The world is marching inexorably towards specialization, where companies build highly specific expertise in one subcomponent and partner with an ecosystem of suppliers for everything else. This journey really got going with globalization in the 80s and 90s. Digitization accelerated that trend by creating new methods for independent organizations to work together in embedded and trusted ways.
Supply chains that were once linear now have the opportunity to become networks. Software is necessary to help orchestrate and reduce the transactional cost of these networks of companies. Software providers have an opportunity to build important platforms that can act as a coordination layer for entire industries. This is a market opportunity that we feel is wide open—there is a lot of money to be made here.
We know, we know—that was a lot of jargon per sentence. But stick with us. We’ll simplify it all and map it back to something a lot more approachable at the end: beer!
Why: Specialization Theory
The theorem predicting the outcomes of today’s specialization race was proposed roughly 85 years ago by Nobel Prize winner Ronald Coase. In Geoffrey Moore’s retrospective on Coase’s “The Nature of the Firm,” he wrote about Coase's theorem:
“…doing any kind of non-core work outside the firm has the advantages of leveraging someone else’s capital investment [scale] and expertise, but procuring the right product or service from the right vendor—and managing the relationship with that vendor and the workflow connecting the two companies—imposes a transaction cost.”
The supply chain has become so complex and so global that transaction costs are much higher than they were in the 90s. Firms are actively looking for ways to reduce costs and make this whole process easier.
A simple example of this specialization is the progression of Ford. When the Model-T was first created, the whole process was vertically integrated. They even owned an iron ore mine! The Ford of today has thousands of suppliers and owns relatively few assets. In technology, the same has occurred with the personal computer. Transitioning from a wholly integrated company to a diversified subcomponent industry has made the market victors far larger—Microsoft and Cisco are much, much bigger than the fully integrated IBM of yesteryear. This pattern of specialization has occurred in essentially every product category, with even something as simple as a bike having 50+ suppliers.
What: The Permeable Enterprise
In the future, any non-strategic asset will be removed from a company's balance sheet and placed on the income statement as an expense. There are the obvious things—corporate kitchens, real estate, cleaning services, and call centers—but there will also be non-intuitive capabilities outsourced; it won’t just be lower-skill work. We propose that outsourced assets will have four types of potential arbitrage. Software companies should look to coordinate any or all of these four levers:
- Access to labor: The temptation in our post-Covid world is to rely on remote work for geographically concentrated labor pools. Whether you’re looking at Vietnam for textile labor, Utah for call centers, or New York for vest-wearing bankers, it can be alluring to just say, “We’ll hire remote.” However, when labor requires seasonal spikes or has high variability, it requires significant company resources to manage and may be better off handled by a specialized provider. The more transient the labor is, and the more ephemeral the activity, the more it should be transitioned to an expense.
- Risk pooling: The more consumers that can be added to a risk pool, the more performative its actuarial prediction will be. If a firm only sees the data it is involved in, it has a limited algorithm—but if you can utilize data across firms, you get better risk or credit data. By pooling risk, you can also have an underwriting specialization, which naturally leads to a cheaper cost of capital. The easiest example of this to grasp is in insurance products, where a “brand company” can focus on distribution, and a specialized provider can pool risk from many customers. By adding some ML/AI spice to the mix, scaled data pools can result in significantly more performative risk profiling.
- Data and benchmarking: A third party can not only reduce risk but also provide explicit benchmarking and scoring to calibrate risk across a broader set of data; this helps to make better credit decisions, evaluate operating performance, etc. These companies can emerge as data intermediaries. Organizations like FICO, Avetta, EWS, and Verisk have all opted for this path.
- Capital formation: If an activity requires a lot of capital, it needs a lot of scale. This scale is achieved by a supplier of many customers (vs. within a customer itself). In the 90s, this took the form of electronic manufacturing services companies like Wisetrom, Flextronics, and Solectron taking over manufacturing for companies like HP. Arguably, AWS is a version of this service too! Outsourcing servers is a great business, and the activities coordinating all this labor would be best coordinated by software.
So how is a company to take advantage of these coordination opportunities?
How: The Opportunity for Software
There is a delicate balance to strike between specialization benefits and transaction costs when outsourcing assets. As the boundaries of a firm’s focus narrow, the borders also need to become more permeable, to allow multiple parties to work together to build a product.
The answer to this problem? Software! Digitization decreases transaction costs. As the move to company specialization and task-specific scale continues, it naturally propagates the need for more software. It is a virtuous circle of software proliferation.
The benefits to a software provider’s industry are also beyond returns of scale and specialization. The more automated this method of interaction, the more the firm can scale the relationship.
Companies use a variety of applications to automate interactions with suppliers:
- Sourcing and Contracting: Software can dramatically decrease searching and qualifying costs between parties by acting as a central repository of data. The most advanced companies will move beyond the database to an active marketplace that can facilitate negotiations or contracts.
- Procure to Pay: Software can also reduce the cost of ordering, receiving, and paying for goods and services. This is especially important in high-transaction-volume businesses.
- Certification: Tracking vendor qualifications is a pain (particularly in multinational supply chains). Software that acts as the trust layer to ensure that vendors are following technical certifications or ESG standards is hugely valuable.
- Managing Risk: When certifications move from a “nice-to-have” to a legal necessity, software can start to manage risk. An example in the financial sector is companies automating adherence to Know Your Customer (KYC) regulations. It can also go beyond that simple use case—whether you’re dealing with insurance, cyber security, or credit scoring, software can dramatically increase the speed and decrease the cost.
