Niches Don't Matter Anymore...Kinda

I've flipped over 150+ deals in the past 14 years across three completely different real estate verticals, and I was never the expert in any of them.

Most of them I sold. Some I kept.

Then I got tired of watching other short-term rental managers fumble managing my Airbnbs, so I partnered with my friend Isaac Moore and built a short-term rental management company to run them properly. Still alive today, called Buy More Time.

Then I started buying hotels and converting them into multifamily.

Three different verticals. Three different operating models. Three different capital stacks.

None of them required me to be the smartest person in the room. They required me to find the smartest person in the room and partner with them.

I had very few problems raising money for any of those pivots, as long as I was partnered with a real expert in the niche. If my partner was a subject matter expert who has gone "full-cycle" on the deals I'm raising for, the capital eventually showed up.

That was nearly a decade ago, with a fraction of the tools available today.

If it was possible then, what do you think is possible now?

TL;DR: Niches Still Matter. The Niche Operator Just Doesn't Have to Be You.

The top 10 real estate funds raised 40% of all institutional commitments in 2025. The top 15 managers control 45% of private real estate AUM. Sector specialists are getting absorbed in M&A waves they cannot fight. Meanwhile, AI is compressing the cost of cross-sector underwriting from quarters to weeks. CBRE reports 3x faster preliminary analysis. Goldman Sachs estimates 20-35% reductions in due diligence costs. JLL cut lease abstraction labor by 60% and recovered $1M+ in missed escalations. The PE managers winning the next decade will not be those with the deepest single-asset-class expertise. They will be those running the formula: AI fluency plus a full-cycle subject matter expert operator, in every vertical they want to play in.

I. The Pressure on Single-Sector Sponsors

The data on capital flows is stark, and it has gotten worse for specialists in 2025 and into 2026.

According to research firm With Intelligence, the top ten real estate funds raised $68 billion in 2025, capturing 40 percent of total investor commitments. The top fifteen managers now control 45 percent of all private real estate assets. Primior Group's 2026 analysis pushes the figure higher still: the ten largest funds captured 53 percent of strategy allocations last year, up from 33 percent the year prior.

The Catalyst Group, in an April 2026 fundraising review, observed that more than half of real estate funds closed below their target size in 2025, that fundraising timelines stretched to record lengths, and that LPs are consolidating relationships rather than expanding them.

This is not a cyclical adjustment. It is a structural rotation in how institutional capital is being deployed.

With Intelligence makes the consequence explicit. *"Sector, geography, and strategy specialists have been absorbed by larger private equity groups, with a significant wave of M&A over the past two years accelerating consolidation."* The niche shop is increasingly the acquisition target rather than the platform.

The signal at the top of the market reinforces the same pattern. In April 2026, Ares Management announced the final close of U.S. Real Estate Fund XI at $5.4 billion, the largest private equity-backed real estate vehicle in history, alongside its European Real Estate Fund IV. The mandate is explicitly multi-sector: industrial, multifamily, and office-to-residential conversion. Blackstone separately filed for an IPO of its Digital Infrastructure Trust, targeting a $2 billion raise to acquire stabilized data centers leased to investment-grade hyperscalers.

The largest pools of capital in the industry are explicitly building for cross-sector flexibility. The middle of the market is being asked to follow.

Lender behavior reinforces the trend. The Mortgage Bankers Association is projecting $806 billion in commercial mortgage origination for 2026, up from $633.7 billion in 2025. That liquidity is meaningful, but it favors sponsors who can identify relative value across asset classes rather than defend a single sector regardless of where spreads have moved, and it is landing into the same maturity wall and tenant story we walked through in the SaaSpocalypse piece.

None of this eliminates the single-sector sponsor. Many will continue to perform well in defensible niches with durable demand. But the trajectory is clear. Specialization without optionality has become a structurally harder game.

II. Where Institutional Capital Is Actually Going

The shift in capital concentration is paired with an equally important shift in *which* sectors are receiving that concentrated capital.

Cohen & Steers, in February 2026 research, observed that "next-generation" real estate sectors, defined as data centers, industrial, senior housing, self-storage, manufactured homes, single-family rental, and cell tower REITs, now represent more than half of the public REIT investing universe. Twenty years ago, those sectors were small slices of the index.

The PwC and Urban Land Institute *Emerging Trends in Real Estate 2026* report, drawing on input from more than 1,700 real estate investors, developers, and lenders, names data centers, senior housing, and self-storage as the highest-conviction property types for the year ahead. Self-storage holdings in the largest U.S. core funds now exceed hotel market value. Medical office demand has become "more durable than that for traditional office." Student housing has matured into a "liquid, mature rental housing subsector."

