SaaS for DaaS: Solving for Domain-Productization

Bridging the gap from “data asset” to “data-market-fit”

One of the biggest challenges I’ve seen in standing up a new data feeds business is “domain-specificity”. John Farrall has been writing a lot on the concept recently but, in short, it centers on the challenge of getting your data to sing to an end market.

This is nothing new: product-market-fit is always an existential challenge. But the nuance here centers on two things:

1. Data assets tend to have a tremendous number of use cases. Unlike vertically-oriented SaaS plays, a data play may be highly horizontal. Think consumer credit card data and the uses across a) researchers b) Wall Street c) corporate planning teams and d) audience targeting. Each of those use cases has unique buying behaviors, evaluation processes, competitive considerations, and even language.
2. A primary economic advantage of data commercialization is the ability to offset sourcing costs with “asset leverage” — taking a single asset and utilizing it for a number of purposes. The more you have to shape that asset into an end-use specific product, the more you lose that asset leverage.

Therein lies the struggle: how do you make a widely applicable asset resonate with hyper-discerning industry buyers without over-engineering into a narrow product?

Transactional friction, brokers, and middlemen

Given that the struggle above centers on transaction friction, the answer has been the traditional solve: the broker or marketplace. Wherever there are complex buyer/seller dynamics, there exist “smoothers” who reduce friction in exchange for a fee.

In the data world, brokers and marketplaces are commonplace. These can be generic- like the Snowflake marketplace- that simply make it easier to discover and access assets or specific- like ConsumerEdge- who partner with specific companies to make data available to specific end-users (in this case hedge funds).

A breed of “middleware” specific to data products has also emerged: tech that provides value-added, industry-specific functionality to bridge some of the functional gaps between the data product and the industry user. Take Datavant- the leading data ecosystem for the healthcare industry. Datavant doesn’t just “connect” data providers and consumers- it effectively manages the exchange in the most mutually-beneficial way to all parties, standardizing to the patient level while also ensuring full HIPAA compliance (among other things). In the modern data ecosystem, these companies, by providing the connective tissue that makes everything run, are, in my opinion, the most valuable players of all.

Opening new doors

The only downside of middleware plays is that they‘re oriented to connecting data providers and consumers. But what about aspiring data providers who may not be ready yet for an industry-specific connection? And specifically, let’s talk about the thousand pound gorilla in the data world- the Hedge Fund buyer.

Hedge Funds are arguably the de facto “potential buyer” whenever a new asset emerges: If our data can generate just a couple bips of Alpha, of course a Hedge Fund would pay millions for it! This isn’t false — but it’s being said thousands of times over and, by extension, the people listening have seen/heard/evaluated thousands of claims. Being the de facto buyer means you are also a very, very, very (keep going) sophisticated buyer.

So while Marketplaces could help you get discovered and Middleware can drive the ultimate exchange, it’s on the data provider to provide a compelling product. And so we’re back to the dilemma about assets vs products and leverage vs over-engineering.

Enter SaaS for DaaS.

They’re not brokers, they’re not marketplaces, and I don’t know that I’d consider them traditional middleware. Instead, they’re SaaS for Daas- domain-specific technology designed to translate a data asset into the language of the end-user, and in the most compelling way possible. Two companies in this vein, Althub and Covariance, was designed by researchers and hedge fund industry veterans to productize the standard transformation and evaluation process, (e.g. “tickerization”, back-testing against fund performance). This isn’t tech for hedge funds to find new data — it’s tech for data providers to bridge the gap between what they have and what they need to sell.

This concept carries with it a couple of evaluation requirements:

1. Does the tech meaningfully offload the productization burden from the provider? It’s a classic build vs buy question, specifically along the resourcing dimensions of a) onboarding the data into the platform b) translating the onboarded data to the buyer and c) making the post-sale exchange. Solving these burdens creates significant value add — but can you solve them scalably across data providers?
2. Does the tech carry a quality signal? If the end-goal of the platform is to help providers communicate to buyers, then the platform’s reputation might be the most important value-add. If Althub can become the Moody’s of alt data and provide a standard evaluation that carries that level of weight & trust- then they’ll have no shortage of providers banging down the door to use the platform.

The data ecosystem constantly surprises me. We started with “it’s really hard to productize generic data to specific end-users”, journeyed through the traditional solves (marketplaces/brokers), delved into big evolutions (middleware), and are now seeing new takes. I’m excited to see how AltHub evolves and whether this “SaaS for DaaS”, “outsourced productization” model will stick.

And if it sticks, can it scale across industries? Could we see a company out there truly emerge as a “data product transformer”- taking industry-spanning assets and processing them into premium, domain-specific “data products”? Let’s see.