The Great Proptech Convergence

Earlier this year, we published the details of a clash between two proptech darlings: EliseAI and Funnel. Each originally operating in its own lane, they collided through mutual expansion into each other's domains: EliseAI into CRM, and Funnel into AI communication. The collision accelerated into a spat that left multifamily operators — many integrated with both platforms — alarmed.

That tiff is a sign of things to come.

The drivers are straightforward: as AI makes building software easier, the cost of adding marginal functionality drops. And as real estate operators tire of managing dozens of individual point solutions, vendors face pressure to do more in the technology stack lest they be made redundant by another provider that duplicates their functionality.

Today’s letter will explore a growing trend in proptech: every technology provider doing everything. From the property management systems to what-were-once-point-solutions, the Great Proptech Convergence promises an increasingly abundant, and chaotic, world for investors and real estate operators alike.

The Bloat Bonus

Not too long ago in the software world, duplicating another company’s functionality was a de facto declaration of war. Expanding into an adjacent vertical or use case required real effort and investment, and doing so required the new entrant to believe they could do it better than the incumbents.

But a few shifts have changed the logic.

Building software is far easier than it once was. This isn’t a new trend; frameworks and libraries have made software development far simpler over the past 15 years. But AI has supercharged development. Today,a single senior engineer with Cursor, Replit, or Claude Code can produce an order of magnitude more functionality than was possible only a few years ago. 

But for software companies, the declining cost of development has raised awkward questions about where value will be created long-term. Always a weak moat, the idea that a critical mass of features and functionality would insulate a company from disruption long-term is now out the window. 

Three years ago, I wrote a letter on the future of the property management system. In that piece, I interviewed Ender co-founder Jonathan Lonsdale, who is attempting to build a new property management system from scratch, about what it takes to do just that:

There’s probably 10,000 workflows you need to solve for, and you’re not going to get credit for the first 9,000 or so. With 6,000 [workflows], people may use it but you have a shit product. At 9,000, maybe people can like it, but it’s missing major things—core accounting features and things that would be really helpful. But at 10,000, people love it when you add five new workflows.

While Lonsdale’s statement is still true, the “mass of workflows” moat is less compelling than it once was, and it will surely be less compelling in two years than it is today.

But just as the cost of building software falls, the rewards of selling more functionality to customers have increased. 

Specifically, more functionality gives software companies access to more data. Get a unique lens into customers’ proprietary information — the leasing conversations, the financials, the maintenance logs — and you’ll have a moat that can’t easily be rebuilt in Claude Code. AI also makes the data feedback loop more legible: the more training data a company has, the better it can train its model, and the more valuable the company’s product is to its customers.

But the final driver of functionality bloat is coming from the customers themselves. Real estate operators are facing pressure to embrace new technologies just as many are feeling fatigue from dozens of disparate point solutions and tools. The average property manager has more than 20 different logins to various tools and technology platforms, many of which are expected to integrate with each other as well as the source of truth for ledgers and leases. 

For many operators, the solution is to consolidate around a smaller number of technology providers, each of which does more. A given point solution might be 20% better than its peers, but a 20% edge might not be sufficient to justify yet another tool that must be onboarded, integrated, and trained around. Vendor management tools like Revyse have made this a core part of their value proposition: ingest an operator’s full suite of tools, identify areas of overlapping functionality, and propose technology providers that could be cut as redundant. 

While consolidating real estate operators are each making their own rational decision to simplify their technology stack, technology providers are getting the message loud and clear: do enough to justify your existence or get cut in the next budgeting cycle.

The More-Than-Point Solutions

No real estate technology company has drawn more attention through product expansion than EliseAI. Initially focused on bringing AI to top-of-funnel leasing conversations, the company has since expanded to manage the full leasing process as well as most major resident-facing touchpoints including customer support, maintenance, delinquency management, and renewals. And as we covered earlier this year, Elise’s expansion into CRM ran them up against CRM incumbent and erstwhile partner Funnel, temporarily jeopardizing the firms’ integration.

The justification for Elise’s expansion is straightforward: if they are, genuinely, the best in the business at AI-mediated tenant communications, why not apply that skill to all tenant communications and not just leasing? The early results appear to support the bet: AI-mediated delinquency management is generating operating savings and improvements in bad debt. 

