The Numbers an Investor Network Should Track

Ask most angel network or accelerator leadership teams how their organization is doing, and the answer arrives fast: deal flow is up, membership grew, the last cohort was the biggest yet. Those numbers are accurate and worth having. They are also, on their own, close to meaningless, because none of them says whether the network is doing the one thing it exists to do, which is help good companies get funded and help members make better investment decisions.

This is not a niche problem. It shows up at the academic level, in the industry's own trade data, and in the day-to-day reporting that most networks build for their boards, and it points to the same conclusion from three different directions: the easiest numbers to collect are rarely the ones that matter.

The Research Problem Is Worse Than Most Operators Assume

A 2025 review of accelerator performance research found sharply contradictory results in the literature: some studies show accelerated startups are 3.4 percent more likely to raise venture capital and grow sales 2.7 times faster than non-participants, while other research finds many programs fail to accelerate anything and may even hold startups back. The same review put the average success rate across accelerators globally at just 25 percent, and traced the disagreement to specific measurement failures: selection bias, where the startups likely to succeed anyway are the ones accepted into the program, attribution problems from short program windows, and evaluation methods that lean on self-reported outcomes with no control group to compare against.¹

None of that is a reason to give up on measurement. It is a reason to be honest about what a given number can and cannot tell you. A network that reports its cohort's aggregate funding raised, without asking whether those companies would have raised anyway, is repeating the exact error the research identifies. The number is real. The causal claim behind it usually is not.

Activity Metrics Answer the Wrong Question

The instinct to report activity is understandable. Deals reviewed, meetings held, applications received, and membership counts are all easy to pull from a CRM and easy to put in a slide. They also tell you almost nothing about quality. A network that screens five hundred companies a year and funds none of them is not obviously doing better than one that screens fifty and funds ten, and a membership count says nothing about whether those members are showing up, doing diligence, or writing checks.

The distinction that matters is between activity and outcome. Activity metrics describe what the organization did. Outcome metrics describe what happened as a result, to the companies, to the members, and to the portfolio. Most networks over-invest in the first category because it is available on demand and under-invest in the second because it takes longer to show up and requires tracking something past the point where the deal closed.

What Member Engagement Predicts

Queen City Angels ran a six-year internal study on member engagement, defined as active participation in due diligence, governance, and mentoring rather than passive membership, and found a direct relationship between engagement and investment behavior: as the group's membership more than tripled, the average amount invested per deal more than doubled, and members who engaged more deeply with the group's process became meaningfully more likely to write sidecar checks alongside the group's lead investments.² The same study tracked portfolio company health on a simple six-point scale from writeoff to successful exit, and found 85 percent of the group's active portfolio graded at the midpoint or above.

What makes this useful is not the specific numbers, which will differ by group, but the structure underneath them. Engagement was not tracked as a vanity metric about attendance. It was tracked because it predicted something the network cared about: bigger checks, more sidecar participation, and a healthier portfolio. That is the test worth applying to anything on a network's dashboard. Does this number predict an outcome members care about, or does it just describe an activity that happened.

A Practical Set That Does Not Require a Data Team

Four categories cover most of what a network needs to track, and none of them require enterprise tooling.

Conversion, not volume. How many companies move from application to first meeting, from first meeting to due diligence, and from due diligence to a term sheet. A network with a low overall deal count but a strong conversion rate at each stage is doing better screening than a network drowning in applications with almost nothing making it through, regardless of which one looks busier.

Member participation depth. Not how many members exist, but how many are actively reviewing deals, sitting on diligence teams, or mentoring portfolio companies in a given quarter. This is the number that predicted check size and sidecar behavior in the Queen City Angels study, and it is available to any group willing to track attendance and volunteer roles rather than just membership rolls.

Portfolio health past the close. A simple graded scale, even a rough one, applied consistently to active portfolio companies on a regular cadence. The point is not precision. It is having any signal at all about how funded companies are doing after the check clears, since that is the outcome the whole process exists to produce.

Time to decision. How long it takes a company to move through the network's process, start to finish. Founders talk to each other, and a network known for a slow, unpredictable process loses access to the stronger deals before members ever see them, regardless of how good the eventual decision is.

The time cost of tracking this well is smaller than most leadership teams assume. Groups that have built this discipline report tracking fifteen to twenty metrics across fifty to a hundred members in roughly forty to forty five hours a year, well under four hours a month, particularly when portfolio companies themselves supply part of the reporting rather than staff reconstructing it from scratch.³

One trap worth naming directly: benchmarking against other networks is tempting once these numbers exist, and it is usually the wrong instinct to lead with. A network in health care deal flow will have a different conversion profile than one focused on consumer products, and a young network still building its member base will look different from one fifteen years in. The comparison that matters most in the early years is a network against its own history: is conversion improving quarter over quarter, is participation depth growing or shrinking, is portfolio health trending in a direction anyone would be comfortable explaining to a new member. External benchmarks become useful later, once a network understands its own baseline well enough to know what a meaningful gap represents rather than chasing a number that describes a different kind of organization.

Reporting This Well Matters as Much as Collecting It

A network that collects the right four categories and then buries them in an annual PDF nobody reads has solved only half the problem. The value of outcome metrics comes from members and leadership looking at them often enough to notice a trend before it becomes a crisis, not from having them technically on file somewhere. A conversion rate that has quietly dropped for two quarters running is a useful early warning. The same number discovered a year later in a report nobody opened is just a historical footnote.

The practical fix is a short, regular cadence rather than a single exhaustive annual report. A brief quarterly update covering the four categories, shared with the full membership rather than just the board, does more to shift behavior than a thick report released once a year. Members who see their own participation reflected back to them, alongside what that participation has historically predicted, have a concrete reason to engage more rather than a vague sense that they probably should.

Where This Fits for a Network Using Pynn

This is precisely the layer a deal flow CRM and assessment platform should be removing friction from, not adding to. Every one of the four categories above already exists somewhere in a network's workflow: application data, meeting notes, diligence assignments, portfolio updates. The failure mode is not a lack of data. It is that the data lives scattered across email threads, spreadsheets, and individual members' memory, which means nobody assembles it into a picture until someone asks for one at year end, by which point the pattern that would have been useful to see months earlier is already stale.

A network that tracks conversion, participation, portfolio health, and time to decision consistently is not just building a better board report. It is building the evidence base to know whether the network itself is getting better or worse at the one job it has, which is turning good companies and good investors into good outcomes for both.

Sources

¹ Administrative Sciences (MDPI), KPIs for Digital Accelerators: A Critical Review, 2025. https://www.mdpi.com/2076-3387/15/7/258

² Angel Capital Association, How Member Engagement Can Grow Your Angel Group's Investments. https://angelcapitalassociation.org/blog/grow-your-angel-group-investments/

³ Angel Capital Association, Methods To Capture Key Data Elements On Investments and Outcomes. https://angelcapitalassociation.org/blog/methods-to-capture-key-data-elements-on-investments-and-outcomes/