People Who Interview People
Do Stories Speak Louder Than Data?
In this month’s newsletter, I want to explore the limitations and benefits of using data-driven approaches to understand people and their stories. VCs are increasingly focused on data, a phenomenon led by data science-specific funds like SignalFire, with traditional venture funds like Greycroft, Lightspeed, and Sequoia following suit. Sequoia offers various publicly available products like this interesting library of written content for data-informed product building. Tribe Capital, EQT Ventures, and e.ventures also employ data driven tactics according to this TechCrunch article. Each firm has a different data-focused methodology:
SignalFire - Leverages its recruitment prediction and market data analysis engines to make investment decisions and as value-added tools for its portfolio companies. SignalFire raised $500 million across two funds in 2019, and purports to pull from 100 datasets to understand what is happening in the world (e.g. talent flows and consumer spending) (TechCrunch article).
Tribe Capital - Built a quantitative framework for measuring important aspects of early-stage businesses, like growth and product-market fit. This Tribe article, A Quantitative Approach to Product-Market Fit, codifies analytical techniques that help guide investment decision making. Tribe Capital described their framework in this way, “Our approach to product-market fit is not a model that picks for us, rather, it’s a model that aids with seeing the world clearly. Our investment approach uses this as one of several inputs.”
EQT’s data-driven product is called Motherbrain. It uses convolutional neural networks, or CNNs, to review time-series data about companies to help guide investment decisions.
e.ventures (founded 1997) takes a human-plus-machine approach to investing, and in 2019 raised $400 million for two early-stage funds.
I’ve long wondered about the extent to which data drives success in early-stage investing. To me, data insofar as it helps with organization (follow-ups, sourcing, and pinging the right investors at the right time for deals that fit their thesis areas) seems to be of use, but can data really help make better early-stage investment decisions? In thinking like a data scientist, there are myriad arguments in support of data as an additive tool in an investor’s toolbox. Funds can scrape the web to understand when an entrepreneurial individual has left their current job, gain an understanding of key and emerging popularity and business metrics (app store performance, NPS, clicks), and perform competitive benchmarking (customer behavior, team, lifetime value). See Data-Driven VCs: How 83 Venture Capital Firms Use Data, AI & Proprietary Software to Drive Alpha Returns for more information.
Data in venture is more than a trend, it has quietly become part of many venture firms’ systems for assessing companies and determining whether to back or continue to support them. Check out this Forbes article for more information on data-driven VCs.
Data in investing will not be the sole subject of my analysis today. We will also discuss a correlated group of people who don’t make investments - interviewers. I asked numerous interviewers, in this context, defined as podcast hosts and newsletter authors (or, people who interview people), to answer some questions about how they source guests, manage their information, and to what extent they were data-driven in the choices they made. The specifics of who contributed responses will remain anonymous, but I was able to gather responses from podcasts and newsletters that most people in alohomora’s readership will be familiar with. My initial hunch was that interviewers and investors would have similar tendencies when it came to data, but the use of data was actually the primary axis upon which these two groups differed.
My favorite interviewer quote from this research was:
“The jurisdiction of what makes a good story or episode isn’t an exact science, but we can usually sense it during the interview and research phase.”
The key insight here is that an interviewer’s workflow is very similar to that of an investor. Both source, diligence, win, and help companies. You can find Nikhil Basu Trivedi’s thoughts on an extended version of the above process in his Venture Capital Flowchart. Through my chats with interviewers, I realized that interviewers think of themselves and their podcasts or newsletters as VC funds, and think of their guests as investments. The below graphic shows the core functions mentioned above, and how those map to the investor and interviewer workflows.
Below, find key interviewer insights as correlated to the aforementioned elements of VC.
Source: Without prompting, most interviewers compared their processes of finding interviewees to the process of sourcing companies in VC:
“I view myself as a pre-seed investor.”
“I think of my listeners as my board of directors.”
Diligence: Every person I talked to started with, “I don’t really use data in my approach - it is primarily qualitative.” All interviewers were organized and thoughtful in their approaches to finding interviewees, but none leveraged data-driven systems.
Each interviewer had a different focus area. Some prioritized storytelling and evoking a specific feeling, others looked for interviewees focused on a specific market, others looked for individuals with stories that contained elements of struggle and growth, and yet others optimized for quantity of interview content over quality.
Win & Help: This mostly happens post-publication. Many of the podcast hosts and writers I chatted with interviewed entrepreneurs, technologists, and business people. They provided value to those individuals through increased product or company discoverability, and positive press. Like in VC, these interviewers are trying to identify interesting brands and individuals. Some are focused on identifying these brands and individuals when they are early, and others are focused on sharing a story arc from day one to exit.
Takeaways
Despite all of the similarities between interviewers and VCs, none of the interviewers employed data-driven approaches in their work. This could be because podcasters and newsletter authors are often individuals and/or small teams, or because approaches to monetizing podcasts and newsletters for large-scale consumption are still in flux. In contrast, venture capital is competitive and established - VCs have to fight to differentiate and survive. VCs need objective successes to achieve financial return, whereas interviewers can tolerate, and sometimes seek, subjective successes - specific guests that speak to a niche audience. If an interviewer selects a bad guest, it is water under the bridge, but if a VC selects a bad company, it impacts their track record and the future of the fund. Though there are many similar elements between crafting a great podcast or newsletter and making investments, the stakes are different. Thus, VCs take advantage of data to feel more confident in decision making, whereas interviewers believe that stories can defy data, and that it is often the story that counts.
To conclude, I tried to tie these two worlds together with a simple graphic - both interviewers and investors seemed to fall into three categories: relationship driven, sector focus / thesis driven, or metadata & financials driven. This breakdown is by no means perfect, as many firms take hybrid approaches. Feel free to reply with your feedback! Again, keep in mind that the survey conducted was anonymous. Logos in this graphic are merely representative, well known examples of funds, podcasts, and newsletters in each of these categories.
Content of the Month:
Understanding - The best article I’ve read in a while by Nabeel from Palantir.
Window Swap gives you a preview of the world from different people’s windows.
NY Times: Reconstructing Journalistic Scenes in 3D
Notes from a Podcast with Matthew Ball - full podcast here.
