XX.XI. Production and Supply Chain

The 11th system architecture pattern: Designing the production process
Dr. Albert Künstler

roduction and supply chain is the third component of the Business Model architecture, and it is the most hidden and conservative part of your processes. Your marketing and sales activities are public, your service is exposed to every customer, including the agents of your competitors who want to borrow your fresh ideas. But production and supply happens behind the scenes and lets you keep your secrets.

Production is where the wizardry happens, where genius architects and designers work to amaze the customers with mind blowing creations, where secret ingredients are poured into the cauldrons with magical potions and viands, where gigantic engines grind day and night long in clubs of steam.

Once everything is ready, the product is fulfilled to the customers so they can enjoy the experience in their Service cycle. Even in pure “service” business, production chains exist, but they create tools and artifacts that are utilized by the employees who actually serve the customers. Thus both product and service business have their production and supply chains, so we will describe the principles without making very much difference between products bought and used by the end customers, and products used by the service company employees to satisfy the requests of their clients.

Examples: For us there is not much difference between the hammer bought by the end user (a product), and the hammer bought by the repairman who charges the house owners (service) for the minor repairs with the same hammer. Of course the professional worker may buy more advanced tools but the production and supply chain for both will be quite similar.

Similarly, the banks do not sell physical products to the end customers, they provide financial services, however their employees use a lot of software and tools that work behind the scenes, so production and supply chain of banks will include not just all processes required for money transactions, but also processes of creating, upgrading and utilizing the software by their clerks, service agents, and sales people.

For a dancer or an illusionist (pure service, no products), production and supply chain will include the stage props, the costumes, and of course the processes of training and rehearsals.

Basically if during the service cycle we were talking about the application chains, i.e. steps how the product or service is used during the Service cycle. In the production and supply chain we will be talking about the replication chains, i.e. steps required to make the next service cycle possible, replicating (manufacturing) the product or replicating (preparing) the act of service.

The most common architecture for the production and supply chain is the Assembly chain that is already covered in one of our previous guides. So we assume you are already familiar with the elements, flaws and illnesses of that architecture and will not repeat ourselves. Instead we will add more important details in the business environment.

Research landscape of a production and supply chain

Product design cycles

Frankly speaking, you don’t necessarily need to invent new products. There are thousands of businesses that sell common products. They don’t reinvent the wheel, but inherit the replication chains from one another. Very often when a new business is started, the founders will do what they did before as employees of another company. So they will bring their customer base, their skills and production processes to their new business. If they lack a specific skill they will hire someone who had a certain job before and this person will bring the missing processes and know-hows to their company. At a larger scale, a factory who wants to penetrate the new markets will hire an industrial engineering company to design a new production line which includes the best processes and tools.

However, in a competitive landscape nowadays, more and more businesses are under pressure and they have to introduce new versions of their existing products. Only a small fraction of businesses are able to invent something completely new but many more companies need to constantly update their production to stay in business. In any case, product design becomes an important part of production for many companies.

Product design cycles include the processes of understanding what customers want, we covered them in the Demand analysis and Productdevelopment guides.

This is the basic process:

  1. Understand customer needs (customer genotype) through Demanddiscovery techniques and Marketing cycle.
  2. Prioritize solutions (product genotype): select Take-my-money and Must-have needs of the customer and decide which of them should be satisfied at which level (State-of-the-Art vs Good Enough vs Quick & Dirty vs Zero).
  3. Design the selected solutions at the decided above level, which means designing the Service cycle experiences (what uses to to satisfy the need and how they use the product for that). These experiences will require certain product features.
  4. Design the replication chains (supply and production recipes) that are required to implement these features in the product.
  5. Create a small pilot batch of products following the new recipe. Run a closed beta test with a limited number of users and see if the new version actually works correctly and solves the intended problems. Rework the defects and put them into serial production.

