Latent Data Assets
In January 2024, Reddit, an internet chat forum, disclosed that it had entered into data sales agreements with an aggregate value of over $200m. The data in question being forum conversations that Reddit had accumulated on its platform over two decades since its inception. Until that point, Reddit made most of its money solely from advertising and forum membership fees. In one fell swoop, the business revealed a latent asset, one that is now highly sought-after as a training set for large language AI models. In so doing, Reddit also up ended its business model.
In previous thought pieces, we have tried to highlight various aspects of how great businesses are run. We have discussed customer centricity, corporate culture, playing the long game, the benefits of scale and we examined execution. An area that we have not yet broached explicitly is business models.
A business model might be described as a catch-all phrase that basically describes the framework of how a specific company chooses to add value to customers and how it makes money in the process. Razor and Blades, a stalwart of MBA case-studies for decades, has for some time now been over-shadowed by the trendier direct-to-consumer (d2c) and more recently x-as-a-service (subscription) models as the pinnacle of business model evolution. As the Reddit development suggests, business models remain highly dynamic.
Business Innovation
The Oxford dictionary describes innovation as "a new method". Whilst most discussions of innovation centre on technology, i.e. new technological methods – what gets less recognition, is business innovation – that is, new business methods of extracting value. It seems to us that both go hand in hand. In parallel with the extraordinary technological innovation of the 20th century was a series of business practice innovations that mirrored it.
Early 20th century business formation favoured conglomerates. The conglomerate was perhaps an obvious approach to gaining and deploying massive scale in product development, customer reach and frankly, market dominance. The market success of GE, GM, DuPont et al was then held up as proof of the model's power. It didn't last; the burgeoning tech industry of the 1960s and 1970s was a fertile ground for not just disruptive new technology, but also disruptive new business models. The evolution of business models since say the 1970s has been relentless and continues in the 2020s.
The semiconductor industry is as good a proxy as any to review this evolution. Weighed down by massive capital requirements, corporate sprawl and the need for specialisation, once dominant vertically integrated IDMs (integrated device manufacturers) like IBM and Philips became a dying breed. New business models soon emerged to replace these tech conglomerates: the principle one being Foundries (outsourced manufacturers such as TSMC), coupled with the newly termed 'Fabless' model (asset-light design-only companies like Nvidia) facilitated far superior return on capital and created dramatically increased dynamism in the sector. It didn't stop there. EDA (electronic design automation), dedicated Capital Equipment vendors (ASML) and not least IP (intellectual property like ARM) also showed how highly specialised business models with distinct revenue models could be
adopted and restore high ROCEs in the industry. Business model innovation, it seems, was just as important as the technological innovation.
The chip companies' customers were themselves also evolving their businesses rapidly, again mirroring the technology sector's rapid evolution. Apple, originally a business selling hardware systems to consumers, evolved into a business that still sells hardware to consumers, but also subscription services and perhaps most notably, has itself revealed a massive latent asset. That asset was the stealth toll bridge it has built, which extracts over $20bn from Google each year for the privilege of being the default browser on apple's hardware devices.
Silicon Valley has led business innovation over the last two decades. Companies like Google and Netflix have led the charge in not only technological innovation but also adopting smart scalable business practices. The Valley has followed suit en masse across a raft of markets. At one end of the spectrum, Amazon developed AWS and, via a creative business model pivot turned what was a cost centre into a $70bn+ revenue division. At the other end, the CTO of Vizio, a TV company has candidly admitted that its business is about "the post-purchase monetisation of the TV" – in other words, selling customers' TV usage patterns. Customer monetisation is not new of course – Airlines have long relied on a similar but perhaps more acceptable approach to monetising their high- spending customers through rewards programs and credit card partnerships.
Notwithstanding those examples, DELL, one of the survivors of the tech industry is a superb example of constraint-driven business creativity.
A Pioneer
To say that Michael Dell was ahead of his time is an understatement. Mr Dell pioneered the direct-to-consumer business model decades before so-called 'd2c' became a thing in business schools. Dell also understood the power and value of his business' data as a predictive business tool.
Founded in 1987, DELL the business is a fascinating and still evolving business case study. It is fascinating because, like Costco, Southwest and Ryanair, it implemented an extremely smart business model to realise a very high return business inside the constraints of what is a cut-throat and typically low return end-market. Ostensibly, Dell setup an ultra-low-cost business that sold direct to customers. That might sound passé here in 2024, but in the 1990s that alone was massive undertaking in what was then an industry and supply chain very entrenched in its ways. Early-on, Dell (and his unsung-hero CFO Thomas Meredith) discovered and leveraged the subtle but highly advantageous side-effects of their chosen business model. This HBR article extract gives a flavour for the business innovations that Dell and his team pioneered:
"the direct model turned out to have other benefits that even Michael Dell couldn't have anticipated when he founded his company. You actually get to have a relationship with the customer he explains. And that created valuable information which in turns allows us to leverage our relationship with both suppliers and customers. Couple that information with technology and you have the information to revolutionise the fundamental business models of major great companies.....Most important, the direct
“model has allowed us to leverage our relationships with both suppliers and customers to such an extent that I believe it's fair to think of our companies as being virtually integrated. That allows us to focus on where we add value and to build a much larger firm much more quickly.”
Source: "The Power of Virtual Integration: An Interview with Dell Computer's Michael Dell", HBR, 1998
In short, Michael Dell had realised the predictive power of his business' data. DELL, the business, in effect created a virtual vertical-integration business model by "blurring the traditional boundaries and roles in the value chain". DELL famously pioneered just-in-time delivery of stock. Less famously, it was DELL's CFO Thomas Meredith that had realised that the relatively low margin nature of their business, coupled with a ridiculously fast deflating input-stock necessitated a culture obsessed with low stock and cash-collection. The outcome was, in effect, a negative working capital business model and a massive boost to ROCE. CFO Thomas Meredith deserves huge credit for this as he recognised the impact that the business model would have on ROCE.
As Bill Gurley presciently observed in his Fortune article (extracted below), Dell had stumbled upon an interesting new stage in the evolution of business models: negative working capital. They collected money from customers before they have to acquire components or spend money. This phenomenon allowed the company to grow without raising capital, even if day-to-day profitability is zero.
“Dell's incredible five days of inventory allows it to pass on component price declines faster than anyone else in the industry. But perhaps the unique aspect of Dell's business advantage is its negative cash conversion cycle. Because it keeps only five days of inventories, manages receivables to 30 days, and pushes payables out to 59 days, the Dell model will generate cash”even if the company were to report no profit whatsoever."
Source: "A DELL for Every Industry", Fortune, 1998
That very insightful passage was written well over twenty years ago and yet it is fascinating that those hugely attractive attributes of that business model remain quite a rare find in today's businesses.
The Next Wave
The relentless pace of innovation in AI today suggests that business models will continue to evolve rapidly. The reddit example above offers just one example for how latent assets might be discovered and monetised within other companies today. DELL shows that such assets can be monetised internally via higher ROCE.
It is worth pondering what similar assets in other companies might prove to be: similar user-generated content, retail transaction data, or it might even be proprietary know-how such as a codified version of the company's procedures.
There are likely many more Reddits out there.