What does a transaction coordinator actually do?

Ask an agent what a transaction coordinator does and you’ll get a version of the same answer: “everything after the contract.” It’s accurate and useless. The interesting part is the inventory: the repeating set of tasks that stand between a ratified contract and a clean closing, and the ways each one goes wrong when nobody owns it.

This guide is that inventory. It’s written for three readers at once: the agent deciding whether to hire a TC, the TC benchmarking their own role against the full job, and the broker designing operations around one. By the end you should be able to answer the practical questions: what the role covers, where its legal edges are, what it customarily costs, and how much of it is about to change.

A transaction coordinator is the person accountable for the file, not the relationship or the negotiation or the price. The file means the documents, the dates, the signatures, and the hand-offs among the half-dozen parties who each control one piece of the deal. The agent’s job is to win the client and negotiate the terms. The TC’s job is to make sure what the parties agreed to happens, on the days the contract says it must, with the paper to prove it.

The economic logic is simple. An agent’s highest-value hours go to prospecting, showing, and negotiating, the licensed work only they can do. Every hour spent chasing a missing initial or re-counting a deadline is an hour of that work not happening. Coordination gets hired out for a plain reason: it is relentless, detail-hostile, and cleanly separable from selling. A deal generates the same run of coordination touches whether the agent closes four deals a year or forty, and only the low-volume agent can still absorb them alone.

What makes the role harder than it sounds is that it is a standing watch rather than a task you finish. Most TC tasks are trivial in isolation. The difficulty is that a dozen files run at once, each on its own clock and each mutating as amendments land, and the cost of missing one date on one file can exceed the fee for coordinating all of them. The role exists because contract deadlines punish the divided attention that a successful sales practice produces.

Here is the whole job, in the order a file experiences it. Forms and customs vary by state and brokerage, and listing-side files differ from buyer-side files in the particulars, but the skeleton below is stable, and it pairs naturally with a written transaction checklist kept next to every file.

Open the file. Before anything else, verify the contract is actually a contract: every page present, every blank filled, every counter and addendum incorporated, every signature and initial in place, and a defensible effective date. Plenty of timelines have been wrecked by an unincorporated counteroffer rather than by any exotic failure. Then open the file in the brokerage’s system and send the introduction email, the one that tells the lender, the title company, the co-op agent, and the client who is coordinating and how to reach them. That email matters more than it looks: it is the moment the TC becomes the file’s single point of contact instead of its fifth copy-recipient.

Calendar every deadline. Read the executed contract and every addendum, pull each date, and apply the form’s own counting convention, the step amateurs skip. Business days or calendar days, whether the effective date counts, which holidays the form recognizes: the definition of “day” usually lives in a paragraph nobody re-reads, and it sets the whole schedule. Each date goes on the calendar twice, the deadline itself and a reminder far enough ahead that something can still be done about it.

Track the earnest money. Confirm the deposit was delivered inside the contractual window and obtain the receipt. Not “the buyer says they wired it”: the receipt itself, from the escrow holder, in the file. In many form contracts, late delivery gives the seller remedies up to termination, which makes this among the cheapest catastrophic misses to prevent. The mechanics and customs are covered in our earnest money guide; the TC’s job here is narrower: proof, on time, filed.

Run the inspection window. Schedule the inspections, confirm access, and watch the expiration date closely, because under many forms silence at the deadline waives the contingency. The TC doesn’t decide what to ask for. The TC makes sure that whatever the parties decide is signed, dated, and delivered before the window closes, and that the answer, even a simple “we’re proceeding,” exists in writing.

Order title, monitor the contingencies. Get title opened early, confirm the commitment arrives, calendar the objection window, and track the lender’s milestones (application, appraisal, conditional approval, clear to close) against the contract’s financing dates. Those run on two different clocks: the lender moves on its pipeline, the contract on its paragraphs, and only the contract’s clock has legal teeth. A weekly lender status in writing turns the deal’s biggest unknown into a paper trail.

