Data-Driven VC: what we have built at Red River West
In the last two articles we published, I talked about what is data-driven VC and what were the difficulties of building a data platform for a Venture Capital Fund. Today, it’s time for me to explain what we have built at Red River West.
I hope it will inspire some of our peers in order to build their own approach. Happy reading :)
To get a little bit more context, you can find the last articles here:
👉 Part 1
👉 Part 2
Introduction
I will begin this article by introducing Red River West and our investment strategy to allow you to understand where our data-driven approach is coming from.
Red River West is a cross-border fund investing in European startups typically raising Series B or C rounds in order to help them reach global leadership through commercial expansion to the US. Our strategy is to invest in a narrow portfolio of startups with a technological edge in order to help them with their US expansion.
This strategy implies that we spend a disproportionate amount of time on portfolio support. We are also known to do some of the deepest Due Diligence (startup analysis) on the market. It means we had to find a way to spot all the best startups in Europe without having the entire team focused on the sourcing effort. It was possible when we were operating just in France but it became complicated when we extended our scope to Europe.
I mean, how could you know of a promising startup in Poland or Estonia without an intimate knowledge of these markets, little network there, and no tools to track them?
This is why we decided to focus our tech effort on sourcing & screening. As these steps are at the beginning of the investment funnel, everything else follows, and by creating a platform improving and automating significantly these steps we knew we could significantly enhance our performances.
It led, in early 2020, to the creation of the first version of our tech platform: RAMP (Red River Algorithmic Management Platform).
Today, RAMP gathers more than 400 publicly available data points on more than 25,000 European startups and allows us not only to spot some of the most promising startups easily but also to analyze them in a much more efficient way by gathering insights in a single multi-layered Dashboard.
How is RAMP structured?
In order to spot and analyze promising startups, we first gather multiple data points on them from more than 10 different sources. Social networks, web visits, product reviews, tech stack, fundraisings, country repartition, news, etc. we gather all these data points, and much more in a single data lake that we aliment on a weekly basis.
Then, we had to focus on one of the most complicated steps of the data-driven sourcing process: creating a set of tools capable to clean and treat the data points that we collected to take care of missing data, false signals, duplicates, etc. (you can know more about this step and its challenges in my second article). It allowed us to use all the data points collected and cleaned to calculate multiple scores acting as signals for our team to know which startups should be our priority targets (this is step 3 from my 4-step plan in my first article).
As an example, here is a non-exhaustive list of some of the scores that we have built:
- The “Growth score”, as its name suggests, quantifies the growth or the momentum of the startup.
- The “US deployment score”: our thesis at Red River West being to help European startups to reach global leadership by expanding in the US market, we measure each startup’s US presence and assess their US business potential.
- The “ESG score”: by leveraging multiple data sources, we are able to estimate to what extent ESG matters for a given startup or vice versa, identifying potential red flags.
- The “Priority score”: this score rates every startup with a percentage and allows us to rank the startups that we have to contact in priority because they display exciting growth patterns, match our investment strategy, and because it’s probably the right time to contact them!
We can also use our Growth score to identify the fastest-growing startups in specific industries (this is something that some of our LPs who are sector-focused absolutely love by the way!).
Calculating a growth score
In order to calculate thes “Growth Score”, we use multiple data points as a proxy for the real growth of the startups (because startups don’t publish their exact revenues/profitability metrics and publication in official registry are partial and outdated). We combine them with different weights based on a complex algorithm that changes the weights of these metrics depending on the business model of the startup (B2B or B2C) and the importance of each metric in the real growth of a startup.
As a simple illustration, for a B2B enterprise startup the “sales employee growth” is a better proxy than for a B2C startup for which, the best proxy can be “website visits” or “follower growth”. Our algorithm accounts for all these subtleties and automatically adapts thousands of different weights every month.
In addition, this score takes into account the size of the company: it is much easier to grow fast when you start from a low base! To have a meaningful benchmark, we use a statistical approach to take that into account and adjust the growth score.
Competition mapping
On top of all of that, we are improving more and more the “startup analysis engine” of RAMP. Today, it allows us not only to identify startups and perform the first screening but also to find the most relevant competitors and analyze the entire competitive landscape of any startup. We can identify who are the leaders & the most serious (fastly growing) contenders of any market and then compare the features of their products, their metrics, teams, etc. It speeds up greatly the deal flow process and allows us not to miss a single competitor.
Importance of UI/UX:
We spent a lot of time talking to Tier 1 VCs in the US on data-driven approaches: some of them have built extremely powerful platforms but they hardly use them!
Indeed, gathering data, building insights, creating various alerts, etc. is not enough…
It’s not because you’re building an internal product that you don’t need a product manager or a designer. People use platforms they love, not platforms that are just “useful”.
We’ve had a strong focus on the UX and UI of the platform: we wanted the platform to be used by every member of Red River West on a daily basis and we knew that valuable insight wouldn’t be enough. Users would have to love the platform for this objective to be achieved! By using low-code tools and hours of user interviews and feedback, we designed a great interface that allows our team to find all the information they need easily and provides several levels of analysis.
Today, RAMP is used on a daily basis and enabled us to spot hundreds of promising startups all over Europe.
Conclusion
I could expand for hours on all the other modules & automation we have built, what we’ve learned, the data sources we use & the ones that were actually not that relevant, the insights we extract thanks to complex algorithms, etc. but we have to keep a little bit of our secret sauce for ourselves… 🤫
We also have a very ambitious roadmap and are currently in the process of building better tools for market monitoring, deal flow management, due diligence, and much more. Besides, we have identified a few key modifications and additions to implement in order to further improve our sourcing performance thanks to RAMP. We will iterate fast toward building one of the most advanced VC platforms in the world! I’ll keep you posted 🔥
Oh, and one last thing …
RAMP also allows us to share regularly and easily valuable insights with our LPs. Because we’re nice people, we decided to share a bit of these insights with you too. Indeed, thanks to our growth score, we are able to identify the fastest-growing startups in specific industries or sectors.
Would you like to know which are the fastest-growing startups in Food tech ? in gaming? in carbon accounting? and much more?
If so, you can subscribe here: Newsletter