While Funnel may have been the most annoyed by Elise’s move into CRM, they weren’t the biggest target of the move. For many operators, PMS platforms like Yardi’s RentCafe are the obvious place for CRM to live. So the real display of EliseAI’s market power is claiming a chunk of the PMS’s traditional domain. 

But Elise isn’t the only proptech darling taking on more. Loyalty platform Bilt has expanded over the past year to be, in part, the “hospitality platform for housing,” encompassing everything resident- and operator-facing, including access control, package management, move-in/outs, renewals, and even leasing workflows like tour scheduling, lease signing, CRM, and more.

Venn, once a real estate operating company itself, has an even more expansive vision as the “operating system for residential,” describing its resident management system to real estate operators as “the structure that holds every interaction, every workflow, every detail of your business.” In practice, this means workflow management as well as a white-label resident app and communications suite, renewal management, AI-powered leasing, and (you guessed it) CRM.

While all these companies were moving in their separate lanes just a couple years ago, they’re now very much competitive. Each has its own unique color: EliseAI has the best conversational AI capabilities, including voice; Bilt has a consumer brand and more than 6.5 million members; Venn has a deep integration library and white label capabilities. 

But these three companies are just the most visible players. Almost a dozen more firms have similar ambitions and at least a handful of recognizable names on their client rosters. Funnel, for instance, boasts 9 of the top 15 multifamily operators and more than 1.5 million units on its platform. And while platforms like Livly, Amenify, SmartRent, and HappyCo haven’t drawn as much attention as their peers, each has impressive client rosters and shouldn’t be counted out as the existence of software features erode as a moat. Why shouldn’t HappyCo offer a leasing chatbot, SmartRent a resident CRM, or Funnel a work order management system?

The incumbent property management systems, however, get a vote too.

The Empire Stacks Back

When a point solution launches a new piece of functionality, the biggest competition is often not another point solution but whatever baked-in tool the clients’ PMS offers. After all, the major PMS platforms — Yardi, Entrata, AppFolio, RealPage, et cetera — consider themselves full-stack real estate management software solutions. While they may not have the best tenant screening tool or resident onboarding flow or work order management system, those tools do exist within the platforms, and many are sources of incremental revenue for PMS companies.

Through that lens, it becomes more understandable that PMS providers often see demands to “open up” their data model to integrations less as friendly pleas to play nice in the ecosystem and more as competitive threats. If a PMS makes it easier to use a third party app to screen tenants, for instance, revenue that could’ve been generated by the PMS’s own tenant screening tool erodes.

Beyond the core functionality of ledger management, there are few things property management systems do that are not replicated somewhere by a point solution. Even Bilt used its heft to break the PMS monopoly on payments, long thought impenetrable.

To their credit, the major PMS providers are not mere observers in all of this. In addition to their tried-and-true strategy of buying point solutions, the companies have been increasingly aggressive over the past year in integrating AI tools. Last September, Yardi launched a Claude MCP connector, and AppFolio announced an agent-to-agent connector just last month, allowing real estate operators to run much more complex operational work in Claude while hewing to the operator’s specific guardrails and workflows.

Of course, PMS companies still have a big secular advantage: they are the system of record for property financials, tenant ledgers, and, typically, payments. That data moat won’t go away easily, and integration with those core platforms remains just as critical as ever even if functionality moves elsewhere.

But the office sector provides a sobering counterfactual for the property management behemoths. For office operators, Yardi is little more than accounting software. Without control of rent payments — which often happen over wire or check — PMS’ providers moat has proven far less defensible. This has enabled companies like Cove and HqO to capture key parts of the customer interaction layer over the past decade with relative ease.

In multifamily, the PMS platforms have been able to largely run feature-to-feature against the breadth of point solutions, making the adoption of point solutions a question of functionality or preference rather than necessity.

This is all crazymaking for real estate operators. 

“One issue we've run into is that a lot of the companies we brought on for a specialized reason, say vendor credentialing or tenant screening, start branching out and doing other things,” explained Kelley Brine, President of Rose Valley Management, a multifamily operator with more than 10,000 units, in an interview with me for Insights by Blueprint. “Everyone starts at one point and then spreads. The problem is that it can dilute the focal point of where they started. Then you get overlap. The company we use for one thing is suddenly also a CRM, and our CRM is also something else.

“The main thing we hired them for sometimes isn't as good as it was, because they're investing in all these other avenues to diversify revenue.”

