Data Eats the World
The wheel. Electricity. The automobile. These are technologies that had a disproportionate impact on the merits of their first practical use-case; but beyond that, because they enabled so much in terms of subsequent innovation, economic historians call them “general-purpose technologies” or GPTs. The automobile did not just replace the horse-and-buggy, it enabled the suburb and the shopping mall. Some say Machine Learning is a new GPT: we posit that hiding in plain sight, data is itself the latest general-purpose technology.
The idea that software is eating the world has gained currency in the last ten years. In our opinion, this is indicative of the last regime — not the next. See, software is focused on processes. Software automates and simplifies. The low-hanging fruit of automation has been plucked. That does not mean there is no longer any runway in tech to innovate and create value. On the contrary — there’s a lot of runway left, and the key to unlocking innovation is data.
We see a common mistake with entrepreneurs that are software-centric: they tend to latch on to specific industries because they mistakenly believe in a waterfall approach to innovation. These founders see the world a lot like linear, waterfall software development: first software eats this industry, then it eats the next. The same mistaken belief is why many VCs adhere to “themed” funds like fintech or D2C or Web3 or whatever is faddish. Unlike VC investors, we don’t adhere to a “theme.” We find these themes give the illusion of specificity without any common first principles.
If themed funds are like waterfall software development, then our data-driven approach is like agile. We aren’t about lining up industries like dominos in order to knock them over one by one, we are industry-agnostic. We found companies across industries and contexts. Cloud computing, healthcare, HR tech, DevOps, it doesn’t matter to us.
What we do is even simpler and broader than that: we found and build data-driven companies. It isn’t a universal theory, but it sure is close to primeval and applicable across industries and contexts.
“Data-driven” is such a common compound adjective sometimes we fear repeating it so often doesn’t do our intended meaning any justice. But we haven’t found a better way to explain our thinking, so it stays. What do we mean by data-driven? Data has had its fads — “Big Data” was a buzzy concept in the 2010s that has fallen out of favor compared to “AI and ML.” Our founders, Tom and Vivek, started their first company — Rapt - in 1998. If founded today, Rapt would be deemed AI/ML (and probably be in the portfolio of a VC with an AI thesis).
Rapt was all about the algorithmic and model-based processing of data — not just moving data but taking data that already existed in customer systems and creating value — first for supply chains and then the nascent digital ad economy. From first principles, AI and ML companies are data-driven. Data-driven companies derive the core value to the customer from generating, capturing, orchestrating, pipelining, analyzing, or activating data.
Tom and Vivek’s second company before super{set}, Krux, was data-driven because it helped companies collect data and use it for improving their businesses. Seriously, we can say that so simply because data-driven is such a fundamental concept. All super{set} companies are data-driven, and we adhere to this not just because that is our area of expertise as technologists, but for two important reasons:
- Data is an Asset: There is physical capital and human capital — and data capital. The difference is that data never wears out or gets tired. We believe that one day data is going to sit on company balance sheets — an intangible asset, but so important that it has to be there.
- Data is a Flywheel: All platforms feature network effects where more users and friends create a bigger pool of value — data network effects work similarly. The more data generated, the more it can be put to work in better services, automation, and more. This works at the individual level, but there are also second-order data network effects where data is harnessed from the collective to generate ever-increasing value for individual customers. This results in an exponential increase in value as more data = more customers = more value = more customers = more data.
Of course, it isn’t as simple as “get data, have a perpetual motion machine.” For one, using data can be complicated. Personal data must be respected as a human right and protected under privacy laws. Or data may be wildly unstructured and hard to derive value from, like tar sands rather than a gushing well. There are a host of non-technical barriers to scaling data across contracts, customers, and go-to-market. The concept of give-to-get is not a technical problem, but it is an essential condition for turning the data-driven flywheel and must be cemented at design time.
At super{set}, we see data-driven origination opportunities everywhere. The hard work is thoroughly exploring those opportunities to only follow-through with the gems.
Want to work on building a data-driven company? Find out what it takes to become a super{set} co-founder.
