Define the Promise — and Who It Is Not For

Task

Define the customer promise, qualification criteria, and disqualifiers.

Summary

Clarify the buyer, trigger, qualification criteria, disqualifiers, and customer promise.

Define Who the Offer Is For—and Who It Is Not For

Task ID: S2-02

A repeatable offer needs more than a clear description and a price. It needs a customer promise that can be supported, qualification criteria that predict a successful engagement, and disqualifiers that stop the company from accepting work it cannot deliver consistently. This article explains how to define those elements, test them against real customer results, and measure whether they improve customer fit.

The sale that should have been declined

A founder receives a referral from a respected customer. The prospect has money, an urgent problem, and a familiar company profile. The opportunity looks ideal.

During the sales conversation, however, the prospect asks for several exceptions. The work must start before its data is ready. Its executive sponsor will not attend the planning sessions. It expects the service to cover several business units, even though the quoted scope covers one. It also wants a result that the company has achieved only when customers supplied timely information and assigned an experienced internal owner.

The founder accepts the deal.

Revenue arrives, but the offer immediately stops being repeatable. Senior people are pulled into delivery. The team adds undocumented work to protect the relationship. The schedule slips. The customer receives something different from what it thought it had purchased, while the company earns less than expected.

The problem was not simply weak scope control. The company made a promise without defining the conditions under which it could keep that promise.

That is the purpose of customer qualification. It is not merely a sales technique for ranking leads. It is the operating process that determines where the offer is likely to create its intended result, where the company can deliver it economically, and where both parties should decline or delay the engagement.

The central principle is:

The customer promise, qualification criteria, and disqualifiers must be designed as one system.

A promise describes the result the offer is intended to produce. Qualification establishes whether the conditions for producing that result are present. Disqualifiers identify conditions that make the result, delivery method, or economics unreliable.

Without all three, the company may have a named package, but it does not yet have a repeatable offer.

The promise and the fit decision belong together

A customer promise is a clear statement of the useful change the offer is intended to produce for a defined customer in a defined situation. It is not a list of activities, a collection of features, or a claim that the company can help almost anyone.

Research on value-based business-to-business selling distinguishes three connected activities: understanding the customer’s business, constructing the value proposition, and communicating the value the offering could create in the customer’s operations. This shifts the conversation from what the supplier does to what the customer should be able to achieve.

For a productized service, the promise must also support repeatable delivery. Research on service productization describes the work as turning variable or hard-to-understand services into defined, documented, and reproducible offerings. The same research emphasizes several connected decisions: solve a recognized customer problem, define the target group, communicate the benefit, document delivery, and choose an appropriate balance between standardization and customization.

That balance matters. A promise that is so broad that it accommodates every prospect will create exceptions in delivery. A promise that is too narrow may exclude useful variation that could be handled through standard options or modules. The goal is not to make every customer identical. It is to determine which differences the offer can absorb without becoming a new service each time.

The promise should therefore identify five things:

The buying trigger deserves particular attention. A customer profile explains who might have the problem. A trigger explains why the problem has become important enough to address now.

Jobs to Be Done theory similarly places decisions in their circumstances. It argues that people and organizations choose products or services when particular forces create a need to make progress; demographics and product attributes alone do not explain the choice.

A company might therefore have the right industry, employee count, and revenue but still be a poor prospect because nothing has made the problem urgent. Conversely, a company slightly outside the usual profile may be a strong prospect because a merger, regulatory deadline, failed system, new executive, rapid hiring period, or major customer requirement has created an immediate need that closely matches the offer.

Promise, fit, and readiness are different questions

Three decisions are often mixed together:

Customer fit: Is this the type of customer and problem the offer was designed to serve?

Delivery fit: Can the company produce the intended result through the standard scope, process, staffing model, and price?

Buying readiness: Is there a real reason to act, a responsible owner, and a workable decision process now?

A prospect can have excellent customer and delivery fit but no immediate trigger. That prospect may belong in a future follow-up process rather than being rejected permanently. A prospect can also be eager to buy while having poor delivery fit. Urgency does not make an unsuitable engagement suitable.

