Ask the owner of a product business what it does, and the answer is usually a verb: we make, we build, we supply. Ask how much of its work is now organised as separate projects, each with its own budget, timeline and leader, such as product launches, customer installations, new supplier set-ups and range changes, and the answer may be most of it. Those two answers describe different businesses.
Many businesses drift from making things to coordinating others who make things. Production moves to contract manufacturers. Installation moves to subcontractors. Freight, warehousing and even design move outside. Each decision is sensible on its own numbers. Together they change what the business is: from a producer that gets better through repetition to an orchestrator of projects and suppliers whose success depends on choosing well and coordinating well.
This article explains what changes in that shift, where margin and learning go, what new risks arrive, and how to decide which capabilities to keep. It is not an argument against outsourcing. It is an argument for knowing what kind of business you have become.
A historical example
In 2005, The Economist reported on companies that had moved deliberately from production towards coordination. A global footwear brand, it noted, no longer made shoes but managed footwear projects. A global beverage company had handed much of its bottling and marketing to others. A German car maker treated each new vehicle platform as a separate project. BP had reorganised its exploration division as a portfolio of largely independent projects, which the company called an asset federation, and the report noted that project managers then had to build their own self-sufficient teams rather than rely on head office. Siemens, having found that about half its turnover came from project-like work, calculated that completing all of it on time and on budget could add about €3 billion to profit over three years, a company estimate rather than a reported result.
The specific figures are now historical. The structural point is not: when the repeatable part of a business is contracted out, what remains is largely a portfolio of temporary undertakings, and those have different economics.
Writing in Cost Engineering in 2003, Nick Lavingia argued that a company which consistently selects the right projects and executes them well improves its return on capital, and that the difference can separate a profitable company from a takeover target. For an orchestrating business, that capability is most of what it has.
What changes
Moving work from a standing function into projects or outsourced arrangements does three things that rarely appear in the proposal approving it:
- Capability becomes a series of engagements. A function that ran continuously built up judgement about suppliers, failure modes and estimates. A sequence of projects or outsourced orders builds up that judgement only if something outside the individual jobs is designed to hold it.
- Fixed costs become variable. That looks like flexibility on the way in. On the way out, the business may find it has lost a floor of capacity that absorbed variation without negotiation.
- Reliability changes. Repeated work tends to become more predictable. Project work is generally less predictable, with more variation in cost and timing. A business that converts work into project form inherits that profile, and few business cases say so.
Two halves of a business
It helps to think of a business as having two halves:
- The repeated half, where doing something many times makes you better at it: manufacturing, assembly, routine service. Value accumulates through practice and improvement.
- The orchestrating half, where value comes from selecting, coordinating and integrating: choosing products, designing, sourcing, managing suppliers, assuring quality, serving customers.
Two questions follow. Which half does the customer pay a premium for? And which half did the business keep? If customers pay for the quality of the making and the making has been contracted out, the business may be holding the lower-margin half of its own value chain. If customers pay for design, brand, integration or service, keeping those and outsourcing production may be exactly right.
Learning stops compounding unless you build for it
In a producing business, learning accumulates almost automatically because the same people do the same work repeatedly. In an orchestrating business, each project or outsourced order can start from scratch. Estimates do not improve, supplier problems recur and lessons leave with the people who learned them.
Capturing learning needs a deliberate mechanism, funded from a standing budget rather than from individual projects, which have no incentive to pay for the next project’s benefit:
- Short reviews after each significant project, feeding estimates and checklists.
- Supplier performance records covering quality, delivery and responsiveness.
- Standard playbooks for recurring work such as product launches or supplier onboarding.
- Retained technical expertise able to challenge suppliers rather than simply accept their advice.
Selection becomes the core capability
For an orchestrating business, the most valuable capability is often selection: choosing which products to develop, which customers and projects to take on, which suppliers to use and on what terms. Selection is cheap to do badly and expensive to do well, so businesses tend to underinvest in it and overinvest in more visible activities. Give the people who make selection decisions good information, clear criteria and the standing to say no.
What orchestration brings with it
- Dependence you cannot easily see. Suppliers now hold process knowledge the business used to hold. How much that matters depends on how easily they could be replaced.
- A handover at the end of every job. Each project ends, and someone must receive what it produced.
- Capability that lives in individuals. Skills built on projects may leave with the people who hold them unless the business captures them.
- Resourcing arithmetic. Several concurrent projects drawing on the same few people can look fine on paper and be impossible in practice.
Manage suppliers as part of the business
An orchestrator’s performance depends on its suppliers’ performance, so supplier management becomes a core activity rather than a purchasing task. Useful practices include sharing forecasts and plans so suppliers can prepare, agreeing clear specifications and acceptance tests, visiting suppliers’ sites, reviewing performance together using simple scorecards, and working with suppliers on recurring problems rather than simply rejecting faulty goods. Good suppliers value customers who plan well and communicate clearly, and often give them priority when capacity is tight.
Contract for the orchestrator’s needs
Supplier contracts written for one-off purchases often miss what an orchestrator needs over time. Useful terms include clear quality standards and acceptance tests, a requirement to notify the business before changing materials, processes or sub-suppliers, capacity commitments for peak periods, access to the supplier’s site for audits, and agreed procedures for handling defects. These terms protect the business’s ability to deliver to its own customers, which ultimately depends on how its suppliers behave. Take advice on contract wording for significant agreements.
