Decide What to Stop, Continue, and Convert

Task

Decide which consulting offers to stop, continue, or convert.

Summary

Rationalize the existing offer portfolio so the company can concentrate on the selected transition.

Choose Which Consulting Offers to Stop, Keep, or Turn Into Products

Task ID: S1-14

A consulting firm should not keep every offer that has produced revenue. It should decide which offers deserve continued investment, which should leave the portfolio, and which contain repeatable work that can become a clearer service or product. The decision requires evidence about demand, customer results, delivery cost, repeatability, strategic fit, and founder dependence—not revenue alone.

The portfolio problem hiding inside successful sales

A founder-led consulting company can accumulate offers without consciously designing a portfolio.

One customer asks for an assessment. Another needs implementation help. A referral produces a training engagement. A long-standing customer requests an unusual integration. Each project brings in revenue, so each service remains available. Before long, the company appears to offer strategy, implementation, customization, support, training, and several loosely related specialties.

The resulting portfolio may look healthy because customers have bought everything in it at least once. But past sales do not prove that every offer should remain available.

Some offers generate strong fees but consume too much senior time. Some appear profitable until presales, revisions, support, and project management are included. Some delight one unusual customer but have little repeat demand. Others are modest consulting engagements that reveal a recurring problem, a common workflow, and a potentially valuable product.

The operating principle is straightforward:

Keep an offer because the evidence supports its future role, not merely because someone once bought it.

That evidence must cover more than sales. Research on service productivity shows why. A 2024 meta-analysis combined 77 articles, 81 independent samples, and 30,238 participants and concluded that service quality and internal efficiency should be evaluated together. Measuring only cost can reward work that damages customer outcomes. Measuring only satisfaction or revenue can preserve work that is economically unsustainable.

This makes offer rationalization a business-design decision, not a housekeeping exercise. The company is choosing where it will concentrate sales effort, delivery capacity, management attention, and future product development.

What the research says about services worth keeping

Professional services are difficult to compare because they are often described differently for each customer. The proposal may change, the deliverables may change, and the delivery team may quietly adapt the process. Two projects with different names may contain almost the same work, while two projects sold under the same offer may have completely different economics.

Service productization research addresses this problem by making an offer easier to describe, buy, deliver, and invoice. A multi-case analysis by Härkönen, Tolonen, and Haapasalo found that service productization involves systematizing and making both the offer and its delivery process more concrete. It is not simply a marketing exercise; it connects the commercial definition of the service to the technical and operational work required to deliver it.

A later professional-services case study examined standardization, modular service platforms, and service automation. Its central idea was not to eliminate every customer-specific element. It was to define stable service components so that the provider could assemble an appropriate solution without redesigning the whole engagement each time.

This distinction matters when deciding what to convert. A service can remain flexible while becoming more repeatable. For example, a security assessment might have:

• a standard intake and evidence request;
• a defined assessment method;
• a fixed set of core tests;
• optional modules for different environments;
• a consistent report structure;
• clear remediation and follow-up options.

The customer still receives work suited to its circumstances. The company no longer starts with a blank document and an empty project plan.

Research connecting productization with portfolio management makes a related point: portfolio decisions work better when the company has a consistent structure for describing what it sells. Without that structure, leaders compare names and revenue categories rather than comparable commercial and delivery components.

The financial analysis also needs more precision than invoice value minus contractor expense. Time-driven activity-based costing estimates the cost of work from the resources used, their practical capacity, and the time required by different transaction or project characteristics. The approach is useful when standard work and complex exceptions consume very different amounts of effort.

For consulting offers, this means counting more than scheduled delivery hours. The economic record may also need to include:

• qualification and proposal work;
• unpaid discovery;
• project management;
• specialist and founder review;
• revisions caused by unclear scope;
• customer meetings;
• travel and third-party tools;
• non-billable remediation;
• post-project questions and support;
• collection delays and commercial risk.

An offer with a high price can still be weak if it repeatedly consumes scarce capacity, creates delivery volatility, or depends on the founder to rescue it.

