Analizing The Data Gap with Vestberry: Why Venture Capital's Biggest Blind Spot Is the Portfolio It Already Owns

Venture capital has never had more data at its fingertips — yet once the term sheet is signed, many firms revert to fragmented spreadsheets, stale updates, and institutional memory to manage the portfolios they've spent years building. The front end of the funnel is obsessively optimised; the back end is largely improvised.

Analysis of over 850 VC funds shows that structured portfolio monitoring consistently correlates with meaningfully higher IRR and TVPI. Most firms haven't made the leap — not for lack of tools, but because poor oversight compounds quietly, over years, until it's too late.

Marek Zamecnik, Co-CEO of Vestberry, has spoken with over 2,000 VC firms over seven years and seen the full spectrum of how funds operate behind closed doors. In this conversation, anchored in Vestberry's recently published white paper, he explains why the industry has normalised the chaos — and what it takes to finally fix it.

You've spoken with over 2,000 VC funds over seven years. What's the most surprising thing you've learned about how venture firms actually manage their portfolios — versus how they think they do?

The gap between perception and reality is consistently striking. Most GPs genuinely believe they have a handle on their portfolio — they know their companies, they talk to founders regularly, they have spreadsheets. But when you look under the hood, the data is fragmented, often months stale, and living in five different places that nobody fully trusts. The surprising part isn't that firms lack data — it's that they've normalized the chaos. They've built workarounds so elaborate that they mistake them for a system. The firms that have made the leap to structured portfolio management are almost universally shocked by how much they were missing, not because the information didn't exist, but because they had no systematic way to surface it at the right moment.

The white paper argues that portfolio management is where "data can make the biggest impact," yet it remains the least systematized part of the venture stack. Why do you think the industry has been so slow to catch up here, especially given how data-obsessed VC is on the deal sourcing side?

Deal sourcing is competitive and visible — you either win the deal or you don't, and the feedback loop is immediate. Portfolio management, by contrast, is a slow burn. You hold investments for 4 to 10 years, and the consequences of poor oversight compound quietly over time. Nobody gets fired in year two for not having a risk dashboard. The urgency just isn't there — until it is, usually when a company fails in a way that could have been caught earlier. There's also an identity issue. VCs have always prided themselves on judgment, pattern recognition, relationships. The idea that a system could improve on that feels culturally uncomfortable in a way that, say, a better deal sourcing tool doesn't. But the evidence is clear — analysis of 850 VC firms shows that funds investing in structured post-investment support consistently generate meaningfully higher IRR and TVPI. The data is making the case for itself.

"Data-driven" has become something of a buzzword in VC. What separates a firm that genuinely operates this way from one that just bought the right tools and put up a dashboard?

The dashboard is the easy part — and honestly, it can be a trap. Firms that just bought tools often have beautiful visualizations sitting on top of dirty, incomplete, inconsistently defined data. That's not data-driven, that's data-decorated. The real separator is whether decisions actually change based on what the data says. Does the team trust the numbers enough to act on them? Is there a single source of truth that everyone — investment, finance, operations — draws from, or are there still three versions of the cap table floating around in email threads? Genuine data-driven firms have closed the loop between insight and action. They've standardized KPI definitions, they track actuals vs. budget at the company level, they have alerts that fire before a quarterly review. The mindset shift is from "we look at data when we prepare for board meetings" to "data is what triggers our interventions in the first place."

Off-the-shelf data platforms like PitchBook and Crunchbase are now widely accessible. You argue that the real competitive advantage comes from proprietary portfolio data. How does a fund start building that proprietary layer — especially a smaller fund without a dedicated data team?

The good news is that proprietary data doesn't need to be massive to be powerful. For a smaller fund, it starts with what you already have but aren't structuring: your portfolio company updates, your KPI spreadsheets, your qualitative notes from founder conversations. The first step is simply centralizing it — picking one system and being disciplined about using it. You don't need a data warehouse. You need a single place where every analyst, every partner, is drawing from the same source. From there, you layer in consistency: standardized KPI definitions, a regular reporting cadence, agreed-upon metrics by stage and sector. The insight you build from your own companies — their growth patterns, their failure modes, how they've responded to your support — is something PitchBook and Crunchbase can never sell you. That's yours. Off-the-shelf platforms have become commoditized; every fund can buy the same signals. The differentiation now comes from the operating intelligence and founder relationships that only you have accumulated.

Most VC firms are deeply attached to their spreadsheets — and honestly, many are very good at using them. At what point does spreadsheet-based portfolio management become a real liability, and how do you make that case to a skeptical partner?

Spreadsheets are genuinely good tools, and I want to be careful not to dismiss them. Some very smart people run tight portfolio operations in Excel. The liability isn't the spreadsheet itself — it's what happens when the portfolio grows beyond what any one person can hold in their head and manually maintain. Version control breaks down. Data goes stale between reporting cycles. An analyst leaves and takes institutional knowledge with them encoded in a file only they understood. The early-warning system simply doesn't exist — you find out there's a problem at the quarterly review, not when the first signal appeared three months earlier. To a skeptical partner, I'd skip the philosophical argument and go straight to cost. How many hours does your team spend each quarter gathering, cleaning, and reconciling data before you can actually analyze it? What did you miss in the last portfolio review because the information wasn't in front of you in time? The spreadsheet isn't the problem; the question is what it's costing you to maintain it — in time, in risk, and in decisions made on incomplete information.

You write that real-time analytics can surface issues months before they would appear in traditional reporting. What does that early-warning system look like in practice — what signals tend to matter most, and what do you do when they fire?

In practice, we focus on a concentrated set of five to six metrics per company — not thirty, not two. The goal is catching deviation from plan early: when actuals start diverging from budget, even modestly, that's often the first signal of an emerging issue rather than a one-off blip. Revenue deceleration, cash runway compression, rising churn — these are the quantitative triggers. But the system becomes meaningfully more powerful when you layer in qualitative context: notes from board meetings, a founder who seemed stressed in the last call, team turnover ticking up for two consecutive quarters. Neither the data nor the intuition alone is sufficient — it's the combination that lets you catch things before they're visible in a P&L. When a signal fires, the response is calibrated: not every alert requires a call, but every alert requires a decision about whether it does. The discipline is in the triage, not the alarm. Intervention that happens early enough can change outcomes. Intervention that happens at the quarterly review usually just documents them.

If a GP reads your white paper and wants to take one meaningful step toward data-driven portfolio management this quarter — just one — what would you tell them to do first?

Pick one source of truth and commit to it. Not a new tool necessarily — just a decision that there is one place where portfolio data lives, and that everyone on the team uses it. Most firms don't fail at data because they lack technology. They fail because data is everywhere and trusted nowhere. Before you invest in analytics, before you hire a data person, before you build a dashboard — establish that foundation. Standardize what you're tracking, agree on how you're defining it, and stop tolerating three competing versions of the same number. Everything else — the risk signals, the benchmarking, the forecasting, the LP reporting — is built on top of that. Without it, you're just doing sophisticated analysis on unreliable inputs. With it, even basic analytics start generating real insight almost immediately.