The output of this process is the recipe for a new version of the product, a new version of the replication chain. On our Business Model diagram, production flows from right to left, so we will use the same notation to make it consistent with the previous chapters, the next elements will also be depicted and discussed from right to left.

Production and Supply Chain.


The first part of the replication chain is supply. The parts, components that your business buys from suppliers. When designing the supply chain in the Product design cycles above, you need to make decisions on the components and ingredients. Do you build or buy them? If you buy, do you buy from the mass market or from a very specific supplier. When you buy them from a specific supplier you may expect better quality, uniqueness, and customization, but you will fall into dependance and shall order large enough quantities to keep the supplier interested in partnership.

How do you decide? That decision is determined by the class of components and on your prioritization.

Not all the components are the same, in the supply chain the ABC classification is widely used. A items are the most important, these are critical and expensive components. You may have only 10% of those in your inventory, but they are responsible for 60–80% of value. B items are less critical, representing 20% of your inventory they are responsible for 15–25% of value. The rest are C items, they are numerous and less significant, while representing up to 70% of your inventory, they generate only 10% of value or less.

Let’s say you decided to build a Placeholder for a certain experience. If so, you may pick anything. The cheapest components from the mass market or even a fake, foam or paper prototypes. The Placeholders are not intended for real use. They are required for prototyping or for experimentation with the limited number of customers.

What if you decided to build a Quick & Dirty solution for a particular Must-have need of your customers? Do you buy or build the components, if you buy, will you buy mass market or exclusive, cheap or expensive?

Most likely, the best strategy is to buy A items from the mass market in the higher price range, and to buy excel volumes of B and C items from the mass market, low price range. You overpay for A items, because they are critical. So you lose your margins, but you are sure that your product is reliable to meet the Must-have expectations of the customer. B and C items are not so critical, you may save on them, however you should be ready to spend time on additional acceptance of materials testing, and throw away the flawed components and ingredients, so usually you will buy 10 items when you need 8.

Now how about a Good Enough solution? Here your product becomes more mature, but the whole idea of the Good Enough level is to copy, not invent. You don’t need to build something unique here, but you need to be good enough for your clients, not worse than your competitors. At the same time you develop your product and your business and at this point you want to make more margins. That is why you may start thinking of moving up the supply chain, and build several critical steps in-house instead of buying some turnkey critical components. You will still need to buy high quality A items from suppliers but you may assemble lower lever components instead of buying already assembled parts. You may also abandon the mass market and find more exclusive suppliers that already provide components to your competitors. B and C items may be still bought on mass market, but as you establish a higher level, you most likely will not buy the cheapest items, but balance their cost and level of defects caused by these components. That will give you a Good Enough solution that has both a good quality and good margin.

Finally with the State-of-the-Art solutions for Take-my-money needs of your customers, you are in the invent and build mode. You need the best architects and engineers, secret ingredients, and unique processes, very different from your competitors. You will have very customized suppliers of raw components and will have a lot of assembly in house. As you develop a unique solution you will radically change the supply chain. For example you may eliminate some traditional expensive components to drastically reduce costs or significantly improve other elements of the C.U.S.T.O.M.E.R.experience by adding new components that none of your competitors yet employ. In fact if you want to succeed, you need both: reduce costs and increase value. So think about removing the most expensive A items in your designs and substituting them with your own secretive home-built parts, produced with a new set of A class raw materials acquired from new exclusive suppliers.

We should repeat that not all of your features should be State-of-the-Art. So it is never about inventing and reengineering the whole thing. One product may have one State-of-the-Art function that serves a particular Take-my-money need of the customer, but also several Good Enough functions for the most important Must-have needs, and even several Quick-and-Dirtyfunctions. Thus your supply chain should be aligned with these decisions.


Well, you have the right components and ingredients from your suppliers, now you need to assemble them into a finished product. As we discussed in a separate Assembly chain guide, you have the main chain and the branches that bring supplies to the main process.

