In today’s evolving AI market, capital alone is no longer a differentiator. The real edge comes from rigorous, evidence-based decision-making. At Activant, we believe a research-led approach is essential for identifying companies that can endure beyond market cycles. That means going beyond surface metrics to deeply understand:

Relationship Capital → the depth and quality of customer integration

Model Stack Strategy → ownership of key layers in the technology stack with genuine domain expertise

Distribution Ownership → embedding into critical workflows that create long-term resilience

This isn’t about predicting winners from a pitch deck, it’s about building conviction through data, diligence, and a clear view of market dynamics. In this final installment, we explore why deep research into markets, moats, and model strategy has become the edge in both building and investing in the next generation of enduring AI companies, and how applying these principles can help avoid the traps of hype and short-term thinking.

This concludes our 5-part series on navigating AI market consolidation through a disciplined, research-first lens. Catch the full series on our website:

I. The End of the AI Gold Rush? – when hype gives way to discipline

II. Does the House Always Win? – working with and around tech giants’ ecosystems

III. The Great Filtration – market realignment and the shakeout ahead

IV. The Durable Moat – building long-term defensibility

V. The New Mandate – research-driven investing for lasting impact

Thank you for joining us on this 5-part journey. Our goal was to bring clarity to what we believe drives staying power in a consolidating market and insights we hope will be valuable whether you’re deploying capital or building the next enduring AI company.

The narrative of inevitable platform-led consolidation now faces a significant and growing counterforce: regulatory scrutiny. While the tech giants are currently winning the market battle, they could face growing headwinds from regulatory action. Antitrust authorities in both the United States and Europe are investigating these novel consolidation tactics and are launching inquiries that could fundamentally reshape the competitive landscape.

In January 2024, the U.S. Federal Trade Commission (FTC) opened an investigation into investments and partnerships between major cloud providers and leading AI companies, issuing compulsory orders to Alphabet, Amazon, Microsoft, Anthropic, and OpenAI. This scrutiny is not limited to the U.S. In April 2024, the U.K.'s Competition and Markets Authority (CMA) announced preliminary inquiries into the partnerships between Microsoft and Inflection AI, Amazon and Anthropic which have since concluded, with both cleared. The agency specifically examined whether the Microsoft/Inflection deal constituted a "relevant merger situation," indicating a willingness to challenge the asset-light structure of these “acqui-hires”. This escalating regulatory pressure is a material regulatory counterweight to the consolidation thesis.

While founders and investors wait to see if regulators will reshape the external landscape, they must also address the urgent need to fix their internal models. The giants' strategic dominance has not only changed the market, but it has also broken the traditional investment playbook.

The last 18 months have set a trap for many investors. Intense market hype and low barriers to entry have unleashed a flood of AI startups, many delivering applications built largely on public APIs with limited proprietary differentiation. While this approach enables rapid product launches, it often leaves companies vulnerable to replication by larger platforms, making defensibility and long-term value creation much harder without unique data, technology, or domain expertise.

In many ways, “.ai” has become the new “.com.” A slick demo can be the modern equivalent of a flashy Super Bowl ad, attention-grabbing, but not necessarily supported by meaningful revenue or a sustainable business model. This dynamic has created a dangerous illusion of progress, though there are, of course, notable exceptions where teams are building truly differentiated and defensible platforms.

In this environment, where some investors have adopted a broad “spray and pray” approach across portfolios of quickly assembled, API-dependent products, the risk of underperformance in a consolidating market is high.

Many elements of the VC playbook from the mobile and SaaS booms no longer translate effectively to today’s AI market. In that era investors chased signals like user engagement metrics and top-line growth; but in the AI era, these indicators can be dangerously misleading.

This is where the strategy of pattern-matching fails. A VC might see a product with a low customer acquisition cost (CAC) and rapid initial adoption, believing they've discovered the next unicorn. More often, they've likely found a compelling feature, not a durable company. Tech giants can replicate and distribute this feature for free to millions of users, causing the startup's long-term value (LTV) to erode and undermining its business model. Without a research-led approach, the traditional venture model is at high risk of underperformance in this environment.

A research-led investment thesis is not just an advantage, it’s a prerequisite. It acts as a filter against the hype, replacing gut instinct with a rigorous, first-principles analysis of the market's new physics. It demands a deep understanding of the entire technology stack to distinguish between a temporary feature and a durable platform.

Instead of chasing superficial signals, a research-led doctrine focuses on the foundational pillars of defensibility outlined in Part IV: Relationship Capital, a coherent Model Stack Strategy, and a path to Distribution Ownership.

For those who haven’t read Part IV, here’s a quick recap:

Relationship Capital → Building trust and integration with customers that creates a compounding data advantage over time.

Key question: Does the company have a genuine data flywheel that generates proprietary, compounding advantage with every user interaction, or is it simply passing data through a third-party API?

Model Stack Strategy → Owning the right layers of the technology stack, often with deep domain expertise and proprietary datasets.

Key question: Has the company gone “vertical,” building deep domain expertise and proprietary datasets for a specific industry that horizontal platforms cannot easily replicate?

Distribution Ownership → Embedding into critical customer workflows to create high switching costs and genuine integration lock-in.

Key question: Is the company embedded so deeply into critical workflows that replacing it would be disruptive and costly for the customer?

This framework is the direct answer to the challenges we detailed earlier in this series. The "gravity wells" (large platform ecosystems that draw customers inward) of the tech giants? They can be countered by Distribution Ownership. The "wrapper Challenge"? It's mitigated by a real Model Stack Strategy. The threat of the "acqui-hire"? It is mitigated by building sticky Relationship Capital that makes a team and its product inseparable.

Answering these questions takes more than a 30-minute pitch. It requires deep technical diligence and a thorough grasp of industry-specific workflows. It's about pre-emptively answering the critical question that public-facing product roadmaps make relevant: "Why can't this be a feature in Microsoft Copilot or a Google Workspace update next quarter?" In a market defined by the “gravity wells” of giants, this disciplined, research-led approach is the most reliable path we’ve found for avoiding the traps of the past and identifying the truly durable companies of the future.

Capital itself is a commodity; actionable insight, forged through deep and rigorous research, is the only scarce resource. This new reality where the "acqui-hire" phenomenon has exposed that valuation does not always mean value, makes a disciplined approach essential. The road ahead is likely to be more challenging. We expect to see a wave of M&A—mostly “acqui-hires”, and a string of companies’ wind-downs. But we will also see the emergence of a new class of resilient, market-defining companies. The central question is no longer just "What can you build?" but "What can you build that lasts, and what evidence can you show today that it will?"

Click here to read the full series on our website 🔗