Colliers' 2026 Global Investor Outlook reports that data centers alone drew 31 percent of all funds raised in the first three quarters of 2025. Appetite for alternatives, including student housing, self-storage, and healthcare buildings, is rising independently. Nearly half of surveyed investors (49 percent) now favor direct investments and separate accounts over traditional commingled funds, indicating a desire for greater control over sector allocation.

A more specialized example illustrates how fast these niches institutionalize. CRE Daily reported in late 2025 that approximately $300 billion of industrial outdoor storage real estate is now considered institutionally investable, with Blackstone, J.P. Morgan, and Blue Owl Capital actively entering the space. IOS rents have grown 123 percent since 2020, and vacancy rates run roughly half those of bulk warehouse space. Five years ago, IOS was the domain of local business owners and a handful of niche sponsors. Today it is a competitive, institutional asset class, and the same pattern is reshaping the trophy-versus-boring conversation we wrote about earlier this year.

The point is not that any one of these sectors is the right place to be. The point is that the menu of where serious capital is being deployed has expanded dramatically, and a manager whose mandate covers only one item on that menu is functionally unable to compete for the other thirty.

III. The Architecture That Worked

Across each of my own pivots, the structure was the same.

I identified a subject matter expert with full-cycle experience in the target niche. I built the corporate structure, the capital partnerships, and the institutional positioning around them. I sourced the deal flow, raised the equity, and stood behind the strategy with investors. The SME contributed the operational reps that no amount of research can substitute for. (If you are evaluating that partnership architecture for your own raise, book a strategy session and we can walk through the same diligence I run before signing on with a new SME.)

That division of labor isn't new. The sponsor-operator split has existed in institutional real estate for as long as the asset class has been institutional. What has changed is the *cost of assembly*, in time, capital, and information.

A decade ago, identifying the right SME required six months of conferences, bad meetings, and warm introductions through a network you had to cultivate over years. Underwriting an unfamiliar asset class required a ninety-day diligence sprint and a paid consultant. Sourcing the first deal required another year of relationship building.

That sequence is now compressing into weeks. The reason is the third party in the new architecture: AI.

IV. What the Workflow Actually Looks Like

Consider a multifamily sponsor presented with an offering memorandum for a 280,000 square foot industrial outdoor storage portfolio in Dallas. Three properties. Asking $42 million.

A decade ago, the rational response was to pass. No team, no comparables, no underwriting infrastructure, no relationships in the asset class. The optionality was effectively zero.

Today, a single business day looks different.

> The 6-Hour Cross-Sector Underwrite > > Hour 1, Morning: Sponsor uploads the OM and queries an AI workspace for an IOS market overview of Dallas, cap rate ranges by submarket, the major operators, recent transaction comps, operational risk factors that distinguish IOS from traditional industrial. The output is roughly what a junior analyst would have produced in two weeks. > > Hour 2, Midday: A sensitivity analysis runs the deal at three rent-growth scenarios against the asking price, identifying the assumptions that would need to hold for the investment to clear a 14% IRR. > > Hour 3, Afternoon: The sponsor identifies the most active IOS operators in Texas, ranks them by transaction volume over the prior three years, and surfaces what is publicly knowable about their joint-venture preferences and structure terms. > > Hour 4, Close of Business: Outreach goes out to the top three potential operating partners. Not generic. Asset-class-fluent, signaling diligence and capital readiness.

The sponsor has not bought the deal. They have not signed an LOI. But they have moved from "out of mandate" to "in evaluation" in approximately six hours.

V. What the Productivity Data Says

The compression is measurable, and the firms with the best data on it are now publishing those numbers.

CBRE's 2025 Tech Adoption Report found that development teams using AI for underwriting completed preliminary analysis three times faster than those that did not. Goldman Sachs, in mid-2025 research, estimated that AI tools could reduce due diligence costs for large institutional portfolios by 20 to 35 percent.

JLL has reported its own internal numbers. After implementing AI-powered lease abstraction across its global business, JLL reduced manual review labor by 60 percent and uncovered more than $1 million in missed escalation clauses across its existing portfolio. Its teams now handle three times the lease volume without additional headcount.

McKinsey has estimated that AI could generate $110 to $180 billion in value across the real estate sector through productivity gains and improved decision-making.

CBRE has separately deployed AI-enabled facilities management across more than 20,000 sites covering one billion square feet of managed real estate, demonstrating that enterprise-scale implementation is now operationally proven, not theoretical.

The real estate technology adoption picture is also expanding from the bottom up. JLL's Global Real Estate Technology Survey shows that of approximately 7,000 global PropTech companies tracked at year-end 2024, roughly 700 were already shipping AI-powered solutions, with 62 percent of those VC-backed. The pipeline of AI-native real estate tools that will reach institutional managers over the next 24 months is large.

