The economics of data businesses

Today, data has made its way to becoming a key cog in the modern economy alongside land, labor, capital, and oil. But unlike others, data is unique in a way that it can be used by multiple people simultaneously and repeatedly without being depleted. The value of data gets unlocked when it is accessed and churned by businesses to innovate and create knowledge. As we know, every company uses data in some form or the other to run its business. But what does it mean to be a data business today?

In today’s time, an organization can call itself a data business only if data is its core product. Data is central to any organization and without it, there is no organization. We have seen many organizations, which are data businesses that have come up in the last couple of decades that are straightforward examples that come to mind, some of which are old organizations with a (with a 100+ history).

Let us dive a little deeper on some of the key tenets of data businesses.

• Humble beginnings. Many a time, data businesses seem unreliable at the beginning, and this is because data acquisition is not easy. One needs an upfront investment or a certain amount of work to collect data of the scale that is needed to even start work. Data below a certain size is simply not useful and one needs MVD (minimum viable dataset). This essentially means data of a minimum size and of a certain quality to signal value in the market. Along with having minimum data, one also needs to ensure that the data is of a very high quality even though it may be in unstructured form. In the data world, good is often defined by accuracy and being comprehensive, and achieving both is imperative and important.
• Size and network of data as a Moat. The moment you capture unique data, you also capture unique value. The size, quality, grain, and network effects of data are the moats. Just like in wartime moats provided a strong line of defense from enemies, in business parlance and in our world, these moats are the ones that will keep competition at bay. It will take competition that much longer to come close to the same size, quality, and grain of data. What this means is – data is the novelty here, it is the business and it is the value. Once data is captured, computing it is easy. Some of the world’s leading search engines are a great example of crawling the complete worldwide web data and page ranking them basis network of links between them and click history from users.
• Data Businesses are super sticky. As we move forward, the marginal cost of acquiring data begins to decline over time. Since the data corpus is learning on the fly, it is refreshing itself without having to wait for periodic updates and refreshes. This makes the data business super sticky as they are virtually impossible to displace. Churn rates for mature data products are negligible; therefore, they are making established data businesses difficult to disrupt and replace, e.g., today a leading job review website empowers job seekers with straight-from-the-source insights, reviews, and corner-office intel so they can make their next career move. With its existing corpus of data, it shows reviews of companies to prospective employees. Once people sign up to this job review website, they can view reviews of other companies and review companies they are working for and contribute. This in turn allows this website to create a larger corpus of data. The more people use and trust this website, the more organizations take it seriously. And as users, see more and more people contributing to the said website, they can be more confident they will stay anonymous when they add their review.
• Long-haul winner. Like wine, data businesses get better with time. It is a winner takes most market, with the ultimate result being a small handful of companies being able to sustain and control majority of the market share. In SaaS, a particular generation of software can become obsolete in a few years, but data businesses can last for decades. The idea should be to invest in building a data asset once and keeping customers forever. That way, as your data corpus scales, your sales become effective and easier and then you can start slicing and dicing data for effective targeting and price discrimination. Once the data is widely used, it goes from being optional to being essential, wherein we have various use-cases for upsell, cross-sell, and recurring revenue. The ones that fail overtime are often killed by irrelevance with time rather than by competition.

As we have already established, the first fundamental truth of a data business is that it is all about data and they are all built around unique and/or proprietary economic and customer value. There are two basic factors here that are important to understand. First is the source of data i.e., the ecosystem that generates it. There could be various ways to build this asset and some of them are brute-force-based primary data collection, outsourced data acquisition, proprietary data generated as an output of core business processes, standardization and aggregation of nonproprietary data, creation of data content loops where data is captured in exchange of a service or information, and data creation where synthetic data is generated in cases where real data is unnecessary.

The second factor is where this data sits in the value chain. Smart orchestrators understand the data value chain and set their strategy accordingly. You could be a business that offers data, or you could be a business that offers insight, which is undeniably more valuable than just data, or you could be a business that offers a service based on data.

For organizations wishing to jump start their data business, value proposition needs to be constructed keeping the following in mind.

• Gain creators – What suite of data products and offerings will generate the best payoff for the customer? Where should we fit it into the value chain – data, insight, or service? What analytics or additional data would be needed to maximize impact for customers?
• Pain killers – Biggest challenges faced by customers and how data generated by the organization will help address these?
• Enablement – Operating model to deliver value and the organization structure required to deliver the same.

Although the focus should be on all three, gain creators end up creating maximum utility value for customer and economic value (or valuation) for the business. As we have established earlier, there is a need for non-data businesses to become data businesses overtime, because they are sticky and difficult to compete with. Therefore, it is imperative for organizations to start pondering on the kind of data they have access to and what real customer value is. They should use the power of cloud hyperscalers (as opposed to old school on-premises system) to achieve their goals as they offer faster and better return on investment.