This distinction is visible in GitLab’s public lead-management process. Its status definitions separate a qualified lead from a disqualified one, a lead that may be suitable later, and one that is ineligible for the sales process. The categories are company-specific, but the operating lesson is broadly useful: “not ready now” and “not suitable” should not be stored as the same decision.

Qualification should therefore produce more than a yes-or-no label. At minimum, it should distinguish:

• qualified now;
• suitable, but not ready;
• suitable only for a different offer or scope;
• unsuitable for reasons unlikely to change.

These outcomes protect future opportunities without pressuring the sales team to keep weak deals active.

What the evidence says about choosing customers

Choosing which customers to pursue can feel unnecessarily restrictive when a company is still building revenue. The evidence nevertheless supports thoughtful prioritization—provided the company has adequate information and applies the decision consistently.

A cross-industry study of 310 business-to-business and business-to-consumer firms, supplemented by two validation samples, found that customer prioritization was associated with higher average customer profitability and return on sales. The researchers attributed the result to better relationships with higher-priority customers and lower marketing and sales costs. They also found that implementation depended on the quality of customer information, the ability to assess profitability, organizational alignment, management involvement, and planning and control.

That final point is crucial. Writing “ideal customer” at the top of a sales document does not create prioritization. The definition must affect who receives sales attention, which deals advance, what exceptions require approval, and what the company measures after the sale.

Research on knowledge-intensive business services also shows why the promise cannot be written entirely from internal opinion. Effective value communication depends on customer knowledge, and the criteria customers emphasize can change with the stage and depth of the relationship. The provider may need to formulate the initial proposition when a relationship is new, then refine its understanding as the customer supplies better information and the work develops.

This supports a practical sequence:

1. Start with evidence from completed work, lost sales, difficult delivery, and customer interviews.
2. Write a clear initial promise and fit definition.
3. Test the definition during real sales and delivery.
4. revise it when customer results reveal a repeated pattern, rather than rewriting it for every prospect.

The last distinction prevents “customer learning” from becoming an excuse for unlimited customization. A one-off request is not necessarily evidence that the offer should change. A repeated need among profitable, successful customers may be.

Qualification should predict success after the sale

Many qualification systems concentrate on whether a prospect can buy: budget, authority, need, and timing. Those questions matter, but a repeatable service also needs to ask whether the customer can be served successfully.

GitLab’s published opportunity-management guidance, for example, asks sellers to investigate the economic effect of the problem, the customer’s desired results, the decision process, the people involved, the severity and duration of the problem, the cost of inaction, and whether a deadline or event makes action necessary. It treats qualification as something that continues through discovery and scoping rather than as a single form completed when a lead enters the pipeline.

A smaller service company does not need to copy an enterprise sales framework. It does need to answer the same underlying questions in language appropriate to its offer:

• Does the customer have the problem the offer solves?
• Is the problem important enough to support action?
• Does the intended result matter to a named person?
• Can the company deliver the result using its standard method?
• Can the customer supply the access, information, decisions, and internal work the method requires?
• Does the expected revenue justify the sales, onboarding, delivery, and support effort?
• Will the customer accept the offer’s boundaries, price, schedule, and division of responsibility?

These questions combine sales fit with delivery fit. That is essential because a deal can be easy to close and still be a bad customer engagement.

There is no universal qualification model or target

A 2025 case study of lead scoring at a business-to-business software company found that a model trained on the company’s own customer relationship management data identified promising leads more accurately than its earlier method. The factors that mattered included lead source, status, product, account type, and expressed interest. The study also noted that academic evidence on successful business-to-business lead-scoring systems remains limited.

The practical lesson is not that every small company needs machine learning. It is that useful qualification criteria depend on the offer, customer behavior, channel, sales process, and outcomes found in the company’s own data.