Chase what’s missing. Signatures, disclosures, HOA documents, the addendum one side signed and the other forgot: a running share of any TC’s week is polite, persistent follow-up. The skill is systematizing it, keeping a log of every outstanding item, who owes it, and when it was last requested, so that “sent” is never confused with “done.”

Coordinate the week of closing. Confirm the Closing Disclosure went out on the lender’s side; federal law requires it reach the borrower at least three business days before consummation. Schedule the walk-through, circulate the final time and place, reconcile any last-minute amendment, and confirm everyone is bringing what they must. Then, after signatures, confirm recording and funding, collect the settlement statement and final executed set, and close the file out against the broker’s compliance checklist.

The role has a hard legal edge, and it matters more than most task lists admit. In most states, an unlicensed person may not perform activities reserved to licensees, and the ordinary flow of a transaction brushes up against that line constantly.

A TC does not negotiate. Price, repair credits, extensions, who pays for the survey: the moment the conversation becomes “what should we offer back,” it belongs to the agent. A TC can transmit a signed counter. A TC who drafts the counter’s terms or argues its merits with the other side has, in many states, crossed into licensed activity.

A TC does not advise. “Should we waive the inspection contingency?” is a question with one safe answer from an unlicensed coordinator: “Let me get your agent on the phone.” Explaining what a form says is treacherous ground; recommending what to do about it is over the line almost everywhere.

A TC does not interpret the contract’s legal effect. Reporting that a deadline is Thursday is coordination. Opining on whether missing it forfeits the deposit is somewhere between brokerage and law practice, and an unlicensed TC should do neither.

The boundary’s exact location varies by state. Some states publish specific lists of what unlicensed assistants may do, some TCs hold real estate licenses to widen what they can touch, and brokerage policy is often stricter than state law. The safe generalization: communicate, schedule, assemble, track, and remind; never decide, negotiate, or advise. An agent hiring a TC should ask which side of that line each task in the engagement letter sits on, because the license on the hook for the answer is the agent’s.

Every agent runs one of five models, whether or not they’ve named it. The economics below are customs, not quotes; they vary by market, by deal side, and by scope, but the trade-offs are structural.

The agent does it themselves. The default, and at low volume the right answer. The cost is invisible because it’s paid in evenings and in prospecting hours that never happened. The failure mode is success: the model works until the quarter it’s needed most, when the agent has the least attention to give it.

The independent per-file TC. Probably the most common paid model. A contractor charges a flat fee per closed file, customarily a few hundred dollars in most markets and often due only if the deal closes, and carries files for several agents at once. The strengths are obvious: pure variable cost, no payroll, and a seasoned professional who has run a lot of files. The weaknesses are the flip side: the TC’s attention is shared, their systems are their own, and the “paid at closing” custom means the files that die absorb real work nobody pays for.

The in-house employee. Teams and brokerages at volume hire the TC onto payroll. You get control, consistency, and one system across every file, at the price of a fixed cost that doesn’t flex with the market and a single point of failure who occasionally takes a vacation. The break-even is a per-team question: enough monthly files that the salary beats the sum of per-file fees, with coverage solved separately.

The virtual or offshore TC. Remote coordination services, often priced below local independents, handling document chase, calendaring, and status communication. The economics are the appeal; the risks are time zones, turnover, and distance from local forms and customs. The tasks that survive the distance are the checklist tasks. The ones that don’t are the judgment calls, which the next section takes up.

The software-assisted model. Increasingly, the real question is less which human than how much software sits underneath whichever human you choose. A TC, or an agent self-coordinating, working on top of software that reads documents, builds timelines, and watches deadlines carries more files at higher quality than the same person working from an inbox and a spreadsheet. How brokerages are restructuring around that is the subject of our AI back-office guide, and the tools themselves, from checklist platforms to AI review, are compared in our guide to brokerage compliance software.

Ask working TCs about capacity and you get a range, not a number, because the limit is the overhead around the checklist rather than the checklist itself. A file’s visible work (the emails, the scheduling, the chasing) is bounded. The invisible work is not: every fact about a deal lives in one document and is needed in five systems, so the TC retypes it into the transaction platform, the calendar, the status email, the broker’s checklist, and the client update. Then an amendment lands and every copy is wrong.