- Working with Suppliers: Perhaps the highest value, and the most challenging layer to digitize, is the collaboration and coordination of work. Allowing third-party workers to interface with other stakeholders in an automated way is a massive value unlock. Straightforward examples like call center software or physical access are a great place to
start, but the bigger value is in acting as the communication, management, or project collaboration layer.
These tools will be combined depending on a company’s relationships with its suppliers. Simplistically, you can segment these relationships along two dimensions, creating a basic 2x2 matrix. On the x-axis is spend—think of this as a proxy for the money involved or just the sheer transactional volume of a relationship. On the y-axis is the strategic importance, or value of the supplier to the company’s end product or service. In the illustration below, we have used a pharmaceutical company as an example (though the idea applies to any business).
Basics: These are suppliers on whom the example pharma company spends very little—and that spend is not critical to the company relative to other supplier categories. An example would be the outsourced provider of cafeteria services. These suppliers are granted physical access and can request the data directly related to the task they have been given (e.g., the number of employees expected to be onsite on any given day), but the nature of the relationship does not warrant intensive data interchange.
Specialists: These are the suppliers with whom your spend is limited, but they are critical to your product. Ideally, you don't want many of these suppliers because they will likely have leverage over you, but they are unavoidable in many businesses. In our pharma example, a principal investigator in a clinical trial may not be a large part of a clinical trial budget, but will be critical to completing a trial on time.
You need these suppliers more than they may need you, so the software used to manage this relationship needs to have some real value for the supplier—or you need to pamper these suppliers with more help and service, or offer better payment terms than with a basic supplier. Generic sourcing and onboarding techniques may not suffice. With specialists, you have to make the onboarding process as smooth as possible and tailor it to the service being bought.
Partners: These are the suppliers that require higher coordination because the supplier is becoming increasingly involved in the product manufacturing or service delivery process. In our pharma case study, a contract manufacturer would be an example of this supplier type. Partnership can also occur in the relationship with the customer, when the supplier is co-responsible for NPS scores or other performance metrics. This is common in customer support, where third-party organizations in lower-cost regions will answer support questions on your behalf using your call center software and knowledge management system.
The most extreme version of this partnership approach is when you design the product, and third parties build and deliver it. This is common in the semiconductor industry, where most companies lack manufacturing facilities (a.k.a. “fabless” companies) and focus entirely on the design and distribution of chips. We published an essay on this use case if you are interested in diving deeper.
Deep partnerships are not limited to physical goods, either! It has also happened in financial products. Take auto insurance: one company may own the brand and underwriting model, but they will then contract with numerous third parties in the case of an accident. These third-party providers can include the grimy work of the tow company, the mechanics at the auto body repair shop, and the office workers doing appraisals.
Co-Development: These are the suppliers that require the highest level of coordination. At the low end of co-development is design for manufacturing. A great example is how Foxconn will sometimes assist Apple with manufacturing design tweaks for the iPhone. In the context of our pharma example, BioNTech did most of the vaccine development for the Covid vaccine, while Pfizer handled the testing, regulatory requirements, and scaling.
These partnership and co-development relationships typically necessitate workflow solutions, collaborative planning, and even collaborative design solutions. In the software industry, at the most intense level, your partners are helping you build the core product itself. A co-developer
needs complete access to your API or SDK, and adheres to the same design and documentation standards as your employees.
The Supplier Network Opportunity
Most of the problems that are caused by friction can be solved by software. On the buyer side, processes high in cost of friction, time, and spend can have that cost greatly reduced through automation. For the seller, they have an opportunity to do the boring work just once—they can publish real-time credentials, qualifications, financial viability, certifications, and cyber security that all potential buyers can trust. This saves massive amounts of time with things like HIPAA compliance. At its core, it’s about the reputation and professional identity of an organization: being accessible and trusted via software.
If this professional identity for the supplier is both persistent and portable, a buyer can instantly qualify a supplier and onboard them more quickly. This dynamic creates a two-sided network effect where buyers bring suppliers onto the software platform, which brings more buyers to the platform (additional suppliers benefit buyers, and vice versa). Supplier networks are less well-understood than their consumer network counterparts but may offer even stronger opportunities for lock-in and profit. Both stakeholder participation and data assets enforce a two-sided network effect—and in areas like safety, compliance, and risk, even competitive buyers and sellers want a common standard. The software company that can be the host of that standard will significantly reduce transaction costs in the ecosystem.
Web 3.0’s Similarities and Differences
The world is collectively dunking on crypto right now, but Web 3.0 has some powerful concepts that are analogous to the supplier network opportunity of today (which echo a lot of concepts from object-oriented computing thirty years earlier). The idea of creating trust-verifiable and active contracts that allow multiple entities to work together in new embedded and collaborative ways—these are worthwhile ideas. We would argue that in some cases, you can entirely forget the speculative tokens. Supplier networks can utilize cultural cache and ethical behavior to organically build trust. Rather than rely on capital incentives, a company can rely on the revenue streams of a software product with product-market fit. From solving a single problem, they can go on to building a network.
The modern corporation is starting to look like a brand sitting on top of a web of suppliers. The deeper the economy goes into digitization, the more porous and transient a business's capabilities need to be. Even Sam Adams—the official beer of the Boston Red Sox—isn’t brewed in Boston! It’s brewed in a handful of other states, and from there, is bottled by third parties and then distributed to retailers.
This transformation of industry and the building of a “permeable enterprise” will be powered by software. The software providers are naturally positioned to be a multi-stakeholder platform and accrue significant profit. If you want to chat further about these ideas, feel free to contact us at firstname.lastname@example.org.