The assembly process adds the most value. This process is always governed by the classic Cost-Time-Quality triangle, which tightens operations from all sides with the requirement to do everything faster and cheaper while maintaining high quality standards. The triangle dilemma says that you can’t limit all three, and usually if you tighten the quality constraint it will inevitably require more time and budget, while setting a fixed budget will likely require more freedom in timing and possibly will affect the quality.

As such, assembly management requires two things: a constant focus on Plans and schedules to juggle existing assets to meet incoming order flow and deadlines, and a continuous focus on Effectiveness to maintain quality without sacrificing cost.

To reflect reality, one more dimension needs to be added to the triangle. Demand was omitted from classical management theory, which was born in times of scarcity. However, demand becomes more important now in the age of overproduction and continuous innovation.

In times of scarcity, demand has always been insatiable, and orders have always been backed up by real customers waiting for the products to be delivered. But now hundreds of thousands of innovative products are being created driven not by evidence of demand, but by hope and fear. The hope of domination in an emerging niche and the fear of missing out on an opportunity. Another example of orders not supported by explicit demandare production plans based on demand forecasting, which are quite common in almost all industries today. Such make-believe orders are initiated by top managers, not real customers, so there is always a risk of overproduction, unsold stock, and disappointment.

Before demand is identified, fully understood, and validated, there is no reason to confine production to the narrow constraints of lowest cost, highest quality, and fastest delivery. Our recipe is to define clear levelsbacked up by the demand axis: Placeholder, Quick & Dirty, Good Enough and State of the Art, each of which has its own size and shape of the Budget-Time-Quality triangle.

That makes the classical triangle in fact a 3-dimensional pyramid, a tetrahedron if you like, where the fourth corner is the Demand. The 3-dimensional Demand-Budget-Time-Quality pyramid defines the space for balancing the above mentioned plans and schedules and effectiveness of the production process.


The “packaging” part is driven by strategic and marketing decisions. The company decides if it wants to sell the product:

  • a premium product or a mass market product,
  • a finished product or a component supplied to another brand,
  • sell under their own name or join the franchise network.

There are clear advantages for a new company to work as a supplier of a large well-known brand, as they can focus on their production efficiency and use the power of the established marketing funnel and the mature production guidelines. That is especially true when companies expand to the foreign markets. The effort required to introduce a new brand to the crowded market is enormous, and putting a locally familiar label on a well assembled product may be the easiest way to gain market share in a foreign market.

However, that strategy works best if you want to build a profitable and sustainable Good Enough business. If you have an ambition to introduce State of the Art products, that of course will require your own brand on the box, as the product recipe will be very unique and the customer segments will be different from the established Good Enough markets.


The last part of the process is basically delivery of the finished and packaged product to the Service cycle. If the previous process was pretty much converging from multiple suppliers to a streamlined assembly and packaging, the Fulfillment is a diverging Distribution chain, which was covered in one of our previous guides.

The product is being delivered to stores, directly to customers, or to another production company if the company produces components for a larger assembly chain.

Where are you at the large scale supply chain?

A lot of products nowadays have hundreds of components and long multi-step supply and production processes distributed between retail stores, distributors, trade agents, component manufacturers, service providers, franchise networks, and finished goods production houses.

When doing research for a company, the right thing to explore will be to draw the larger scale chain and reconnoiter the surroundings: the suppliers, the sales channels and the subsequent assembly and distribution chains. Where is the most value created? Where is the most severe competition? Which parts of the large chain are consolidated or fragmented markets? Where your company wants to be?

Every segment of a large chain is a market. Your company buys components from upstream, assembles the products of the next stage, and sells these products further downstream. Your competitors, other companies in this stage do exactly the same. How many companies like yours are in the market? Are there a few other companies or thousands of them? Where is your company as compared to the competition? Is it between Top-3, Top-10, Top-500? What effort is required to migrate from Top-500 to Top-100? How will it impact the revenues? Is it a good market to stay in at all?

