Next Generation Conglomerates
A Playbook for Digital Disruption
Note: Read more on Twitter here
Those who know me best know that I’m a historian of business failure and success (having experienced plenty of failure myself). As a student of success stories, I’ve been fascinated with conglomerates for years. My obsession was initially piqued with my growing love affair with Berkshire Hathaway and further amplified by the works of Greenwald and McTaggart (both of which discuss core benefits of conglomerates and case studies of the model in detail).
As we look back to the origins of the conglomerate structure, the model initially became popular in the late ’50s / early ’60s as low interest rates made it more accessible for creditworthy acquirers to lap up their adjacent competitors. Ultimately, this spilled over to acquiring complementary or even countercyclical businesses. The host company could take on debt to buy these assets, leveraging the cashflows of their core business to service debt until the acquisition became accretive, aggregating the cashflows of the individual holdings into a holistic financial picture. Often these targets were too small and/or low-growth to be exciting (sometimes even public companies in their own right), but when aggregated together, the sum of the parts was more enticing than the standalone components. Often this financial profile could be further bolstered via shared infrastructure (e.g., shared procurement) that would give the holdings an additional edge on their competition and further enhance their profitability.
Berkshire Hathaway is the most well-known example of this, having demonstrated persistently high returns over a multi-decade period. But Berkshire certainly isn’t alone. Conglomerates like Danaher ($200B market cap) have created incredible shareholder value, as have aggregators of vertical software like Constellation Software ($45B market cap) and Roper Technologies ($45B market cap). In insurance, we’ve seen this with companies like Acrisure (an aggregator of independent insurance agencies that rose to $23B of value in <10 years) and Ryan Specialty (a conglomerate of specialty insurance products that is worth $11B just 11 years after founding), as well as in food and beverage with companies like Nestle and AB Inbev (who have a combined market cap of >$450B). For all these companies, their value is the accumulation of the future cashflows across a diverse array of products and/or holdings, not just a single business.
The Death of the Conglomerate?
While conglomerates demonstrated incredible success, the model lost some of its luster over the last two decades as corporate strategy shifted to favor focus and as the shared services of these conglomerates became commoditized. This itself increased competition for potential acquisitions, making it difficult for conglomerates to find opportunities capable of yield, especially as capital markets expanded and made credit significantly more accessible. Only the special few (like Berkshire) were able to drive alpha from this strategy, executing on opportunistically mispriced assets and/or companies that were willing to take a discount to be in the orbit of the “Oracle of Omaha.” Many conglomerates also started to become inefficient, given the high managerial requirements of tracking multiple, diverse and distinct businesses (especially when doing so across multiple geographies with different corporate and tax structures), further eroding the economic synergies of the model. As middle management bloated in these structures, the cost synergies of the model broke, making M&A value-destructive more often than not. Berkshire defied this trend, managing an enterprise worth over $680B that did $355B of revenue in 2021 with a miraculously small HQ team of 20–30 people.
As we look to the tech universe, however, a resurgence of conglomerates is occurring. We’re seeing platform companies like Amazon and Apple emerge as conglomerates, not by acquiring assets, but by expanding their empire into new business lines. These companies are using the cash flows of their dominant business lines to extend into new product lines that make them even more dominant, rather than acquiring wholly independent assets like the conglomerates of the past. Amazon expanded upon a once-tiny niche (e-commerce for books), to now find itself in a host of businesses that are completely separate from their e-commerce core (web services, third-party logistics, media & entertainment, etc.). While the old-school conglomerate structure focused on acquiring and integrating existing assets under a single umbrella, the next generation of this structure will focus on creating platforms for development of these assets based on a shared core “capability”.
Conglomerate 2.0 — The Platform Conglomerate
As we look at the root causes of deterioration of the conglomerate model, today’s startup landscape provides the opportunity to recapture its glory. At the heart of this is identifying a sustainable mechanism for continuous competitive advantage (aka a “capability”) — how can the startup generate edge on the competition for EVERY new product it launches? Apple is a perfect example of this. Apple isn’t a consumer products company, it’s a “platform” business — the ecosystem of developers that have built applications on top of their platform is its core engine for competitive advantage. The depth and breadth of Apple’s developer ecosystem drives captivity over its customers, further reinforcing the attractiveness of its platform to developers (even amidst folks like Elon Musk raging about the fees). With that, Apple is able to extend a nearly limitless array of products and services with extremely limited expenses beyond maintaining the core platform.
The results are simply phenomenal with the company’s App Store generating >$133B in revenue in 2021 (more than 3x what it was just 5 years earlier) and has corresponded with the company’s growing Return on Equity and Net Margins over that time span (see below). This has been done almost entirely organically, with the company’s last major acquisition happening in 2014 with Beats Audio.
