A Distant Look at the Whole

From gestures to a sketch to the screen in two easy steps.

Mar 02, 2024

One of the purposes of this newsletter is to work in public on a book project about the market for data. There’s a discipline in writing a newsletter (he says, already two days late this week), and trying to explain the project leads to diagrams that I need to create. If they don’t work, better to discover that now than after committing them to print.

Markets or a market?

When I introduce the project to someone new, they’re usually familiar with a particular market for data. They know a set of customers or a category of data sources. The easy shorthand is to frame these as vertical markets, which operate more or less independently of each other. But there’s another dynamic that suggests less vertical markets, and more single market.

Vertical markets are a familiar concept, so if I characterize, let’s say, retail, shipping, and agriculture as verticals, it’s a familiar framing. We understand separation between sectors, and seeing information products that align with that understanding isn’t a surprise. Go to an event for one vertical, expect everyone to frame their own offerings around the vertical.

Even at the single-sector event, you see companies who are clearly active in more than one vertical market. Data buyers look across diverse sources for what they need, and data sellers package their data for diverse customers. It’s the focus of the event that makes it look like an isolated vertical.

For hedge fund data buyers it’s even harder because you’re sourcing data for every investable industry – consumer, industrial, financial, b2b software, media, telecom, healthcare, etc.. Then there are the investable assets like commodities which have their own endless list of datasets.
— Matt Ober, The Rollup

I still find that vertical is useful shorthand to describe different markets, but lately I’ve been following up with my own definition for the purpose of the project. I’m less convinced that vertical markets for data are all that separate, and I’m thinking more of specialist markets defined by endpoints in the supply network. There are markets defined by who the customers are, and there are markets defined by where the data came from. To the extent that isolated vertical networks exist, their isolation will be noteworthy and interesting.

The view from the industry event is the customer-defined market: the customers are in the same industry, and the vendors position their wares for success with those customers. The source-defined market is more technological; those I’ve identified are built on large capital investments and significant technical expertise, and they sell to diverse customers.

Why it matters

I’ve met a lot of people who work with data over the years, most of them classically focused specialists. My view is that ignoring boundaries leads to discovery. If you’re the customer, you discover unconventional—should we say alternative?—sources. If you’re a vendor, you discover unexploited markets.

The point of the Data Market Study is to explore across these specialist fields to identify patterns and to discover sources, analytical methods, applications and customers. Ultimately, every market has to align with the needs of its customers, but I’m finding both customer-defined and source-defined useful as framings for choosing the deeper dives.

If customers are buying from sellers in multiple categories, and vendors are selling to customers in multiple sectors, it starts to look a lot less like vertical markets and more like one big market for data. And wouldn’t that be interesting.