Why VCs need to embrace a data-driven approach

At Balderton, we strongly believe that the value of good data and research in VC — from both an investment decision-making and portfolio support perspective — cannot be underestimated.

I joined Balderton Capital as Research and Data Lead within the Investment Team in August 2022, to continue evolving Balderton’s approach to research and data.

A number of firms have made great progress when it comes to data and research, but there’s still a lot more we could be doing as an industry to make the most of the opportunity at hand. I wanted to open up that conversation, and share our thinking behind why data is so invaluable in VC, and the opportunities and challenges it poses moving forward.

I. Where Research and Data adds value in the VC space

Rigorous research and data is fundamental to our approach to venture at Balderton. I wanted to highlight three key examples in which a solid research and data function can add value:

1. A reliable research and data function is instrumental to navigating the growing volume of investment signals available, enabling investors to source deals earlier

The source of these investment signals can broadly be split into several categories:

• Signals from data-providers with a paid subscription, coming with data extracts and API access. These can include market data, financial data, mobile data from well-known providers such as Pitchbook, Dealroom, Apptopia, Sesamm etc. New players emerging like Expert Focus are innovating in this space.
• Signals from other public datasets or metadata possible to scrape with open company financials or open-source software activity, such as Companies House, Github, etc.
• Signals from social media that might include further plugins or APIs like LinkedIn, Twitter, PhantomBuster etc.

2. As the number of companies leveraging AI, ML and data-science continues to grow, so too does the need for a deep technical understanding from investors

Both our early and growth teams are highly interested in startups within the dev-ops, ML-ops, data-ops sectors. In fact some of Balderton’s most recent investments include PhotoRoom (Generative AI-based photo editing app, based on Stable Diffusion), Ramp (full-stack payment in the cryptocurrency space partnering with more than 400 developers), and Levity (AI-powered workflow automation using no-code).

A deep technical understanding of how these businesses work, and the challenges they may face as they scale, is essential to being a good investor.

3. A dedicated research and data function can provide invaluable support to existing portfolio companies

There has been an increased appetite to benchmark investment metrics across the portfolio, both at Series A and Series B+, to more easily compare financials, growth metrics, and unit economics at scale. Some of these can be benchmarks in SaaS, but also consumer marketplaces, helping companies to compare metrics, and measure progress compared to peers in subverticals.

By opening up the conversation around data and research, not only in venture but more widely, one also creates valuable communities. Bringing together different roles across the portfolio, from data scientists to heads of analytics, can unlock strong networks and conversations to help portfolio companies on their journeys.

II. Growth of data teams in VC and Private Equity sectors

While the idea of trying to integrate data into VC investment decisions is not entirely new, recent breakthroughs by funds have come about in the past few years. One of the most obvious examples is EQT Ventures, who developed their software ‘Motherbrain’ using NLP to measure the closeness of companies. ‘Motherbrain’ uses numerical representations coming from PAUSE technique to measure closeness of companies. Firms such as InReach Ventures also have data engineers within their investment team to source deals earlier and build out proprietary software. A/O Proptech is building interesting data-driven methodologies to support their investments at Series B+.

Equally, funds don’t necessarily have to be investing in hugely expensive AI or ML engineering — the challenge for firms looking to build out their data teams today, however, is the widely-known shortage of technical talent. In the UK, 68% of digital analytics leaders highlighted a lack of skills as a blocker, and 57% said they will never have enough access to enough tech staff.

What’s more, the definition of what constitutes successful data-driven sourcing at early-stage is often grey: should we be measuring conversion rates of startups obtained from specific investment signals? Better adjusting valuations based on perfect market multiples? There’s a lot yet to be determined in terms of what industry best practice will look like.

III. Community-Building at Balderton

At Balderton, we love to bring together different communities from across our portfolio and wider ecosystem. Last week, it was an honour to host our data community for a dynamic discussion on ‘Growing a Data Analytics Function’ with panelists Markus Frise (Head of Analytics at Beauty Pie) and Chris Kelly (Head of Analytics at Wayve).

It was a full-house at our HQs, with a mix of audience from portfolio, friends of portfolio and investors in the space. It was great to hear from our panellists about their respective approaches, including their preferred tech stacks (DBT, Snowflake, Dataiku etc), preferred organisation of teams (‘hub-and-spoke’ model), and Wayve’s ML models. Stay tuned for more insights and events to come.

IV. Final Thoughts

We believe the trend in research/data within VC is just beginning, with increasing European and US-based players investing in technical talent and functions in order to make better investment-decisions and provide deep technical support to their portfolios.

If you are a technical founder or investor helping build out research/data capacities within the VC or Private Equity space, please feel free to reach out to mwehr@balderton.com.