At the same time, a smaller portfolio is not automatically better in every situation. Research covering thousands of public companies found that narrower product offerings were generally associated with better performance in the sample, but broader operational scope could be useful in unpredictable environments. The implication is not “cut everything.” It is to preserve deliberate breadth where it provides resilience, learning, account access, or a meaningful response to uncertain demand.

The goal is therefore coherent variety, not minimal variety. Every continuing offer should have a reason to exist.

The difference between stop, continue, and convert

The three decisions are not rankings from worst to best. They describe different roles in the future business.

Stop means stop selling, not abandon customers

Stopping an offer normally begins with removing it from future sales. It does not necessarily mean terminating work immediately.

Existing agreements, promised deliverables, customer data, support commitments, warranties, subcontractor arrangements, and transition needs still require attention. The company should identify each active obligation, assign an owner, and decide whether to complete, migrate, refer, renegotiate, or wind it down.

A stopped offer may still produce occasional requests. The leadership team therefore needs an exception rule. Without one, every large prospect or friendly referral can quietly restore the service.

A useful rule is:

An exception requires a named strategic reason, explicit economics, an accountable approver, and a clear statement that the work does not restore the offer to the standard portfolio.

Continue does not mean leave unchanged

A continuing offer may be valuable precisely because it requires human judgment. Complex diagnosis, executive advice, incident response, negotiation, architecture, and high-risk implementation may resist full standardization.

That does not excuse vague scope.

A continuing consulting offer should still have a recognizable customer, problem, result, entry criteria, delivery method, pricing logic, and set of exclusions. The company should know which elements require expert judgment and which elements are merely inconsistent because no one has documented them.

The research on service productivity is important here. Efficiency improvements should not be made by stripping away the parts of the work that create the result. Cost and quality have to remain connected.

Convert does not necessarily mean build software

Conversion has several possible forms:

• a fixed-scope diagnostic;
• a standard workshop;
• a modular implementation service;
• a recurring managed service;
• a training or certification program;
• a template, dataset, or licensed method;
• a software-assisted service;
• a standalone software product.

The appropriate form depends on what repeats.

If the same customer problem repeats but the work still requires substantial judgment, the next step may be a productized service. If the same data transformation, calculation, monitoring task, or workflow repeats with limited variation, software may eventually carry more of the work.

The company should not begin with “Which service can we turn into software?” It should begin with “Which valuable customer problem, decisions, inputs, and delivery steps repeat with enough consistency to justify a different delivery model?”

Build the evidence before making the decision

A portfolio review fails when it begins in a meeting with a list of offer names and a request for opinions.

The work should begin by reconstructing what the company actually sold and delivered. In a founder-led firm, the accounting categories, proposal names, and website descriptions may not match. Several offers may be disguised as custom projects, while one named package may contain several fundamentally different engagements.

Create a normalized offer inventory

Review proposals, statements of work, invoices, project plans, time records, support records, and customer notes. Group engagements that share the same basic pattern:

customer → problem → promised result → delivery method

Do not group projects merely because they used the same technology or employee.

For each offer, record enough history to represent normal demand and delivery conditions. Twelve to twenty-four months is often a practical starting window for an active consulting business, but longer sales cycles, seasonal work, or a small number of high-value engagements may require a longer period. Recent structural changes—such as a new team, pricing model, regulation, or delivery method—may justify separating older and newer records.

Evaluate each offer on comparable evidence

The decision record should cover the following dimensions.

This structure deliberately combines performance against expectations, performance over time, strategic alignment, stakeholder response, competitive alternatives, and the effect of removing the offer. Those questions prevent a single metric from deciding the portfolio.

Calculate economics at the offer level

A practical starting calculation is:

Offer contribution = recognized revenue − direct delivery labour − subcontractors − offer-specific tools and travel − material presales, onboarding, support, and remediation effort

Labour should use a loaded cost appropriate to the decision, not a nominal hourly wage or the amount billed to the customer. The calculation should also make capacity visible. One hundred hours of a readily available delivery role and one hundred hours of scarce founder or specialist attention do not create the same constraint.

