Acquisitions Teams Want AI for Sourcing, Not Underwriting: The Demand Data | NextAutomation

If you run acquisitions at a CRE firm and you are deciding where to point AI first, the default assumption in the market is that you should point it at underwriting. Speed up the pro-forma, screen more deals, get to a number faster. It is a reasonable assumption. It is also not what your peers are actually reaching for first.

In our own first-party demand, the clearest signal is that buy-side operators want AI for sourcing before underwriting. They want it to help them find the deal and find the owner before the deal is competitive. Underwriting automation is real demand, but in our data it is the second thing operators reach for, not the first. This piece lays out exactly what we measured, how to read those numbers honestly, why sourcing wins the priority fight, and when underwriting AI still deserves to go first at your firm.

This is one cut from a larger read. If you want the whole picture, the full state of AI in commercial real estate for 2026 covers the demand census, the stuck-points, and the industry context. Here we go deep on the single most counterintuitive finding in it.

The Finding: Buy-Side Demand Runs Roughly 4.5 to 1 Over Brokers

Start with who is asking. When we sort demand for our free CRE-AI templates by the kind of operator raising their hand, it skews hard toward the buy side. Investor and acquisitions sessions outnumbered broker sessions roughly 4.5 to 1: 4,853 distinct sessions versus 1,066. Developers registered 311 sessions. The institutional and LP bucket showed 48, which is real but tiny, and we read that as under-distribution to that audience rather than proven under-demand.

Two honest notes travel with that ratio, because leaving them out would make it a worse signal, not a better one. First, we publish the ratio, not percentage shares. These buckets are topic-affinity groupings, not a clean partition, so a single operator can appear in more than one bucket. Shares would imply a precision we do not have; the ratio is the load-bearing, honest figure. Second, this is a LinkedIn-audience-weighted census. A little under 60% of the underlying views came through LinkedIn distribution, so this reflects the operators in that audience, not a random sample of the whole industry. Read inside those limits, the finding is durable: buy-side operators are reaching for practical CRE-AI at several times the rate of the brokerage side.

That matters for benchmarking your own priorities, because it tells you who your competitive peers are on this. If you are on the acquisitions side, the operators most actively arming themselves with AI are the ones sitting across the table from you on the next deal, not the brokers marketing it. The interesting question is what, specifically, they are arming themselves to do. That is where the surprise is.

Sourcing Beats Underwriting in Real Calls, 13 to 6

Demand for a template tells you what an operator will click on. A recorded conversation tells you what they will actually spend an hour describing to you. So the sharper signal comes from the calls, and in the calls the priority is unambiguous.

Across our recorded operator conversations, off-market sourcing and owner-finding came up in 13 distinct real calls. Underwriting, pro-forma, and deal-analysis automation came up in 6. When acquisitions people tell us what they want AI for, more than twice as many of them name finding the deal as name modeling it.

A word on how to read those two numbers, because the discipline behind them is the reason they are worth trusting. They are floors, not estimates: each is the count of distinct conversations where the theme was explicit, not a guess at how many people quietly care about each thing. And they come from a single corpus, our recorded calls, which we keep separate from every other data source. We do not sum sourcing demand from the calls with sourcing demand from a chat log to manufacture a bigger, cleaner-looking total. Thirteen against six is a like-for-like comparison inside one body of conversations, which is exactly why the gap means something.

There is a second corpus that points the same direction from a different angle. In 150 messages that operators typed into our website chat, off-market and deal-sourcing intent was the third-largest theme, ahead of build-versus-buy questions, reliability worries, and data-security concerns. Two independent bodies of first-party language, recorded calls and typed questions, and in both of them sourcing sits at or near the top of what operators are actively trying to solve. We walk through the mechanics of that problem in how to find off-market properties.

The Off-Market Surge Is the Fastest-Moving Demand We Have Measured

It is not just that sourcing leads. It is that sourcing is accelerating past everything else in the library right now. Our off-market operating system is the hottest-velocity asset we have ever shipped, running at 18.4 sessions per day across 184 sessions, 100% of them in the last 30 days. And the off-market persona bucket went from 0 to 182 sessions in July alone.

One honesty guardrail on that, because it cuts against a claim we are deliberately not making. The overall demand for our library is not a clean up-and-to-the-right growth curve; it is a distribution spike followed by a long tail, which is what content that travels looks like, not organic month-over-month adoption. So we do not claim "AI adoption is accelerating" from the census as a whole. What we can say precisely is narrower and more useful: inside the library, at this moment, off-market sourcing is the vector pulling the most fresh demand. It is where operator attention is concentrating right now, broken out by asset type and region in our off-market deal-flow index for Q3 2026.

