justify action; the current alternative is costly or seriously limited; the differentiated capability is required rather than merely appreciated; the buyer can recognize and measure the outcome; budget and authority exist; the route to market reaches the account efficiently; and the product can deliver without exceptional founder labor.

Do not define the segment by company size, industry, or persona alone. Combine observable account characteristics with use case, trigger, current alternative, technical environment, and buying process.

6. Choose the category and write the thesis#

Choose a category that helps the customer understand the value without creating fatal expectations. Evaluate a candidate on seven dimensions:

Then write the internal thesis:

For [best-fit customer in a specific buying situation] who would otherwise [use the primary alternative], [product] is a [market category] that [delivers the most important differentiated value] because [differentiated capability and proof].

The sentence is a compression tool, not the analysis. If the underlying evidence is weak, polishing the sentence will not repair the position.

The positioning evidence map#

Maintain one table that sales, product, marketing, finance, and customer success can audit:

How do you measure positioning?#

Positioning affects several funnel stages at once, so use an evidence stack rather than a single score.

Qualified response rate — qualified target accounts that respond ÷ qualified target accounts reached. More useful than total response, because a sensational message attracts people outside the intended segment.

Opportunity creation rate — qualified opportunities created ÷ qualified target accounts reached. Define "qualified opportunity" before the test: buying situation, authority, alternative, timing, and a plausible value case.

Win rate against the named alternative — wins where alternative A was in the decision ÷ decided opportunities where A was in the decision. Always report the count. A 75% win rate from four decisions is not stable evidence.

Value realization rate — activated customers achieving the promised outcome in the stated window ÷ activated customers eligible for measurement. This catches positions that sell well but promise the wrong outcome.

Price realization — net contracted price ÷ target or list price. Persistent discounting can indicate weak differentiation, the wrong buyer, missing proof, poor packaging, or simply an unrealistic list price. It is not automatically a positioning failure.

Customer acquisition payback —

CAC = fully loaded acquisition cost ÷ new customers acquired

Monthly contribution per customer = monthly revenue − variable delivery and service cost

CAC payback (months) = CAC ÷ monthly contribution per customer

Include paid media, sales labor, founder time at replacement cost, commissions, trials, and onboarding incentives. Use contribution after variable delivery cost, not revenue. Finally, compare retention and expansion by position, segment, channel, and cohort: a message that lowers CAC but worsens retention destroys value.

Worked example: two positions for the same product#

ClausePilot is a fictional early-stage B2B software company. Its product drafts security-questionnaire responses from a company's approved policies and past answers, attaches source citations, and routes uncertain responses to control owners. The founders are choosing between two composite hypotheses:

• Position A — broad productivity: "An AI operations platform that automates repetitive business work."
• Position B — focused revenue blocker: "Security-questionnaire automation for B2B SaaS teams closing enterprise deals."

Position B names a trigger, an alternative workflow, a segment, and an outcome tied to revenue timing. It also creates narrower expectations: the product must be excellent at questionnaire workflows, evidence controls, export formats, and review.

The founders run two sequential 60-day market tests. Each receives $6,000 of fully loaded acquisition effort and reaches 200 prequalified accounts through the same outbound channel. This is not a clean A/B test — target definition, frame, price, and offer all differ, so it compares two complete go-to-market hypotheses rather than isolating one variable. Within each test, qualification and outcome definitions are held fixed.

Observed results#

Funnel and economics#

Assume the one non-retaining customer in each cohort churns after month 3 and the rest stay a full year. This is a scenario, not a lifetime-value estimate.

• Position A: 2 × 12 + 1 × 3 = 27 customer-months × $225 = $6,075 first-year contribution, or $75 net of the $6,000 acquisition effort.
• Position B: 8 × 12 + 1 × 3 = 99 customer-months × $400 = $39,600, or $33,600 net.

What this does and does not show#

The correct conclusion is not "narrow positioning always wins." It is:

In this channel, period, sample, and offer, the focused security-questionnaire position produced stronger qualified response, conversion, value realization, retention, price, and payback than the broad productivity position.

Three caveats keep the numbers honest. First, the denominators are tiny — nine customers cannot distinguish a 50% win rate from a 35% one. Second, the 12-month retention assumption does most of the work in the contribution figures; shortening it to six months roughly halves both results. Third, the 80% contribution margin quietly assumes onboarding and review labor stay inside that margin, which is exactly what breaks first when a founder-led motion scales.

Before scaling, the founders should repeat the test with another cohort; interview wins, losses, and non-responders about the actual alternative; verify that questionnaire cycle time genuinely improved; measure the service labor hidden in the margin assumption; separate the effects of segment, category, claim, price, and sales execution; and check whether the position holds outside founder-led selling. For causal copy or campaign tests specifically, hold everything else constant — a clear hypothesis, one variable, success metrics chosen in advance, and stable base conditions.

How do teams use the position day to day?#

Founders use the position to reject work as well as choose it. A request is strategically relevant when it improves the chosen capability, proof, segment, buying process, or delivery economics. A large logo can still be a distraction if it requires a different position and a custom product.

Product chooses depth over accumulation. If source-cited, approval-controlled answers are the differentiation, then evidence freshness, permissions, auditability, and export reliability matter more than a general chatbot. Measure whether new capabilities improve the promised outcome for the target segment.

