The Commoditisation of Alpha: Why AI Makes Primary Research More Valuable, Not Less
AI has democratised quantitative analysis to the point where everyone has access to the same models, the same datasets, and increasingly similar conclusions.
I've watched the investment landscape transform over the past decade. AI has democratised quantitative analysis to the point where everyone has access to the same models, the same datasets, and increasingly similar conclusions.
The edge isn't in the algorithm anymore.
68% of hedge funds now employ AI for market analysis and trading strategies. The AI in asset management market was valued at USD 3.4 billion in 2024 and is projected to grow at a CAGR of 24.2% through 2034. This isn't experimental anymore. This is table stakes.
But here's what the data also shows: only 26% of companies have developed the capabilities to move beyond proofs of concept and generate tangible value from AI. Two-thirds of firms report only small or moderate returns on their AI investments.
The gap between AI hype and AI value is widening.
Public Data is Becoming Worthless
Data creation has increased by a factor of 100 over the past 15 years. Raw intelligence is commoditising fast. Frontier labs release models that are "good enough" for most tasks, open-source alternatives catch up within months, and inference prices collapse.
Everyone is looking at the same information.
As alternative data goes through commoditisation, it becomes traditional data. The moment a dataset achieves broad distribution, its alpha decays. 98% of investment professionals agree that alternative data is becoming increasingly important to identify innovative ideas, yet many datasets that deliver useable information at the outset don't maintain their value after months or years.
The signals erode. The edge disappears.
Proprietary, non-public information is the new moat. Not publicly available datasets that everyone can access.
AI Shows You "What" — Primary Research Tells You "Why"
AI excels at pattern recognition. It can process massive datasets at speeds that make human analysis look glacial. It can tell you consumer sentiment is declining, foot traffic is dropping, or inventory is building.
But AI cannot explain why.
62% of asset management firms identify the absence of clear regulatory guidelines as a top challenge in AI adoption. Incorporating AI into workflows introduces ethical concerns like algorithmic bias and overreliance on automation, which leads to reduced accountability.
You need qualitative reasoning to understand causation, context, and the human behaviour driving the numbers. That comes from channel checks, customer surveys, and direct conversations with people inside the market ecosystem.
Primary research validates AI-generated hypotheses. It acts as a fail-safe against the limits and risks of relying solely on algorithms for investment decisions.
The Expert Network Model is Broken
The expert network industry grew just 1% in 2023 to $2.28 billion. Growth is slowing despite increased demand for specialised insights.
The structural problems are mounting.
You're paying $1,200 per call. 40% of those calls are useless. The middleman takes 50-70% margin. Experts are recycled across databases, so you're often speaking to the same people your competitors already consulted.
Poor quality data costs the US $3 trillion each year. The cost isn't just financial. It's time. Analysts waste roughly 14 hours a month vetting experts, scheduling calls, interviewing, taking notes, and chasing transcripts.
You're also carrying compliance risk. As expert networks become more ubiquitous, the regulatory landscape grows more complex. In 2024, the SEC penalised a firm for inadequate controls around material non-public information (MNPI).
Compliance risk is no longer theoretical. It's actively being enforced.
Finished Intelligence Beats Raw Access
The winning model isn't about access anymore. It's about outcomes.
Modern primary research platforms manage the entire process: survey design, expert recruitment, data collection, verification, and analysis. You submit a 10-minute brief and receive decision-ready intelligence that you can drop straight into IC memos.
No scheduling. No calls. No note-taking. No fraud checks. No data cleaning.
AI leaders achieve 1.5 times higher revenue growth, 1.6 times greater shareholder returns, and 1.4 times higher returns on invested capital. But these leaders follow a rule: 10% of resources into algorithms, 20% into technology and data, and 70% into people and processes.
The moat isn't the model. It's the closed-loop human feedback system.
The Future is a Feedback Loop
AI develops hypotheses. Human analysts validate them using primary research. The insights feed back into the model. The loop compounds.
This is where competitive advantage comes from now.
As AI automates basic analytical functions, the value of human analysts shifts to asking incisive questions and designing efficient primary research projects. You explore human behaviour, competitor strategies, and market dynamics that algorithms can't parse from public datasets.
The need for finished intelligence—vetted, contextualised insights—will increase, not decrease. Platforms that bridge the gap between raw information and actionable intelligence become essential infrastructure.
The question isn't whether you need primary research in an AI-driven world.
The question is whether you're still paying middlemen for access when you should be buying verified answers.