Bisnow has framed the trend in industry terms: AI is now being deployed across *"valuation, underwriting, and due diligence, areas long dominated by manual analysis and spreadsheet-based models."*

VI. The Trust Gap Is Real, and It Reinforces the Thesis

A meaningful caveat is warranted, and it does not weaken the central argument so much as sharpen it.

The Center for Real Estate Technology and Innovation, in research published in early 2026, found that investment committees continue to distrust AI-generated analysis for high-stakes financial decisions. AI is broadly accepted for efficiency tasks. It is not yet accepted as the final word on a forty-million-dollar acquisition.

Deloitte's 2026 Commercial Real Estate Outlook found that despite headline enthusiasm, more than 60 percent of CRE firms still rely on legacy technology infrastructure, and respondents to its global survey of 850 C-suite real estate executives offered "more grounded" responses about AI's transformational power than they did the year prior.

The Commercial Observer summarizes the dynamic as well as anyone. *"Trust, not technology, is the gating factor. Despite persistent narratives around automation-driven job displacement, the data suggests a far more pragmatic reality. Most firms report using AI to improve efficiency rather than reduce headcount."*

That skepticism is appropriate. AI compresses the analytical work. It does not replace the judgment, the relationships, or the fiduciary discipline. As Robb Gilman, an accounting and audit partner at Anchin, put it bluntly to Bisnow: *"I wouldn't give AI $20M to invest."*

He is right. No competent investment committee would.

That is exactly why the SME partner matters more, not less, in the current environment. AI compresses the cost of becoming dangerous in a new asset class. It does not eliminate the need for a partner who has lived through a full cycle in that asset class. The two functions are complementary, not substitutable.

The mistake the next decade's losing managers will make is treating AI as a replacement for the operator, rather than as the thing that makes the operator partnership work in five new sectors instead of one.

VII. The Operating Model

The PE and asset management groups likely to outperform over the next decade will not be those with the deepest single-asset-class expertise. They will be those that have institutionalized a repeatable formula:

> AI plus a subject matter expert operator. In every vertical the firm intends to underwrite.

The required steps are straightforward.

First, develop fluency in a new vertical quickly enough to evaluate whether the opportunity is real. AI compresses this from quarters to weeks.

Second, identify and evaluate the right SME operator to partner with. AI compresses this from months to days.

Third, contribute the capital structure, the strategic conviction, and the LP relationships. This remains a relationship business, and AI does not change that.

Fourth, allow the SME to execute the business plan. The operator runs the operator's job. The capital allocator runs the capital allocator's job. AI does not collapse those roles. It eliminates the friction between them.

That is the architecture I executed across three businesses using technology that was meaningfully less capable than what is available today. The deals closed. The capital came. The investors were made whole. Not because I was a deeper operator than the partners I built around. Because the structure was correct.

What used to take six months can now happen in two or three weeks.

VIII. The Counterargument

The most credible counterargument to this thesis is not that AI fails to compress cross-sector underwriting. The data is clear that it does. The credible counterargument is that LP demand will continue to favor specialists with verifiable track records, regardless of how cheap optionality becomes.

There is something to that. PwC's *Emerging Trends* report notes that institutional investors have *"demonstrated an intense preference for managers with scale and proven track records."* Capital is concentrating, and the managers with single-sector ten-year track records are not going to be displaced by an AI-augmented generalist with a spreadsheet.

The argument here is not that the specialist disappears. It is that the *capital allocation pyramid* changes shape. The top of the pyramid (a small number of mega-managers running multi-strategy mandates) absorbs more share, as the data shows. The middle of the pyramid (single-sector specialists with $250M to $2B AUM) compresses. The new growth tier emerges in the orchestrator model: managers with deep relationships, AI-augmented underwriting, and a stable of full-cycle SME operating partners across multiple verticals.

That tier did not exist a decade ago because the assembly cost was prohibitive. It exists now because AI has lowered that cost by an order of magnitude.

A second counterargument worth taking seriously: AI-augmented underwriting may actually *increase* deal competition, compressing returns. If everyone can get smart on cold storage in two weeks, the alpha in cold storage gets bid away. That is also true, and it is part of why the SME partnership is the durable component of the model. The AI fluency is replicable. The partnership with a particular full-cycle operator is not.

IX. Conclusion

For three decades, the industry has rewarded specialists.

The next decade is likely to reward orchestrators: managers capable of pairing AI fluency with full-cycle subject matter experts across multiple verticals.

The competitive question is no longer whether a manager has the deepest expertise in their chosen niche. It is whether they have the operating model to credibly enter new ones at the speed the market now demands.

Niches still matter. The operating expertise is real, the relationships are real, and the cycle scars are real. What no longer matters is whether the capital allocator and the niche operator are the same person, or even the same firm.

That is a meaningfully different game than the one most managers have been playing.

Want a deeper dive on how to structure a multi-sector raise around an SME operator? Book a strategy session here.

See you in the arena.