A high-priced service that requires executive involvement, data migration, and months of implementation should use different criteria from a low-risk service that starts in a day. A referred prospect may arrive with more trust and context than a cold outbound prospect. A regulated customer may require controls that are irrelevant elsewhere. A new offer will rely more heavily on explicit hypotheses; an established offer should increasingly rely on observed results.

For that reason, neither the criteria nor the qualified-fit target should be borrowed uncritically from another company.

Build the offer one-pager from real customer work

The required evidence is a one-page offer definition covering the ideal buyer, trigger, promise, and no-fit signals. To make that document operational, it should also record the conditions and evidence behind those statements.

Begin by examining a complete recent period of serious sales and delivery activity. Include successful customers, disappointing customers, lost opportunities, deals the company declined, and prospects that never reached a decision. If the company has many transactions, use a sample that represents its major channels, customer types, and outcomes.

For each case, collect evidence such as:

• the event that started the buying process;
• the problem described in the customer’s own words;
• the person who owned the result;
• the promised and actual scope;
• customer inputs required for delivery;
• sales time and delivery effort;
• delays, changes, exceptions, and rework;
• gross margin or another useful contribution measure;
• time until the customer received a meaningful result;
• complaints, support effort, renewal, expansion, and referral behavior;
• the reason the customer bought, did not buy, or should not have been accepted.

The purpose is to find combinations, not isolated attributes. “Customers with 50 to 200 employees” may be less useful than “customers that recently appointed an operational owner, have a documented process, and must meet a fixed external deadline.”

Write a promise with conditions

A useful working form is:

For [ideal buyer] facing [problem and trigger], this offer is designed to deliver [specific result] within [relevant period or stage] through [standard scope and method], provided that [customer commitments and operating conditions] are met.

The language should match the strength of the evidence.

If the company has repeatedly produced a result under comparable conditions, it can describe that result and cite its evidence. If the evidence is early, the statement should remain a testable working promise. If the result depends heavily on customer behavior or external events, those dependencies should be explicit.

In the United States, the Federal Trade Commission states that objective advertising claims must have a reasonable basis before they are published, including claims communicated by implication. The appropriate level of support depends on factors such as the type of claim, the product or service, the consequences of error, and what relevant experts would consider reasonable.

That does not mean every customer promise requires a legal document. It does mean that “double revenue,” “cut costs by 30%,” or “deliver in ten days” should not be treated as harmless sales language when the company lacks supporting evidence or has omitted material conditions.

Separate the supporting evidence into three classes:

A promise becomes more credible when the team can explain both why the method should produce the result and where it has already done so.

Turn successful conditions into qualification criteria

Review the strongest customer outcomes and ask what was present before or at the start of the engagement. Then compare those conditions with unsuccessful work.

Qualification criteria commonly fall into five groups:

Write criteria as things that can be observed or verified.

“Committed customer” is vague. “The customer has named the internal owner who will attend the kickoff and approve weekly decisions” is testable.

“Has budget” may also be too simple. A prospect can afford the price but still expect more work than the price covers. The better question is whether the customer accepts the standard commercial terms and whether the company can meet the promise within them.

Define disqualifiers before the next sales conversation

A disqualifier is not merely the opposite of a desirable trait. It is an observed condition that creates an unacceptable likelihood of poor customer results, unreliable delivery, weak economics, legal or ethical risk, or damage to other customers.

Useful disqualifiers often include:

• the primary problem falls outside the offer’s defined use case;
• the prospect requires capabilities the company does not provide;
• the result depends on custom work that would materially change the delivery model;
• the customer will not supply a required owner, data set, access, decision, or approval;
• the requested schedule makes responsible delivery impossible;
• the buyer expects an outcome the available evidence does not support;
• the customer’s security, regulatory, contractual, or geographic requirements cannot be met;
• the expected revenue does not cover the likely cost and risk of service;
• the customer rejects a material boundary of the standard offer;
• the parties’ expectations about responsibility or success remain incompatible after clarification.

Some no-fit signals concern the customer. Others concern the request.