Keep the ability to bring work back
Outsourcing decisions should remain reversible where possible. That means keeping complete, current specifications and drawings, owning the tooling and moulds the business has paid for, holding copies of test methods and quality records, and knowing what it would take to qualify another supplier or bring work in-house. Without these, a supplier relationship that starts as a choice can become a dependency the business cannot escape. Check contracts for ownership of tooling and intellectual property, and take advice where significant value is involved.
Three tests for each line of activity
Most businesses are producers in some areas and orchestrators in others. Apply these tests line by line:
- The repetition test. Would doing this many more times make us materially better at it? If yes, does anything in our current structure actually capture that improvement? Repetition without capture is the most expensive combination: the business pays the cost of repetition without collecting the benefit.
- The margin test. What does the customer pay a premium for here, the making or the orchestrating? Which did we keep?
- The reliability test. If this activity now runs as projects or outsourced orders, have our plans and business cases been updated to assume project-style variability, or do they still assume the reliability of an in-house function?
A fourth question underlies all three: what could we still do next Monday if our largest supplier stopped work? The honest answer measures how much of the business you still hold.
Decide what you will not outsource
The pressure to outsource one more thing never stops, because each case looks marginal on its own. The cumulative effect is not marginal, and individual cases never provide a natural stopping point. The stopping point has to be a deliberate statement of which capabilities the business will keep at any price, usually those that customers pay for and that would be hardest to rebuild.
A worked example
This is an illustration. An Australian consumer products business once made most of its range in its own workshop. Over ten years, it moved about 80% of production to contract manufacturers, mostly overseas. Its team of twelve now handles design, sourcing, quality, marketing and customer service.
Problems have crept in. Three of the last five product launches were late. Quality problems recur each time a new manufacturer is used. Estimates for launch costs and timing have not improved. One manufacturer now makes about 60% of the business’s volume.
The owner applies the tests:
- Repetition: the business launches about six products a year, but nothing captures what each launch teaches. The owner funds a launch playbook, a short review after each launch and a supplier scorecard from the operating budget.
- Margin: customers pay mainly for design, durability and service. The business decides it will never outsource product design or quality engineering, and hires a quality engineer who can audit manufacturers and challenge their advice.
- Reliability: launch plans assumed the reliability of the old in-house workshop. New plans include buffers for first-production samples and shipping delays.
- Residual capability: if the main manufacturer stopped, the business could not supply 60% of its range. It begins qualifying a second manufacturer for its three best-selling products.
Over the next four launches, three are on time, first shipments have fewer defects and launch estimates become more accurate. The business still orchestrates most of its production, but now it governs itself as an orchestrator.
How this applies to a small Australian business
Many small businesses orchestrate: product companies using contract manufacturers, builders managing subcontractors, installers coordinating trades, and service firms using contractors. Practical steps:
- Classify each line of activity as repeated or project-based, made in-house or coordinated.
- Identify what customers pay a premium for, and keep it.
- Build a learning mechanism for repeated projects and fund it centrally.
- Keep enough expertise to challenge suppliers.
- Track supplier concentration and qualify alternatives for critical items.
- Update plans to reflect the variability of outsourced and project work.
- Protect your designs and know-how with appropriate agreements, taking legal advice.
The articles on make or buy and local or offshore manufacturing cover related decisions.
Signals worth watching
- A rising share of work organised as projects or outsourced orders, with no learning mechanism.
- Estimates for repeated projects not improving.
- Rising dependence on one supplier.
- Staff unable to judge whether a supplier’s technical advice is right.
- Suppliers beginning to offer the coordination service themselves.
- People who select work rewarded for volume rather than outcome.
Common mistakes
- Outsourcing one more thing at a time without a stopping point.
- Keeping the low-margin half of the value chain.
- Assuming in-house reliability from outsourced or project work.
- Letting learning leave with each project or person.
- Underinvesting in selection.
- Losing the expertise needed to challenge suppliers.
Frequently asked questions
Is orchestration a weaker business model? Not necessarily. Coordination is a real and scarce capability, and many successful businesses are orchestrators. The risk is becoming one without realising it and continuing to manage as a producer.
How much in-house capability should we keep? Enough to specify, assess and challenge what suppliers do in areas that matter to customers, and to recover if a key supplier fails.
Should we bring production back in-house? Sometimes, when volumes, quality problems, lead times or strategic value justify it. Use the make-or-buy analysis on current numbers, and remember that rebuilding a capability takes time and investment.
How do we fund a learning mechanism in a small business? Keep it simple: a short review template, a supplier scorecard and a shared folder of lessons, with a small amount of someone’s time allocated each month.
Questions to ask
- For each of our main lines, are we the producer or the orchestrator?
- What does the customer pay a premium for, and did we keep it?
- What could we still do next Monday if our largest supplier stopped?
- Where does learning from one project reach the next estimate, and who pays for it?
- Have our plans been updated to assume project-style variability?
- What will we not outsource at any price, and why?
Bringing it together
The shift from making to orchestrating is rarely a single decision. It is a sequence of sensible outsourcing choices that eventually change what the business is. Orchestration can be a strong model, but it moves where margin sits, stops learning from compounding unless you build for it and brings new dependencies and variability. Say clearly, line by line, which business you are now, keep what customers pay for, build a mechanism that carries learning forward, invest in selection and decide what you will never outsource.
Source: KEVOS notes, drawing on a 2005 report in The Economist on project-based companies and Nick Lavingia’s 2003 article in Cost Engineering. Historical company figures are as reported at the time. Examples in this article are illustrations.
