When I started Groove Capital, I had a number of convictions around the types of startups and the qualities of the founding teams that were appealing. Those criteria ultimately became our investment thesis, which has remained largely intact five years and 90+ investments later. Despite our north star remaining constant, things have changed—we’re always learning and adapting as the results of our earliest investments give us more data to refine our investing model.

Along the way, we’ve been fortunate to be surrounded by likeminded investors who have served as mentors, teaching us how to build a learning organization that gets smarter with time. When it comes to operating systems, Daren Cotter is the G.O.A.T. (hear Cotter on episode 131 of By All Means); the data he collects and the rigor he applies to his investing is next level. Having benefited from his tutelage, I thought it would be value-ad to our angels and early stage investors to share a behind the scenes look at some of the tools, methodologies, and processes that we both use to enhance our combined portfolios of 200+ investments.

The early days: Build the habit of documentation

When Groove first started investing, I created a standardized application that would allow our team to evaluate and compare responses across a variety of teams and market opportunities. The app was built from lessons learned from my own startup experiences, along with observations from running BETA (a local accelerator). From the beginning we were tracking things like…

• Company characteristics: category, stage, business model, and competitive landscape
• Founder attributes: experience, passion for the problem, team composition
• Traction: customers, revenue, growth rate
• Fundraise terms: valuation, structure, prior capital raised

Following a pitch and due diligence, our team would consider all of the above attributes (and more) and score each deal in a weighted rubric. That scoring exists today, and regularly comes in handy to support our team in making new investment decisions. And as the results of our early investments reveal wins and losses, we have the ability to tune the model for even greater performance.

Daren’s documentation started similarly. Both of us began with basic tools like spreadsheets, folders for storing updates, pitch notes, and diligence memos. But over time, what we tracked and how we tracked it has varied by the stage and role of our investing (angel/LP vs. VC). In addition to the criteria above, Daren’s documentation includes:

• Subsequent fundraises and transactions, which enables tracking a variety of performance related metrics
• DPI (distributions to paid-in capital), TVPI (total value to paid-in capital), and IRR (annualized return)
• Founder communication/updates
• Graduation rate (companies raising additional capital in formal, subsequent financing rounds following the completion of stage-specific milestones)

Without good notes, it’s tough to recall why you invested in the first place. The key: start early, even if it’s simple. Know that adding new variables is a normal part of the process, but be mindful that it can skew your data when you try to apply things with hindsight.

The patterns that matter (and some that don’t)

The goal of documentation is to leave yourself with a breadcrumb trail that can lead you back to the criteria that most reliably leads to a favorable outcome. With 200+ investments between us, here’s the one thing that has led to a 100% likelihood of a 100X outcome…c’mon, even if this did exist, do you think we’d share it?

Here’s what’s been worth tracking:

• Founders with direct experience solving the problem at hand outperform founders with general startup experience
• Founders with a deep passion for the problem outperform founders with a casual or surface level connection to the problem, likely because the former are more inclined to stay committed when things do not go to plan
• Timing is everything. The question to ask here is, assuming the team is able to execute, is the market ready for this?

And what surprised us?

• Demographics haven’t correlated as strongly with outcomes as one might expect
• Our own conviction level at the time of investment hasn’t always been predictive—which is an argument for fixed-check sizing over conviction-based investment sizing
• Founders with previous exits may be great bets, but may also be quick to kill a project that isn’t going to plan, or overestimate their skills as they operate in new industries or alternative business models

No matter how good your dataset is, it can’t beat your access

It’s easy to think that if you simply improve your picking—better diligence, better scoring—your returns will follow, but here’s the uncomfortable truth: access > analysis. You can improve your picking process tenfold (and you should), but without access to the best opportunities, it won’t matter.

Knowing what to do with a deal once it is in front of you is the point of this piece, but it’s paramount to consider the quality of the deals that are being introduced to you. To enhance both functions, we’d both recommend keeping track of where you are getting your deals. By monitoring the source of your deal flow, you may start seeing themes that can either continue to lead you to better investments, or expose a lack of access to the very founders who are most likely to produce that portfolio-returning event.

Learning is a (fun) part of the process

As professional investors, both Daren (as a full time angel) and I (as a VC) need to generate outsized returns over the long haul to continue to do what we love, and thus the insights shared here reflect the processes that we’ve developed to achieve that outcome. If you’re new to early stage investing, a simple way to improve your investing is to take good notes. From there, your systems may evolve as your investments mature, and it’s this evolution and learning that makes the whole experience so rewarding.