The competitive markets are almost always described by a power lawdistribution. If you draw a chart with the revenues of the competitors, the revenue of the largest competitor, then the second-largest, then the third-largest and so on, you’ll see what’s called the long tail. It is always the same shape for every market. The difference is however in how fat the tail is, or mathematically speaking in the value of the power law exponent.

Consolidated market. In some markets the tail is very thin. That means the market is dominated by several large players, who produce extremely capital-intensive A class components or capital-intensive multipurpose finished products for a very broad audience. Several big players own more than 60% of the entire market forming an oligopoly and dictating the rules and prices. One largest player may dominate with 15–40%, sometimes even 80% of the market, while the other small companies can not survive the competition. The leaders usually invest huge sums of money in innovation and creating unbeatable State of the Art products.

Examples: chip manufacturers, e-commerce, software operating systems, blockbuster movie studios, social networks, search engines.

Fragmented market. Other markets are much more fragmented with a fat tale of hundreds of companies that manufacture B and C class components, narrow niche tools and services for local markets. In such markets there can also be large players but each of them owns not more than 5% of the market, with the largest player with 7–12% of market share. Such markets are quite frequently connected to certain locations and traditions that make it harder to expand to global scale. Fragmented markets have low entry barriers, so almost any new player can create a small profitable business and even capture a local niche market, but it will be difficult to grow fast even with exceptional products. This is where in most cases being Good Enough is literally enough.

Examples: real estate, oil & gas, brick-and-mortar retail, agriculture, bakeries, spa salons.

On the large scale supply chain diagram you should look at your buyers, your suppliers and your competitors. Are your buyers consolidated or fragmented? Are your A class suppliers consolidated? Is your own market consolidated or fragmented, where are you as compared to your competitors? Are you protecting your market share or having aggressive expansion plans?

Fragmented markets are easier to enter but harder to protect in the long run. Consolidated markets require huge investments and State of the Artproducts but can give the advantage long term domination once the company becomes one of few leaders of the oligopoly.

Problem Analysis

In production there are common problems similar for all chains that we discussed in previous chapters. We also touched them recently in the Demand-Cost-Time-Quality pyramid and the main production management activities — Planning and scheduling and Effectiveness optimization. These common problems are:

  • Costs
  • Throughput
  • Delays and wait time
  • Defects

So let us now explore some of the critical problems that are particular to the production processes. Among them are:

  • Variability of parts and modifications
  • Overload and underload
  • Downtime
  • Continuous improvement of effectiveness

Variability and the curse of dimensionality

Designing and redesigning the production and supply chain is a complex problem. Modern products may easily have hundreds or thousands of parts and ingredients.

Examples: Cars have 20000 or 30000 parts, aircrafts have millions of parts. This explains the possible complexity of the supply chain.

The situation appears even more sophisticated when we realize that these parts are evolving, and the newer modifications of almost every component appear every year. That makes it a real nightmare to track all these changes, to store all possible spare parts for the new and sold products, and identify phantom defects in a complex product that can be due to incompatibility or interference of different modifications of different parts.

Overload and Underload

Planning and scheduling are required to allocate resources and do the right volume of things on time and serve the customer orders. Planning becomes especially critical when the demand and the supply are fluctuating.

Overload: Too many orders means stress and the production works extra hours. The parts, tools, and ingredients run short, which means longer wait time for the customers. As we know, waiting (I Wait) is a frequent reason to lose clients, especially those who try the product for the first time (First Experience) or who are in a rush with the purchase (I Rush).

Underload: Too few orders means no revenues. The assets are expensive, every idle hour means direct loss. Expensive equipment is usually bought on credit and requires regular payments on the loan, whether it is utilized or stands idle. The human resources also require salary payments no matter if they have work orders or not.