Jamie Gorski of Catchpoint Collective echoes Brine’s concern. “The question for property management firms isn’t simply, ‘How much can we consolidate?’” 

“Yes, everyone wants to simplify their tech stack. But AI is also accelerating innovation, and many of the most valuable capabilities are coming from specialized point solutions. The winning strategy is to consolidate where it makes sense, while remaining flexible enough to adopt best-in-class solutions, particularly for customer-facing experiences and advanced data and intelligence capabilities that many PMS platforms still don’t address well.”

The End of Software?

A new generation of technology entrepreneurs believe this debate is moot: AI doesn’t simply make the development of software easier, it makes it so easy that the entire concept of a software company will change.

Outcome, which we covered a few weeks ago, is one of the clearest examples: rather than bringing a pre-built software product to its customers and requiring them to adapt their data models and workflows, the company is using AI to build custom tooling around the workflows and data models that already exist. From last month’s interview with Outcome founder Prasan Kale:

Software is three things: Data, business logic, UI. We have the data, we have the business logic because of the ontology layer, we know how one field or cell affects the other, then we need to figure out the user layer, and tap the power of AI to quickly deliver what their software wants to be." Strip a SaaS product to its parts, and the UI, the thing customers actually pay a per-seat fee to touch, turns out to be the least durable layer. The data and the logic are the assets, and the interface is merely the container.

This is simply point solution feature bloat taken to its logical conclusion. If it’s increasingly inexpensive for a software provider to add an additional feature to serve a marginal need or use case, why not just build software per-client as needed? 

And Outcome is not the only newcomer taking this angle; former NestPick founder Omer Kucukdere is launching his own platform, Runy, later this month. “We’re taking a Palantir-like approach with a digital twin orchestration system, running companies' own workflows and AI agents,” Kucukdere explains.

The rise of companies like Outcome and Runy raises another interesting question: in a world where AI can build any custom software for any operator at any point, is there any value to be found in the technology stack at all? 

The emerging consensus among technology providers is that value can still be created through access to data at scale. That data, after all, can be used to train proprietary models on everything from tenant communications to interior design to deal intelligence. In theory, this proprietary training data could give vertical AI companies a moat — at least temporarily — against the large foundation models.

But real estate operators are beginning to wise up to this and set their data policies accordingly. Multifamily operator AvalonBay, for instance, recently mandated that software providers silo their data, only using it to train AvalonBay’s version of the technology. So there’s no enduring defensible advantage for companies serving AvalonBay, and it would surprise me if similar restrictions didn’t gain popularity among operators in the coming years. 

Everything Is Title Insurance

Sure, Outcome has a defensible advantage today. They’ve built replicable workflows and an engine that can deploy AI to real estate operators at scale, and fast. But given the speed at which AI is moving, I’m skeptical we’ll see a scenario in which AI moves aggressively enough to erode software incumbents’ advantage but not so ambitiously that it doesn’t also erode Outcome’s. Either nobody is cooked or we’re all cooked.

In that world, a software firm’s existing codebase or data library matters little-to-none. All that matters is the individual implementer’s ability to use commodity tools to build me exactly what I want when I want it. Prices fall into a fairly narrow range, and purchasing decisions are largely made based on reputation and relationship.

Software would work a bit like title insurance.

One could imagine a favorite technology expert or brand hired to build them custom software designed specifically for their workflows, their portfolio, their data. Everything is spec built, so trusting the individual or firm implementing the “software” is paramount, leading to the emergence of brands — perhaps ones that have little to do with software today. It doesn’t take too much to imagine someone like Jamie Gorski herself, who clearly has the ear and trust of multifamily operators, choosing to become the implementer-of-choice of the leasing tech stack. While “implementation” has traditionally meant customizing and tying third-party tools together, AI could enable the implementation layer to go far deeper into the technology itself.

Not every proptech tool is equally vulnerable. AEC tools like LightTable, TestFit, and OpenSpace are building harder things that can’t easily be vibecoded or replicated by a foundation model. 

Perhaps that's the new bar for companies of all types. At Thesis Driven, we have to ask whether the insights in any letter we write go beyond what the reader could get from an LLM. That bar keeps rising.

Increasingly, every software firm will face the same question: if an LLM could build this, why are we?

–Brad Hargreaves