Join the syndicate

AI technologies in VC industry


How web3 and AI technologies will fit together and how will they affect the investment decision-making process in general? The topic is vast and interesting, and it is obvious that there are not so many intersection points yet, for example, the monetization of digital assets, i.e., company data will represent a special type of assets, and they will have to be stored in distributed and verified networks, since the world around will be full of fake data generated by the same AI networks, and only distributed registries and trained AI algorithms will be able to understand where the truth is and where the fake is, so investors will look for sources of verified information and data. [Examples of emerging startups working in data verification: Golden.com with more than $50M USD in total funding, and Optic.xyz, which raised $11M USD seed round in mid 2022].

It turned out that sentiment analysis of news and search for signals in the market is the most expected and still missing. So, for example, with an increase in the volume of site visitors and an increase in the company’s mention in the media, as well as the appearance of mentions or reposts of news from the CEO of the project by some angels, the system could issue a signal that can be interpreted as ‘important’, and the company would appear on dashboard of the investor marked ‘study in detail’.

Another important part of the product could be finding the ‘right’ founders, and there are many different approaches to valuation, of which I liked Ali Tamaseb’s study, where he breaks down 50 signs of the ‘right’ founders who built billion-dollar companies. One could look at early investors mentioned in this study, such as Elad Gil (Mixer Labs, Color Genomics), David Sacks (PayPal, Yammer), Biz Stone (Twitter), Peter Thiel (PayPal), Alexis Ohanian (Reddit), Mark Benioff (Salesforce), Matt Oko (Yes Vinci), Kevin Hartz (Eventbrite) and many others. Such founders are also successful business angels, but it is only possible to monitor their investment activity through automated systems because there are a lot of such people, and it is even better to track when several of these angels entered into one of the early transactions.
We can also recall the study conducted by Andre Retterath and @Reiner Braun, where they described information asymmetry, when founders provide the market and investors with exaggerated indicators and try to be better than they are, while investors are trying at this time to find reliable information about the market and companies, and because of the difference in numbers and expectations, there is an additional risk, which is covered by excess venture capital.The main conclusion is that the market is still in the very beginning of its journey, and we are not seeing a large adoption and most likely will not see it since the entire industry still lives on warm inbound connections. However, there are several tasks or core cases when data-driven approach may also be appropriate:

  1. Intelligence or identification — when the investor wants to conduct a macro screening of markets, founders or companies and find general information about how a particular segment or company is developing.
  2. Research work — when the investor monitors already identified opportunities and conducts a deeper analysis, scoring or additional information that will help make the final investment decision.
  3. Current portfolio interactions — when the investor has already made some investments and does not have enough data or time to track how these investments are developing in accordance with his expectations.

Each of these cases can be partially closed with current solutions, for example, points 1 and 2 through Pitchbook.com, Harmonic.ai, CBInsights, etc., and point 3 through edda.co or visible.vc, but all of them are either very expensive — $30k+ per year, or do not close all the functionality, so what is the next big thing of the product that can work through all 3 cases on the nascent data-driven trend? In my opinion, it is the ability to provide market signals and sentiment analysis that can show the activity of a particular company in the market, compare it with successful patterns and build various predictive models.

Sentiment analysis is systematized information about the connection and context the company is mentioned in, what its digital footprint is, including user comments, publications of the founders, custom PR articles to raise the round, etc. All this can be analyzed and compared with the most successful benchmarks, and this can greatly help all current players, including small funds and business angels, when making certain investment decisions.

The report by Andre Retterath and his team on the ongoing changes and dispositions in the @data driven VC market clearly shows how an Army of investorsis building their systems for collecting and processing disparate data, improving their awareness of technologies, founders, market niches and much more, which can significantly affect the success of potential investments. Such developments will allow in a couple of years to significantly change the way of analyzing projects and forming pipelines within funds (non-obvious deals will be enriched with data), build competitor buying strategies, identify the best matching co-founders and promote more accurate marketing campaigns. These trends will become an integral part of Venture 3.0.

So, what can we expect from the introduction of AI in venture capital investment?

  1. Openness of data about companies and investors, both public and more closed, will become ubiquitous and there will no longer be a question of who does what and whose investment thesis it is, or how much this company earned from its clients, and who those clients are — all this will be available especially in the early stages, when the founders will seek to disclose their performance to potential investors in order to get on the radar of desired funds, and automated scoring systems will notice certain market signals and signal partners to contact a specific company.
  2. Fundraising will become super automated, as the funds themselves will talk about their preferences and issue public APIs so that companies can upload their metrics and data for analysis. Such robots will evaluate the focus, stage, publications of a specific founder, match with the right partner or investor, and help find a common relevant background and experience.
  3. As a result of super automation, being present in media and on social networks will become even more important for up and coming startups looking for fundraising: the more you share, the higher chance to be found and tracked by some VC fund’s AI robot. There will be no need to spam inboxes of hundreds of funds, if you’re visible enough and have valid business growth, they will find you faster than you think.
  4. Do you remember how public stocks of companies were traded thirty years ago? There were faxes and landlines, people wrote notes or made decisions based on some expert’s belief in economic macro indicators. Now, most retail terminals look like a NASA mission control center, with built-in robots and scrapers and much more. The same path will be trodded by private capital markets. Soon the best deals will be on the radar of trading terminals, and the accounts of companies will be automatically replenished based on the allocations opened by the company, and if one can, he/she will become an investor either through a system of automated robots tuned to a specific type of company, or through distributed investment evergreen funds.
  5. Data will become more valuable and scarce: Big platforms, data carriers like LinkedIn, are changing the rules for collecting information. So, for example, if you read their T&C, it becomes immediately clear that you get a ban for automated data collection. Other large aggregators will protect their information even more strongly, because it will become the main source of business insiders. Generative AI also adds up complexity to the data problem: with Gen AI the web data volume will increase dramatically, and finding and fact-checking “ground truth” data among terabytes of robot-generated content will be another challenge to be solved.
  6. Predictive models that evaluate company performance, market segment sizes and potential opportunities will also become ubiquitous and will describe the sales volume of both a small company and a corporation so that founders and CEOs can write in their presentations more accurate market numbers and sales potentials for these markets, for example, to determine the most accurate TAM/SAM/SOM and due to this, to better see the startup business itself.
  7. A company’s M&A strategies will change completely as end-to-end data will match companies that are best suited in terms of business and size, and cumulative effect when combined, so that they can merge and build a more meaningful business. There will be no need to grow super-fast on your own, it will be possible to find companies that are most suitable for specific tasks, and buy them, thereby closing more and more niches and building potential synergies, which will provide faster returns to investors (exits).
  8. Diversification with automated investment decisions will increase, which will give a new impetus to the development of entrepreneurship, because every company that is able to provide a certain set of potential successful triggers or market signals can receive its share of investments. This will allow even the most inexperienced entrepreneurs to launch their craziest ideas and skillfully manage their sales by seeing the market and potential customers.

June 01, 2023