Some of the largest companies in interesting, big industries (insurance, banking, real estate, distribution, manufacturing) have functioned in a homologous manner for years. I continue to be interested in these markets because the innovation to date in the categories has been largely about external experience over internal structure.
Thoughts on Scale
It’s important to layer in interesting and unique customer / stakeholder experiences, but I don’t believe that’s what caters to scalability, which is what venture-backed startups are about. They need to reach scale quickly in order to make sense as investments given the cost of capital for VCs.
“A startup is a company designed to grow fast.” — Paul Graham, Startup = Growth
However, the problem is that the inherent structure of some of these businesses don’t really cater to rapid scale like a software business with minimal marginal cost.
- A bank’s P&L looks different than an operating company’s. How do neobanks scale? When do equity investments make sense given the structure of balance sheets?
- Insurance companies need so much money to work effectively. How much would you need to invest in customer acquisition and capital reserves?
- For manufacturing-centric companies, how do you compete with the economies of scale of the industrial giants? What does the Honeywell ($HON) or Caterpillar ($CAT) of tomorrow look like?
So, true innovation will require some second-level thinking. In some cases, it could require financial engineering. I’m increasingly interested in the product design of securitization. It could also entail betting big — here’s an example outside the venture ecosystem.
Architecting the Organization
Until recently, I didn’t know theorists and scholars had been expounding upon the theory of the firm within the context of networks and markets for decades.
Peter Drucker is considered a “founder of modern management,” and he often wrote about decentralization and outsourcing. Charles Handy introduced the ideas of the “portfolio worker” and the “Shamrock Organization.” There’s a seminal paper written in 1987 called “Electronic Markets and Electronic Hierarchies” (email me for a PDF) that discusses the changing nature of coordination costs and what impact that would have on business structures. Kevin Kelly wrote extensively about networked organizations in the 1994 book, Out of Control.
There are so many different spins to these lines of thinking. Maybe you transfer the components within a linear corporation into nodes in a network — don’t have a single company distribute, design, and produce a product. I’ve written previously about 3 sided platforms — I think lighting products is a cool application of this idea. Internet-native bodies could aggregate and build upon collective power of independent agents. For example, I think a better group purchasing organization for food and restaurant startups should exist.
If companies and processes across industries (especially old-school ones) become networked and programmatic, how does one design trust into these new systems? People speak of “APIs eating the world,” where human functions and processes become programmatic and oriented towards developers. Matt Hackett had a pretty cool idea in this newsletter.
Why is there no Amazon Web Services for physical space? An ever-expanding bundle of services that let developers provision and deploy physical space. Spin up an r3.xlarge for four hours to host your pop-up, deploy a cluster of showrooms in us-west1-seattle and us-east2-washingtondc to sell your new merch line.
Furthermore, take something like corporate restructuring. You could run simulations, machine learning, and APIs to settle on levels of debt and equity to inject into businesses, but corporate leadership would never trust advisors outside the current investment banking establishment.
Where does trust emerge from? How do you onboard institutions and actors into new structures of work?
Sometimes, startups doing business in legacy markets don’t necessarily need to radically restructure their way of doing business — they just have to accept initial metrics that may fall out of the industry norm. New insurers like Lemonade maintain early heavy loss ratios with the hope that they can reap the returns of the data assets that come with initial customer cohorts. Entrants into pharmaceuticals like Thirty Madison and Ro are building a portfolio of brands/drugs like big pharma companies without some of the advantages of scale. They’re betting they can build moats in other areas concerning technology and data in the long run.
In the next year, I hope to continue developing my thoughts business models, structure, and legacy markets. I believe there’s still plenty of opportunity if the right approaches are taken.
If you enjoyed this article, I would love to hear from you!
I can be reached at aashaysanghvi[at]gmail.com.