At ValueStream Ventures (VSV), we are often asked how we approach value creation within our portfolio companies. We broadly categorize our efforts into three key areas: growth, operational improvements, and product evolution.
Growth and Operational Improvements
The first two categories, growth and operational improvements, are exactly what they sound like. We work closely with management to drive higher KPIs through customer introductions, improved operational efficiency, team dynamics, go-to-market strategies, and customer retention. Each company’s needs are unique, so rather than deploying a single “playbook,” we leverage the experiences of our partners, our network, and our co-investors to address specific gaps in each company. No single individual can tackle every challenge, so connecting management teams with the right experts is a crucial part of our work.
This is a lot of blocking and tackling, but these value-creation tools are table stakes for most investors in early-stage companies. What truly differentiates VSV’s approach is our focus on the third category of value creation: product evolution.
Product Evolution
Our core thesis is that novel datasets can drive better outcomes for our companies. However, for this to happen, the datasets need to catalyze a product evolution that can increase TAM, revenue, and liquidity. The vast majority of datasets (even highly unique ones) do not have this potential, but sadly, we also see many that do have this potential but where it is never realized.
To visualize our approach conceptually, we like to think about how the value of data increases and how that data can create increasingly valuable products. Below is a conceptual chart illustrating our approach:
We start with data that is useful because any data without an apparent use case will, of course, have limited (if any) value. Assessing the size of the problem that can be solved using a particular dataset is not always obvious at first, and so we attempt to cast a wide net in the types of companies we engage with.
We then move on to whether the data is hard to come by. Data such as SEC filings can be useful for things like company financial analysis but is not hard to obtain. If data is too easy to access, we consider it low value for generating enterprise value despite its utility. Hard-to-get data, like multi-spectral imaging from orbiting satellites, requires a more significant expense to obtain, limiting the number of companies that can access it.
Network effect data is obtained by having a network of customers or partners, either through voluntary contribution, like a data co-op, or as a byproduct of tool usage provided by a company. This data is more defensible because it requires a go-to-market motion to acquire. Our portfolio company Windfall operates a contributory data co-op to aggregate consumer attributes, enabling highly accurate wealth estimation models. Our portfolio company Alchemy, on the other hand, provides laboratory management software for capturing data on results of chemical compound development. When aggregated across a large set of customers, this data can be used to develop prediction models for chemical properties of compounds, drastically reducing the time to develop custom products for corporate clients.
The holy grail of datasets is the feedback loop, where the use of the data results in supplemental outcomes data that can refine the dataset over time. Our portfolio company Ocrolus, for instance, combines OCR with a human-in-the-loop AI system to accurately extract financial data from documentation. The human reviewers correct any OCR AI mistakes, which then improves the dataset over time, creating significant defensibility. Further, the company leverages outcomes data from clients to develop fraud detection models for things like fake statements and loan stacking (applying to multiple lenders simultaneously in an attempt to circumvent credit limits).
From Data to Services
Once you have a dataset, the simplest and lowest value product is a data product, sometimes called Data as a Service (DaaS). These products can be incredibly useful and profitable but can be difficult to drive large enterprise values. Selling raw data places the onus on customers to know how to leverage the data, have the right people to analyze it, and operationalize it within their company—all skills that are in short supply.
Using your data to offer unique SaaS features powered by the data expands TAM. More individuals within an organization can leverage these tools, typically resulting in higher contract values. Leveraging datasets to train AI models creates even more value by reducing the need for customers to analyze the data themselves. The best AI models can generate predictions, recommendations, and insights that are useful across a broader set of employees.
Our portfolio company, Windfall, began as a data co-op that helped its customers increase marketing spend ROI through data sharing. Over time, by combining this data co-op with public records, the Windfall team was able to develop models for estimating the wealth of an individual based on nothing more than an email address. After pushing deeper into their data partners, the team was able to start obtaining outcomes data (e.g., donations) to build their first propensity models, which could predict the likelihood of an individual donating to a particular campaign, organization, or school. All of these AI models have proved invaluable to their partners and others in optimizing their marketing spend, so each dollar spent on customer acquisition has a greater chance of success.
The ultimate TAM creation is in services, where a vendor performs automated actions for customers, so they don’t have to do anything. Historically, services have been a dirty word in Venture Capital due to lower profit margins and lack of scalability. However, with novel data and customized AI models, vendors can offer services with gross margins as high as SaaS and infinite scalability by removing humans from the process. This approach allows software companies to solve customer problems directly rather than selling tools for customers to solve their own problems, resulting in the highest TAM potential for any dataset.
Verikai was a VSV portfolio company that was a strong example of this evolution. The business initially began as an analytics tool that combined behavioral data elements with de-identified health insurance claims data. Insurance brokers could use this tool to support their customers in negotiations with health insurance carriers. While data businesses like this can be highly profitable, there is typically a limit to how much a company can sell its data for as well as the number of potential customers. VSV worked with the team at Verikai to outline a plan and introduce key stakeholders to develop a new type of stop loss insurance based on behavioral data rather than traditional census-based data. This led to a significantly increased TAM (the stop loss insurance market is >$30BN annually), the sale of Verikai to a leading insurance group, and the launch of a new specialty insurance company called Radion Health, which became part of the VSV portfolio.
Our Value Proposition
99% of early-stage technologies don’t fall on the earlier chart because they either don’t own their data or it isn’t inherently useful. For the 1% that do, our unique value creation lies in pushing those companies as far to the upper right as possible. Not every company will reach the upper right, but large businesses can be built by moving one or two steps in either direction.
It’s worth noting that this is a framework for value creation, not value calculation. Different markets have different TAMs, which involve traditional research and due diligence. Even in smaller markets, we often find opportunities for TAM expansion using this methodology. Smaller markets, such as Vertical SaaS, provide some of the best opportunities for us, as they are less contested by large incumbent capital providers who are heavily focused on growth as the primary source of value creation.
This strategic engagement is what companies come to VSV for and why they often feel we have an outsized impact on their trajectory. It’s also the type of creative exploration on which our partners thrive.