The company is also able to extend its product lines into higher margin areas (Airpods have gross margins of ~60% compared to 40% for the iPhone), further demonstrating the captivity its platform has over both consumers and developers. Unlike traditional companies, Apple can 1) deploy that core benefit to future project development(s) at extremely low cost, 2) leverage the success of that product to reinforce its core competitive engine and 3) quickly launch products that achieve profitability early in their lifecycle (like AirPods). While “Platform” companies like Apple and Salesforce have generally leveraged their lured third-party providers to create offerings that extend beyond its own service, Apple’s success in developing its own product line extensions gives way to a new model — the “Platform Conglomerate”. Unlike “Platforms,” “Platform Conglomerates” are more likely to internalize their platform capabilities to pursue de novo development of adjacent business lines at rates unattainable for those off the platform. These companies develop a core capability for fueling new product development that can be deployed not just for a single product, but that is infinitely scalable to other future products. These companies leverage learnings around each successive product launch to hone their engine, creating a flywheel that generates increasing returns to scale. By pursuing high-margin categories (e.g., software) that require limited upfront investment (particularly in niche markets with more limited competition, where profitability can be achieved earlier in the lifecycle of the launch), the company can harvest early cashflows to invest in future product launches. Doing so creates a layer cake of cashflows across these businesses, while reinforcing the core engine — the ultimate value driver of the business.
This approach, however, is unique to companies of the Digital Age. Imagine an industrial steel conglomerate attempting this approach. It would have to pour out a tremendous amount of capital into an asset that might not be cashflow positive for years to come (or ever, for that matter). While the conglomerate may have an advantage in procurement for its core business or market, those advantages would likely be far less significant in others (as GE witnessed as its business experienced scope creep). Even if those advantages were attainable in the market, the legacy shared services would require additional investment and lead time to meet these needs and produce any real benefit (while economic synergies may exist, these types of services incur variable costs that generally scale relatively linearly with revenue). The company would also require additional management oversight and expertise as it extended into new markets. All of these investments would extend the j-curve of profitability beyond a realizable window, making synergies impossible.
Digital Disruption of the Real Economy
While developing a business with the dominance of Apple and Amazon is easier said than done, we’re seeing examples of this model in our portfolio, with companies like Starday. We discussed Starday briefly in our post, “Niche is the Next BIG Thing”, but as a reminder, the company leverages a core data engine to identify niche food and beverage categories with unmet consumer intent and simulates countless iterations of products to digital consumers to optimize for launch success. With that, Starday has launched four brands in its first 18 months of operation, all of which are seeing massive pull from major retail outlets. Each of the brands operates in a niche market where they can be the market share leader within the first few years of launching, and each is capable of generating long-term contribution margins greater than 50% (far higher than in traditional food categories given the limited competition and higher price elasticity of specialty products). It traditionally takes “Big Food” roughly $20M and several years to launch a new brand with a failure rate of ~70%. While Starday’s results are early, we’ve cut that cost to nearly 1/100th and slashed the timeline to months from years, and we are seeing very promising results on the success rate of brand launches.
The brands, however, are not the core product. Rather, Starday’s core product is the engine that enables them to successfully launch brands at a fraction of the time / cost of legacy food companies and with much higher success rates. This core asset enables Starday to 1) deploy the insights of its simulations to additional projects at extremely limited cost, 2) leverage the learnings of each successive simulation / product to further strengthen the engine and 3) achieve profitability on core SKUs incredibly early in the lifecycle of their brands given the high margin profile and minimal upfront expense. As an added benefit, outsourced services and modern data infrastructure enable Starday to operate with a fraction of the assets and headcount that it would normally take to manage the brands it owns, demonstrating potential for them to operate far closer to how Berkshire does than heavily verticalized, asset-intensive players like Nestle.
We’re seeing other examples of this in our portfolio as well and believe this new conglomerate model demonstrates the potential for immense digital disruption. The core engines behind companies like Apple and (hopefully) Starday drive a meaningful edge on Return on Equity, enabling them to disassociate themselves from commoditized products and generate real cash flow, not just revenue. These engines are defined by network effects (whether they be customer, user, data or platform-oriented) rather than the industrial conglomerates of yesterday that largely dominated on economies of scale. The truth is that network effects are far more powerful than economies of scale if given the chance to fully realize their potential, so the advancement of this model should terrify corporates feeling comfortable in their current positioning. We believe this model has applicability across each of our core sectors, but it requires a nuanced understanding of these legacy markets to determine how to yield a truly cohesive, competitive advantage (as opposed to the accumulation of adjacent businesses that are likely to yield results more akin to those industrial conglomerates have seen in recent decades).
While there is plenty to be uneasy about headed into 2023, lessons of the past often have a way of telling us about the future. Seeing the prior success of the conglomerate model and its reemergence in the digital economy, we see tremendous promise for this model in the years to come. We look forward to seeing the disruptive impact the next generation of conglomerates can have as they expand beyond digital into the real economy.