Precision should match the importance of the decision. A company does not need a perfect cost model before acting. It does need a model accurate enough to distinguish genuinely attractive work from offers whose apparent profit depends on ignored labour.

Separate facts from interpretations

For every offer, label the information as one of four types:

• Observed fact: recorded revenue, hours, wins, losses, cycle time, or customer feedback.
• Calculated result: contribution, average delivery time, or variance.
• Interpretation: why the offer performs as it does.
• Hypothesis: what could improve if scope, price, staffing, or delivery changed.

This separation is especially important when data is sparse. “Customers value this” is an interpretation unless it is supported by purchases, references, continued use, outcomes, or credible feedback.

Use a staged decision process

flowchart LR
    A[Normalize the offer inventory] --> B[Collect sales, outcome, delivery, and cost evidence]
    B --> C[Compare offers on common criteria]
    C --> D{Evidence supports a decision?}
    D -->|Yes| E[Stop, continue, or convert]
    D -->|No| F[Run a limited test or collect missing data]
    F --> C
    E --> G[Assign owner, actions, date, and measures]
    G --> H[Review results and exceptions]

The process first creates a comparable inventory, then assembles evidence, makes the portfolio decision, and assigns implementation work. Where the evidence is genuinely inadequate, the company runs a bounded test rather than leaving the offer permanently undecided.

Use a stop, continue, or convert matrix

The matrix is the decision artifact. It should be brief enough to use but detailed enough to show why each decision was made.

A practical structure is shown below.

The entries above are placeholders, not recommended decisions. The matrix must be completed from the company’s own evidence.

A weighted score can help organize the discussion, but it should not make the decision automatically. Scores conceal important differences. An offer with excellent margins but serious delivery risk is not equivalent to an offer with moderate margins and dependable customer outcomes, even if both receive the same total.

Decision rules are more useful than a single composite score.

A strong stop case

Stopping is usually supported when several of the following are true:

• qualified demand is limited or declining;
• the customer and problem differ from the company’s intended direction;
• customer outcomes are inconsistent;
• margin remains poor after plausible changes;
• the work creates frequent exceptions or disputes;
• scarce people are required for routine delivery;
• little knowledge, capability, or customer access is reusable;
• no credible conversion path exists.

One weak quarter should not automatically eliminate an offer. The pattern matters, as do market conditions and data quality.

A strong continue case

Continuing is usually supported when:

• a recognizable group of customers repeatedly buys;
• the business problem is important and clear;
• customers receive a dependable result;
• delivery economics are acceptable;
• the work strengthens a useful capability or customer relationship;
• customization reflects necessary judgment rather than uncontrolled scope;
• the company can explain why the offer belongs in the portfolio.

A continuing offer may still require repricing, tighter qualification, standard modules, a new staffing model, or clearer boundaries.

A strong convert case

Conversion is usually supported when:

• demand and customer value are already credible;
• the same problem appears across multiple customers;
• substantial inputs, decisions, tasks, or outputs repeat;
• scope variation can be expressed as a manageable set of modules;
• knowledge can be encoded without destroying the result;
• the new delivery model can reduce delays, variation, or scarce-person dependence;
• customers can understand and buy the converted offer;
• the likely economics justify the investment.

Conversion should usually begin with the smallest repeatable form. Standardize the diagnostic before building a complete platform. Automate a proven workflow before building a broad software suite. Test whether customers will buy the new form before treating the conversion as complete.

What Basecamp’s history illustrates

Basecamp provides a useful public example of both conversion and portfolio focus.

According to the company’s official history, 37signals began as a web design firm. Its team built an internal project-management tool because coordinating a growing number of client projects had become difficult. Clients saw the tool and asked whether they could use it. After Basecamp was released and began producing more revenue than the web design business, the company stopped doing web design and concentrated on the product.

The important lesson is not that every consulting firm should abandon services. The sequence matters:

1. The team experienced a repeated operational problem.
2. The tool solved real work inside the service business.
3. Customers recognized the same problem.
4. The company released a product.
5. Revenue evidence supported the decision to leave the original service.