Design Pickle’s public scope documentation demonstrates the distinction. Its help center lists supported creative work, explicitly excludes categories such as software development, advanced web design, research, and copywriting, and states the inputs required for certain work. Its brand-guide documentation, for example, requires customers to provide finalized concepts and text because full brand strategy and copy development are outside that service.

The lesson is not to reproduce that company’s boundaries. It is to make limitations visible enough that sales, customers, and delivery teams can make the same decision before work starts.

Give each no-fit decision a next path

A disqualifier should lead to a defined action:

• Decline: The engagement is outside the company’s capabilities, standards, economics, or risk tolerance.
• Recycle: The prospect appears suitable, but the trigger, owner, information, budget, or timing is not ready.
• Redirect: Another package, partner, or supplier is more suitable.
• Escalate: The exception may be strategically useful, but a named leader must approve it with explicit cost, learning purpose, and delivery ownership.

This process can be represented simply:

flowchart TD
    A[Serious customer enquiry] --> B{Can the standard offer produce the promised result?}
    B -->|No| C[Decline or redirect]
    B -->|Yes| D{Is there an active trigger and responsible owner?}
    D -->|No| E[Recycle for later]
    D -->|Yes| F{Will the customer accept the scope, price, and responsibilities?}
    F -->|No| G[Decline, redirect, or approve an explicit exception]
    F -->|Yes| H[Qualified fit]

In plain language: first test whether the offer can work, then whether the customer is ready, then whether both parties accept the operating agreement. Enthusiasm or available budget should not bypass the earlier decisions.

The finished one-pager should contain:

One page forces choices. Supporting notes, interview records, proof, and scoring guidance can sit elsewhere, but a seller should not need a lengthy manual to explain who the offer serves.

Measure qualified fit without rewarding the wrong behavior

The primary measure is the qualified-fit rate:

The denominator matters. It should not be every name in a marketing database, because many records have not supplied enough information for a fit decision. It should not be only opportunities a salesperson has already decided to pursue, because that hides earlier rejections.

Use a defined event such as “completed first discovery,” “completed application,” or “fit review recorded with all required fields.”

Report the measure separately for referrals, targeted outbound, direct inbound, and any other meaningful channel. Those channels bring different levels of prior trust, information, targeting, and buying intent. Combining them can hide whether the qualification system or the source of demand is changing.

A high qualified-fit rate is not automatically good

The rate can rise for several reasons:

• targeting has improved;
• the promise and ideal buyer are clearer;
• referrals are producing better-aligned prospects;
• sellers are applying the criteria less strictly;
• poor prospects are being removed before the denominator begins;
• the team is qualifying almost everyone to protect pipeline numbers.

The first three may represent improvement. The last three may represent measurement failure.

Qualified-fit rate must therefore be checked against downstream evidence:

This creates a feedback loop. Sales records the expected fit. Delivery and customer-facing teams later record what happened. The company then compares the prediction with the result.

A prospect classified as qualified but requiring extensive custom work is a false positive. A declined prospect that later succeeds with the same standard offer may be evidence of a false negative. Both should lead to review, not automatic rewriting of the criteria.

Document a target from the company’s own baseline

“Target documented” is the appropriate working requirement because there is no defensible universal qualified-fit benchmark.

The right rate will vary with the breadth of the market, quality of targeting, source of leads, price, sales effort, customer readiness, and implementation complexity. Research on lead prioritization also indicates that useful predictive factors can be specific to the company and its data.

A practical target-setting method is:

1. Define the exact decision event and denominator.
2. Calculate the current rate by sales channel and customer segment.
3. Compare qualified and unqualified groups using downstream delivery and financial results.
4. Identify whether capacity calls for more volume or tighter selection.
5. Set a target or target range for a stated period.
6. Record the assumptions behind it.
7. Review it after enough customers have passed through both sales and delivery.

For example, the target might be expressed as:

Increase qualified-fit rate for targeted outbound from the current baseline to the agreed range by the end of the next two complete sales cycles, while keeping delivery exceptions below the separate operating limit.