That re-keying and reconciliation loop is the real capacity governor. At comfortable load, a good TC re-reads every incoming document against the calendar and the file. At overload, the re-reading is the first thing to go, and it goes unnoticed, because nothing visibly fails on the day it stops. The checklist still gets checked. What disappears is the verification behind the checkmark: the amendment gets filed without being reconciled, the calendar keeps showing the closing date from two addenda ago, and the file drifts from the documents while looking well maintained.

This is why capacity numbers vary so widely between operations that look identical on paper. A TC doing intake, extraction, and deadline math by hand tops out at a couple dozen live files; the same TC with automated intake and computed timelines can carry considerably more, because the hours that re-keying consumed go back into the judgment work. The constraint was rarely the coordinator so much as the copying.

Everything so far describes competence. It doesn’t describe why certain TCs have waiting lists. The difference is judgment at the seams, the places where the file passes between parties and a mechanical coordinator sees nothing wrong.

Knowing which slipped date matters. A title commitment that’s two days late is usually a shrug. An appraisal that hasn’t been ordered ten days before the financing deadline is a fire, even though nothing has technically been missed yet. Both look identical on a checklist: a pending item with a future date. A great TC has learned, over many files, which pendings are noise and which are the first visible symptom of a deal going sideways, and escalates the second kind while there is still time to act.

Hearing what the silence means. When a lender who answered within an hour starts taking two days, no deadline has passed and no rule has been broken, yet an experienced TC starts asking pointed questions anyway, because that pattern precedes bad appraisal news often enough to act on. The gap between “nothing is wrong” and “nothing is wrong yet” is where good coordination earns its keep, and it is the core of preventing closing delays rather than narrating them.

Managing people who don’t report to anyone. A TC has authority over nobody: not the lender, not the co-op agent, not the client who won’t sign. The job runs on tone, timing, and the goodwill of being the most reliable person on the deal. That is a real, scarce skill.

Look back at the inventory and sort it into two piles. Pile one: reading documents, extracting parties and prices and dates, applying counting conventions, building the timeline, noticing that an amendment moved a date, checking a file against a compliance checklist, watching the calendar. Pile two: deciding which slipped date matters, reading the lender’s silence, getting a reluctant co-op agent to move. The first pile is mechanical reading and watching. The second is judgment. For as long as the role has existed, both piles have been done by the same person, and the first has consumed most of the hours.

That is the part that’s changing. Software can now read a transaction file. That is the problem Ratifyly was built on. You forward the paperwork the way you’d send it to a coordinator, and it lands with Ezra, our AI coordinator. He reads every page, extracts the parties, the price, and every date with its counting convention, and builds the transaction and its timeline from the documents themselves. When an amendment arrives, he re-reads it against the schedule and the schedule re-flows. The file is audited against compliance rules; every party (agent, client, broker, lender, closing office) watches the same live timeline; and deadlines escalate before they hit. A human approves what counts, because pile two was never the software’s to own.

The read on what this means for TCs: it is not a replacement, it is a reallocation. The re-keying and reconciliation loop, the thing that caps a coordinator’s capacity and degrades under load, is what automation removes. What’s left is the part that was always the point: the judgment at the seams, applied across more files, with the mechanical watching done by something that does not get busy in June. The brokers thinking hardest about this are designing compliance operations where the software does the first read of every file and the humans spend their hours on the exceptions. The best TCs we know ask for that arrangement first.

That reallocation now has a product category attached: software vendors sell the first pile under the name “AI transaction coordinator.” The label covers a wide range, from calendars with an AI intake step to systems that read the contract itself, and the claims deserve the same scrutiny you would give a hire. We wrote a full companion guide on what an AI transaction coordinator is, and what it still can’t do: which of the five capabilities (intake, deadline extraction, reminders, document chasing, status updates) each kind of tool actually delivers, how the honest products keep a licensed human approving what counts, and how to tell a contract-derived timeline from a template wearing the name.