The product did not begin as a speculative attempt to “turn consulting into software.” It grew from a recurring problem and a working solution.

The same company later narrowed parts of its product portfolio. When it stopped offering the standalone Campfire product to new customers, it explained that the capability had been incorporated into Basecamp and that existing customers could continue using the service.

That illustrates a second distinction: stopping new sales, consolidating capability, and shutting down existing service are separate decisions. A company can remove an offer from its future portfolio while managing customer continuity deliberately.

Basecamp is a private company, so its public account does not provide the detailed project economics required to reproduce its decision. The case is useful because it shows the logic of evidence-led conversion and focus, not because it supplies a universal formula.

What completion looks like

A portfolio decision is complete when the matrix changes behaviour.

A document that labels every offer but leaves the website, proposal library, referral conversations, pricing, staffing, and delivery plans untouched is not offer rationalization. It is classification.

For each stopped offer, the company should have:

• a date after which it will no longer be proposed;
• an active-customer transition plan;
• an exception policy;
• updated sales and marketing materials;
• an owner for remaining obligations;
• a plan for released people and capacity.

For each continuing offer, it should have:

• a named customer, problem, and expected result;
• defined entry and qualification criteria;
• scope boundaries and exclusions;
• pricing logic;
• an accountable owner;
• delivery and outcome measures;
• a list of known improvements.

For each converted offer, it should have:

• the repeated customer problem being addressed;
• the standard core and optional modules;
• the parts that still require human judgment;
• the parts to document, automate, or move into software;
• a limited pilot;
• investment and stopping limits;
• measures for sales, delivery, customer results, and economics.

“Portfolio decision made” is therefore a useful operating state, not a universal industry benchmark. A small firm may decide the whole portfolio in one review. A larger company with several markets or delivery teams may need decisions by segment, region, or service family. Poor historical data may require provisional decisions and explicit tests.

Offer rationalization can be measured through a small set of indicators:

The measures should be reviewed together. A converted offer that lowers delivery effort but weakens customer outcomes is not an improvement. A continuing offer with good margin but rising founder dependence may be consuming the capacity needed for the next stage of the company.

Before the company relies on the result, the following should be true:

• leadership can state which work the company will no longer pursue;
• sales can recognize and qualify the offers that remain;
• delivery teams understand the standard work and permitted variation;
• active customers affected by a stop decision have a responsible transition path;
• conversion candidates have bounded tests rather than unlimited development commitments;
• exceptions are visible and approved;
• the financial and customer evidence is credible enough for the size of the decision.

The final result is not merely a shorter service list. It is a portfolio in which each offer has a supported role: stop work that distracts or destroys value, continue work that deserves expert delivery, and convert work where repeated customer value can be delivered more clearly and consistently.

Sources

Primary sources

• Basecamp, “Where we came from,” official company history.
• Basecamp, “A note about Campfire,” official product-retirement notice.

Open research

• Hofmeister, Johannes, Dominik K. Kanbach, and Jens Hogreve, “Measuring and managing service productivity: a meta-analysis,” Review of Managerial Science, 2024.
• Härkönen, Janne, Arto Tolonen, and Harri Haapasalo, “Service productisation: systematising and defining an offering,” Journal of Service Management, 2017.
• Shamsuzzoha, Ahm, Hannele Blomqvist, and Josu Takala, “Service productisation through standardisation and modularisation: an exploratory case study,” International Journal of Sustainable Engineering, 2023.
• Lahtinen, Niko, Janne Härkönen, and Harri Haapasalo, “Productization as a link to combining product portfolio management and product family development,” Procedia CIRP, 2022.
• Kaplan, Robert S., and Steven R. Anderson, “Time-Driven Activity-Based Costing,” Harvard Business School Working Paper, 2003, and Harvard Business School Working Knowledge overview, 2007.
• Kovach, Jeremy J., et al., “Firm performance in dynamic environments: The role of operational slack and operational scope,” Journal of Operations Management, 2015.