The exact numbers must come from the company. The second condition prevents the sales team from raising the fit rate by relaxing the definition.

The document should also state who owns the measure. Sales may record qualification, but delivery, finance, and customer success should help validate whether the criteria predict the result they were designed to predict.

Common failure modes and the decision to move on

This work often looks finished before it is useful.

The promise is a slogan. “We help companies grow” does not identify the customer, problem, result, method, conditions, or boundary. It cannot guide a fit decision.

The ideal buyer is only a demographic profile. Industry, revenue, location, and headcount may help find prospects, but they do not establish that the problem is urgent, the use case matches, or the customer can participate in delivery.

Qualification means ability to pay. A solvent buyer can still be a poor operational fit. The company must also test scope, readiness, responsibilities, risk, and likely delivery economics.

The criteria describe feelings rather than evidence. Words such as “serious,” “innovative,” “collaborative,” and “high quality” permit each seller to make a different decision. Observable actions and conditions are more reliable.

Disqualifiers are kept private. If the website and sales conversation imply broad capability while the delivery team knows the actual limitations, the company is creating avoidable expectation gaps.

Every negative decision is permanent. Treating “no active trigger” as the same as “cannot be served” wastes future demand. Permanent no-fit and temporary not-ready need separate categories.

The founder can override the system without cost. Strategic exceptions can produce useful learning, but they should be named as exceptions. The company should record why the work was accepted, what extra cost or risk it carries, who owns it, and whether it is intended to change the standard offer.

Sales success is not checked against delivery. A qualification process is not validated because sellers use it or because qualified-fit rate rises. It is validated when customers classified as good fits receive the promised result more reliably and with better delivery economics than customers that lack those conditions.

Before the company depends on the one-pager, the following should be true:

• sales and delivery use the same definition of a good customer;
• each qualification criterion can be supported with observable evidence;
• the promise is matched to proof and states its material conditions;
• no-fit signals are separated into permanent, temporary, and redirect categories;
• the decision and reason are recorded consistently;
• exceptions require explicit approval;
• qualified-fit rate has a precise denominator, owner, baseline, target, and review date;
• downstream results are used to test whether the classification was correct;
• repeated exceptions trigger an offer review rather than quiet customization.

At that point, the company can answer a more useful question than “Can we sell this customer something?”

It can ask:

Can we make this promise to this customer, keep it through our standard delivery system, and earn an acceptable return without redesigning the offer?

When the answer can be reached consistently before the sale—and verified after delivery—the offer is becoming repeatable.

Sources

Open research

• Shamsuzzoha, Ahm; Blomqvist, Hannele; and Takala, Josu. “Service Productisation Through Standardisation and Modularisation: An Exploratory Case Study.” International Journal of Sustainable Engineering, 2023.
• Terho, Harri; Haas, Alexander; Eggert, Andreas; and Ulaga, Wolfgang. “It’s Almost Like Taking the Sales Out of Selling—Towards a Conceptualization of Value-Based Selling in Business Markets.” Industrial Marketing Management, 2012.
• Heikka, Eija-Liisa. “Value Propositions in KIBS: How to Facilitate the Communication of Value?” Knowledge and Process Management, 2023.
• Homburg, Christian; Droll, Mathias; and Totzek, Dirk. “Customer Prioritization: Does It Pay Off, and How Should It Be Implemented?” Journal of Marketing, 2008.
• González-Flores, Laura; Rubiano-Moreno, Jessica; and Sosa-Gómez, Guillermo. “The Relevance of Lead Prioritization: A B2B Lead Scoring Model Based on Machine Learning.” Frontiers in Artificial Intelligence, 2025.

Primary and official sources

• Federal Trade Commission. “FTC Policy Statement Regarding Advertising Substantiation” and “Advertising and Marketing.”
• Christensen Institute. “Jobs to Be Done Theory.”
• GitLab. “Lead Lifecycle Management” and “MEDDPPICC,” GitLab Handbook.
• Design Pickle. “Official Scope of Design Service” and “Scope: Branding & Brand Guides.”