Switch: Every order is unlike the other. Even in a very standardized production, there may be several products with different ingredients and required tools. Then a resource — human or machine — switches between tasks, it is always a loss of productivity. Humans need to reset their context, refresh their memory, or physically move between locations. Machines need to be turned off, cleaned, readjusted, and warmed up. Every switch hurts the asset, causing increased wear and rate of defects.


When something goes wrong in production that is bad not just for the current operation. When an incident happens and the production process is stopped to fix the problem, that means all other works depending on the critical output of the current operation are also suspended. Thus a small problem and downtime on the critical path causes the domino effect of delays and downtimes.

Example: In construction, just one typo in the project documentation may cause millions of expenses even if this misprint was identified early, and nothing was wrongly constructed yet. How is it possible? When the error is found, it may require an additional round of documentation approvals, some signers are out of the office, others are waiting for someone else to sign, so the round may easily take 3 weeks. This three-week delay is a downtime for the expensive rented equipment, construction workers and subcontractors who stay idle waiting for the greenlit, for a large project this downtime may cost millions. And if the misprint was identified during the construction, the impact may be tens of millions because some parts of the building being constructed must be demolished, the errors fixed, the materials resupplied, the developments reworked, and the idle time of all other construction equipment and workers waiting to the work on critical path to be finished.

Continuous improvement of effectiveness

When you care about profits, you need not just fulfill the customer orders, but make everything faster, cheaper, while sustaining the quality of the products. A wise production manager establishes the cycles of continuous improvement of their processes, identifies the leaks and gradually upgradesthe process. The most painful gaps and leaks usually appear between functions, so they should be considered first. Another source of low performance are outdated assets, some of them are really hard and expensive to replace, so additional corrective measures may be re required.

Leveling guide — Production and Supply Chain

Let us take a look at 4 levels of production maturity.

At the Placeholder lever, production is not serving actual customers, but may create prototypes or components for the Product design cycles. For this purpose it makes sense to have very multipurpose tools and people. It’s a research phase with many uncertainties and the value is to test as many hypotheses as possible as fast as possible. Hackers with swiss army knife-like tools may be very productive in generating and testing ideas. More traditional production people who are accustomed to order and neat work may not be the best fit here. They will be too accurate, too diligent, and consequently too slow.

At the Quick & Dirty level, there is a real demand already and real customers who await deliveries. However the demand is unstable and not yet enough understood. The Customer acquisition funnel will not be able to bring long-term contracts and will jump from client to client trying to get at least something to stabilize utilization of the expensive assets. The production will suffer from periods of underload. Having expensive idle multipurpose tools and employees will always push the owners towards finding ad hoc orders for them and plug the holes in the budget. That defocuses the business, but gives extra time for survival. Many businesses can never escape this trap.

The Good Enough level of production can not be implemented without the Good Enough understanding of the customer needs (genotype) without the Good Enough Customer Acquisition Funnel, as it requires a more stable and predictable flow of incoming orders. The production is able to design an assembly line with dedicated resources, people, and tools. Quality becomes an important must-have factor. Hackers and multi-purpose tools are being replaced with highly specialized and predictable resources able to deliver with minimum possible variance. The company makes money the most efficient way.

The State-of-the-Art level creativity is required again. The company can not rely on copying the processes from competitors and engineering companies, but needs to invent its own new way of production. When built on top of the Good Enough process and culture, the company is able to run a lot of pilots with new pieces of technology, reorganize the old production chains despite the natural resistance and significantly improve the value and cost structures of their products. Quality is still important but leadership becomes even more vital. The production shall work in a very tight conjunction with the sales processes as the new revolutionary products will most likely target the unconventional customer segments and distribution channels.

As we see, there are gaps between all the described levels, they require different approaches and different execution cultures. Quite often the reason for change is outside of production: in sales. If the Customer acquisition funnel is capable of generating a sustainable flow of clients, there is a good chance to build a Good Enough production process. If marketing visionaries are able to identify a new trend and growing customer segment, there is a chance to dominate this new market with a revolutionary State-of-the-Art product.