Put the three cuts together and the operator priority is legible. Buy-side firms, at several times the rate of brokers, are reaching for AI to find deals and owners first, and the freshest wave of that demand is concentrated on sourcing specifically. That is a materially different first move than the industry story of "AI will speed up your underwriting." The underwriting use is happening. It is just second.

Why Sourcing Wins the Priority Fight

The data is clear on what operators are doing. The more useful question for your own sequencing is why, because the reason tells you whether the pattern applies to your firm. In our reading of the calls, it comes down to edge versus efficiency.

Underwriting AI is an efficiency play. It makes a job you already do faster and cheaper. That is valuable, but it does not change the set of deals you get to look at, and in a market where everyone is buying the same underwriting tools, a faster pro-forma is quickly table stakes rather than an advantage. Sourcing AI is an edge play. If it helps you find and reach an owner before a deal is broadly marketed, it changes your actual opportunity set. It gets you into conversations your competitors are not in yet. For a buy-side operator, that is the difference between doing the same work faster and doing different, better deals. Edge beats efficiency in what people prioritize, and sourcing is where the edge is.

There is a supply-side reason too. A large share of the operators we talk to are already self-serving on the analysis side. Six distinct ICP operators told us on recorded calls that they are running AI in their own ChatGPT and Claude accounts today, and several of them described pointing it straight at underwriting: dropping an offering memorandum in and getting a filled-in model or a first-pass letter of intent back out. In their own words, one described building a prompt that turns an OM into an LOI; another said he had templates for different deal types and a virtual assistant loading the OMs. If the analysis side is already partly solved with off-the-shelf tools, the unsolved frontier, and therefore the loudest demand, moves upstream to sourcing, where a generic chatbot does not have the data or the workflow to help.

When Underwriting AI Still Comes First for Your Firm

This is a benchmark, not a prescription. Peer demand pointing at sourcing does not automatically mean sourcing is your firm's right first move. Read the pattern against your own constraints, because there are clear cases where underwriting AI should still go first.

• Your deal flow is already strong, but your throughput is the bottleneck. If you see plenty of deals and the constraint is how many your team can actually screen and model without dropping quality, underwriting automation is the higher-leverage first move. There is no point sourcing more deals you cannot process.
• Your underwriting is inconsistent across the team. If two analysts model the same deal three different ways, an AI-assisted, standardized underwriting workflow buys you reliability and comparability before it buys you speed. That is worth sequencing first.
• Your edge already comes from proprietary relationships, not search. If your sourcing advantage is a broker network or an owner rolodex that a tool will not replicate, then the efficiency play on analysis is where AI adds the most on top of an edge you already hold.

The honest reading of the data is not "do sourcing, everyone else is." It is that your peers are pointing AI at sourcing first because, for most buy-side firms, that is where the edge is and where off-the-shelf tools cannot help. If your firm's binding constraint is genuinely somewhere else, sequence to that. Just make the choice deliberately, against your own bottleneck, rather than defaulting to underwriting because it is the assumed answer.

What This Means for How You Sequence AI

Zoom out to the field for a second, because the sequencing question sits inside a broader adoption reality. Per JLL's 2025 Global Real Estate Technology Survey, published October 28, 2025 and drawn from more than 1,500 decision-makers across 16 markets, 88% of investors and owners have started piloting AI, up from under 5% in July 2023. And per Deloitte's 2026 Commercial Real Estate Outlook, published September 29, 2025 across more than 850 C-level respondents in 13 countries, roughly 75% of respondents plan to increase their real estate investment over the next 12 to 18 months. Nearly everyone is piloting, and the capital is expanding. In that environment, the question is no longer whether to use AI. It is which workflow to make reliable first.

Our first-party data gives buy-side firms a clean default answer to that question. If you have no strong internal reason to sequence otherwise, point AI at sourcing first, then underwriting, because that is where the demand, the edge, and the unsolved problem all line up. If you have a strong internal reason, the throughput or consistency or relationship cases above, sequence to that instead. Either way, the move is deliberate.

Two more practical notes for whichever workflow you pick. First, the operators ahead of you are not laggards you are outrunning; they are running AI in their own accounts already and are stuck at the same reliability ceiling you would hit, which means the real work is not access to AI but turning it into a process you can trust without checking every line. Second, whichever workflow you choose, the tooling matters. If sourcing is your first move, we compare the options in the best off-market deal-sourcing software.