Marketing builds a message house from evidence: a category and short description, the primary problem and trigger, two or three value pillars, proof points for each pillar, the alternative and differentiated capability, audience-specific objections, and claims that must not be used without qualification. Public templates from GitLab and Atlassian follow this structure — positioning statement, target audience, value pillars, pain points, proof points.

Sales qualifies around the position. Discovery should test whether the account is actually in the assumed buying situation, and should capture trigger and deadline, current alternative, cost of the status quo, decision roles and budget, required proof, technical prerequisites, reason won or lost, and discount reason. That creates evidence for product and marketing instead of reducing win/loss analysis to a reason code like "price."

Pricing anchors to the right alternative. If the product replaces 40 hours of scarce security and sales-engineering work and unblocks a revenue-critical cycle, the relevant reference value is not the price of a generic AI seat — but do not automatically capture the full estimated value. Consider uncertainty, implementation cost, switching risk, bargaining power, competitive response, and the surplus the customer needs to justify acting. See Strategic Pricing and Pricing Metric / Value Metric.

Customer success makes the promise operational. Onboarding should reach the value event named in the position, and the success plan should track the same metric used in the claim. If sales promises a 30-day reduction in questionnaire cycle time but customer success tracks only logins, the company cannot learn whether the position is true.

Investors test whether the narrative matches behavior: Do wins come from the named segment? Is the stated alternative present in real decisions? Are claims supported by product and customer evidence? Do price realization, CAC, retention, and margin improve in the target segment? Can the wedge expand without abandoning the source of advantage? Would an incumbent neutralize the difference cheaply?

What are the common mistakes?#

• Treating a tagline as the position. "Work smarter" is a headline. It identifies no customer, alternative, difference, value, or category.
• Starting with the founder's favorite category. Category-first reasoning forces evidence into a predetermined frame. Start with alternatives and differentiated value, then pick the category that makes those facts easiest to understand.
• Defining competitors by product resemblance — and treating "do nothing" as no competitor. Technical similarity does not prove substitution, and no action often wins because change carries implementation cost, political risk, and uncertain value. The position must explain why acting now beats delay.
• Claiming universal superiority. "Best," "only," "fastest," and "most accurate" create substantiation obligations and hide the boundary within which the claim might be true. A vivid customer quote communicates value but does not substantiate an objective performance claim.
• Optimizing for response rather than customer quality. A broad AI message may generate more leads while lowering activation, price, retention, or margin. Follow cohorts through realized value and economics before declaring a winner.
• Repositioning after every lost deal. One loss is evidence, not a mandate. Review patterns across comparable decisions, and separate position from product gap, price, timing, procurement, trust, and sales execution.

When does positioning break?#

• Very early products with little customer evidence. Before customers make real decisions, positioning is necessarily provisional. Use problem interviews, prototype tests, pre-orders, design partnerships, and observed alternatives — and label assumptions as assumptions rather than hardening early copy into strategy.
• Genuinely new categories. A new capability may not fit a familiar category, but an unfamiliar one imposes education cost and lacks budget. Start with the nearest understood alternative and a frontier customer with an urgent use case; let repeated behavior justify broader category language later.
• Marketplaces and networks. Each side may need a different position, value proposition, and proof. A demand-side position can attract users while supply economics collapse — measure liquidity, quality, trust, and contribution on each side. See Two-Sided Markets.
• Regulated, health, financial, and safety-critical claims. The most persuasive outcome often requires the strongest substantiation and disclosures. Legal review, clinical or technical evidence, certification, and approved wording may constrain the position. Do not treat regulatory friction as a copy problem.
• Fast-moving AI markets. A model capability that is unique today can become a platform feature next quarter. Durable positioning may need proprietary workflow, data rights, distribution, trust, integration, cost control, or service design — and the alternative map should be rechecked after major model or platform releases.
• Small samples and long cycles. Enterprise and regulated startups may simply not have enough decisions for reliable rates. Use account-level evidence, publish denominators, distinguish directional from stable results, and avoid false precision.

Above all: positioning cannot repair missing value. Clear framing improves comprehension and comparison. It cannot create a needed outcome, acceptable quality, trust, willingness to pay, or workable economics. If customers understand the offer and still do not care, change the product, segment, price, or business model — not the words.

Founder positioning review checklist#

Customer and situation — Can we name the buying trigger and required outcome? Do we know the user, champion, economic buyer, and blocker? Which observable characteristics make the value unusually important? Which customers should we decline or defer?

Alternatives and differentiation — What would the customer do if we did not exist? How often does each alternative appear in qualified decisions? Which capabilities are factually different? Could an incumbent copy or bundle the difference cheaply?

Value and proof — What process change connects capability to outcome? What baseline and time window make the claim meaningful? Do we hold the evidence before publishing? Does the customer success plan measure the promised outcome?

Category and expectations — Does the category create the right comparison set? Can we meet its table stakes? Does it point to the correct buyer, budget, and price anchor? Is the education cost acceptable?

Funnel and economics — Does the position improve qualified opportunities, not just responses? What is win rate against the named alternative, and on what denominator? Do value realization, price realization, retention, and contribution improve? Does the result persist outside founder-led selling?

Governance — Is the current thesis written and dated? Are rejected positions and reasons documented? Do all functions use the same logic? Which market event triggers a review?