By Dwi Wanna 
By [Your Name/Editorial Team], based on insights from Henrik Landgren
Venture capital has long been defined by a duality: it is 50% rigorous science and 50% intuitive art. For decades, the industry has relied on the charisma of founders—the elusive "it" factor that convinces a partner to write a seven-figure check. While this human element remains an essential component of backing the next generation of industry-defining entrepreneurs, a growing movement suggests that the traditional VC model is being undermined by a reliance on fragmented, curated, and ultimately opaque data.
As artificial intelligence becomes the industry’s favorite buzzword, a dangerous trend has emerged: VCs are using powerful AI to speed up the processing of the wrong data. In an era where pitch decks are "vibe-coded" in hours and financial models are polished to perfection by LLMs, investors are increasingly flying blind. The result is a system that creates the illusion of diligence while masking the reality of information asymmetry.
Main Facts: The Crisis of Asymmetry
The fundamental flaw in the modern venture capital model lies in the "curation gap." When an investor identifies a promising startup, they are handed a collection of data points specifically selected and packaged by the founder. This is not necessarily malicious, but it is inherently biased.
In the current landscape, the barrier to entry for crafting a compelling, data-heavy narrative has collapsed. Founders can utilize AI to synthesize growth metrics and project hockey-stick trajectories that look pristine on paper. For the investor, the challenge is no longer finding information—it is discerning which information is authentic.
Henrik Landgren, co-founder and CPTO at Gilion and former VP of analytics at Spotify, argues that the current approach to AI in finance is, to put it charitably, misguided. "An AI stack is not a checkbox to be ticked off," Landgren notes. "Using an LLM to cut down the time needed to put together reports or summarize pitches is instantly gratifying, but it seldom changes the efficiency of your work in the big picture."
Chronology: From Spotify’s Data Revolution to the VC Stagnation
To understand where the venture industry went wrong, one must look at the evolution of corporate analytics.
- The Early 2010s (The Data Awakening): At companies like Spotify, the shift from static Excel spreadsheets to granular, real-time data tracking revolutionized decision-making. By capturing every user interaction and click, firms began to understand the mechanics of their business in unprecedented detail.
- The Mid-2010s (The Information Age): As data became more accessible, venture capital firms began to hire larger analyst teams to process incoming information. However, the reliance remained on "requested" data—reports provided by founders during the diligence phase.
- The 2020s (The AI Hype Cycle): With the explosion of Large Language Models (LLMs), the industry scrambled to implement AI. Instead of using technology to source raw, objective data, firms used it to summarize the subjective, pre-processed documents provided by startups.
- The Present Day: We have reached an inflection point. The industry is now trapped in a "garbage in, garbage out" cycle, where AI is merely accelerating the speed at which investors consume biased information.
Supporting Data: Why "Vibe-Coding" Diligence Isn’t Enough
The core of the problem is that investors are often reviewing the story of a company rather than the pulse of it. In a professional house-buying scenario, an investor would never accept a property survey conducted solely by the seller. Yet, in venture capital, this is the industry standard.
The True Source of Truth
Investors need to pivot toward "direct-to-source" data. This means bypassing the glossy slide deck and tapping into:
- Payment Records: Real-time visibility into cash flow and revenue cycles.
- Marketing Performance: Raw data from advertising platforms to assess true customer acquisition costs (CAC).
- Accounting Systems: Direct access to ledgers to understand burn rates without the "smoothing" effect of management accounting.
- Board Reports: Historical internal documentation that reveals the actual trajectory of the firm, rather than the aspirational path presented in a pitch.
By shifting to this model, an analyst can begin their work at a 70% level of clarity. This leaves the human element—the "art" of assessing the team and the "it" factor—to the domain where it actually belongs: human intuition.
Official Responses and Industry Sentiment
The venture community is beginning to feel the pressure of this transition. While traditionalists argue that personal relationships and "gut feeling" remain the gold standard, a growing cohort of data-native investors—led by firms like Gilion—are pushing for a radical shift in infrastructure.
The sentiment is shifting from "how can we use AI to work faster?" to "how can we use technology to see more clearly?" There is a growing recognition that in five years, the startups of tomorrow—focused on AI-driven hardware, deep tech, and complex infrastructure—will not fit into the traditional income models of the past. Historical indicators of success, such as simple revenue growth, will become insufficient metrics for evaluating companies that operate on entirely new technological paradigms.
Implications: The Future of Investment
The implications of failing to modernize are severe. Firms that continue to rely on manual, founder-provided data will find themselves at a structural disadvantage for three reasons:
1. The Speed of Conviction
In highly competitive deals, the advantage goes to the investor who reaches conviction first. If you spend weeks cleaning up, verifying, and questioning a founder’s data, you have already lost the deal. Better infrastructure is not just a "nice-to-have"—it is the primary competitive moat for the modern firm.
2. The "Forgotten" Businesses
There are thousands of capital-efficient, high-retention companies that are passed over by VCs simply because their growth doesn’t look like the "hyper-scale" models of the past. Many of these businesses are highly viable, but they lack the polish or the "vibe" to attract standard VC interest. With better data access, these overlooked gems can be objectively evaluated and funded, broadening the scope of the venture ecosystem.
3. The End of Blind Betting
The industry must move away from the "spray and pray" model, which relies on the hope that one massive outlier will cover the mistakes of poor due diligence. By building infrastructure that allows for transparent, objective oversight, investors can move from making "blind bets" to making calculated, data-backed decisions.
Conclusion: A Call for Infrastructure
The venture capital industry is at a crossroads. It can either continue to use the latest, most powerful AI tools to polish the same flawed, subjective data, or it can fundamentally rethink its relationship with company information.
The future belongs to those who view data infrastructure as the precondition for success. The "it" factor in a founder will always be important, but in the future, it will be validated by cold, hard, and—above all—unfiltered data. Anything less is simply speed-running the process of failure.
As the landscape of technology shifts beneath our feet, the firms that win will be the ones that stop asking how to speed up their current process and start asking what a superior process should look like. The era of the "vibe check" is coming to an end; the era of the data-validated visionary has begun.
Jia Lissa 
Siti Muinah 
Nana 