ROI on automation is the value an automation returns to your firm, minus what it costs, expressed against the time it takes to pay for itself. In plain terms: if a task eats ten hours a week and a tool takes that to two, the ROI is the eight reclaimed hours priced at what your people cost, set against the price of the tool and the work to set it up. That is the whole idea, and for a 4–20 person commercial real estate firm it is worth being precise about, because the number is easy to fake and the honest version is what keeps you from buying the wrong thing. This piece gives you a model simple enough to run on a napkin, a worked example on a real CRE task, and the part most calculators skip: how to turn the number into a decision.
What ROI on Automation Actually Means
Return on investment is an old idea with a simple shape: what you got back, divided by what you put in. Applied to automation, “what you got back” is almost always time — hours your team no longer spends on a repetitive task — converted into money at what those hours cost you. “What you put in” is the full price of the automation: the subscription or build, plus the setup, training, and upkeep to make it stick.
Two framings answer different questions. The ratio — net value divided by cost, times 100 — tells you whether the thing is worth more than it costs. The payback period — how many months until the automation has returned more than you spent — is the one a small firm should lead with, because it answers what an owner actually asks: not “is this a good ratio” but “how long until this stops costing me and starts paying me.”
The trap is treating ROI as a single figure you compute once and file away. It is more useful as a lens. The number tells you whether to act, how fast to expect a return, and — the part almost every online calculator ignores — which kind of move is worth making. A payback of two months and a payback of two years call for completely different decisions.
The Four Inputs You Need
You do not need a finance background or a spreadsheet with thirty rows. You need four honest inputs.
- Hours per week the task consumes. Count the real time your team spends on the specific repetitive task, across everyone who touches it. Be honest and count the whole task, including the cleanup and the back-and-forth, not just the obvious part.
- Loaded hourly cost of the people doing it. Not the billing rate — the cost. Take salary plus taxes, benefits, and overhead, and divide into an hourly figure. A person on a $90,000 salary costs the firm roughly $60–70 an hour fully loaded, not the $45 their base rate implies.
- Realistic time compression. How much of the task disappears once the automation is real and the team is using it well. Model something you can defend — reclaiming a third to a half of the hours on a well-chosen task is credible; reclaiming ninety percent is a vendor fantasy you will have to walk back in front of your partners.
- Total cost, including the parts nobody quotes. The sticker price is the smallest line. Add setup and integration, the hours your team spends learning it, ongoing maintenance, and — for anything touching confidential deal data — the cost of doing it securely. A subscription that costs $40 a month can carry hundreds of dollars in setup and training before it saves you a minute.
Run those four and the arithmetic is grade-school. Weekly hours reclaimed, times the loaded rate, times fifty-two, gives the annual return. Divide the total first-year cost by the monthly return and you have the payback period in months. That is the model. Everything else is honesty about the inputs.
A Worked Example on a Real CRE Task
Take a task every commercial real estate firm knows: abstracting leases and drafting the routine documents around them — pulling key terms out of a lease into a summary, turning that into a letter of intent, cleaning up a market write-up. Illustrative numbers only; the point is the method, not the figures.
Say two people at a brokerage spend a combined eight hours a week on lease summaries and first-draft LOIs. Their loaded cost is $65 an hour. A general assistant like ChatGPT or Claude, applied by a team that knows how to prompt it on real documents, credibly compresses that by half once the habit is real — call it four hours a week back.
Four hours a week at $65 is $260 a week, or roughly $13,500 a year. The tool itself, at business tier for two seats, runs a few hundred dollars a year. The real cost is getting the team genuinely fluent — a hands-on workshop on the firm’s own documents, which at market rates sits in the $2,000 to $15,000 range depending on depth. Take a middle figure of $6,000 in first-year cost, almost all of it the training rather than the software.
That is a first-year return of about $13,500 against a cost of about $6,000: the automation pays itself back inside six months and returns more than double its cost in year one, with the return continuing at near-zero marginal cost after that. Notice what did the work. It was not a $60,000 custom build — it was the general tools your team can already open, plus the fluency to use them well. That pattern, where the cheapest move carries most of the return, is the one small firms miss most often, and it is the reason we argue for sequencing AI spend so fluency comes before software.
The Two Returns the Calculator Cannot See
The four-input model captures hard-dollar hours, and hours are the honest core of any ROI claim. But for a small firm, two returns that do not fit the hours-saved box often matter more than the ones that do — and every calculator drops them because they will not go in a cell.
Compounding team fluency. The first time your team automates a task, they learn how these tools behave. That skill transfers. A team fluent enough to compress lease summaries can turn the same judgment on email, comps, and market research next quarter without a new purchase. The return on the first automation is partly the hours it saves and partly the capability it builds, and the capability keeps paying on tasks you have not touched yet. The hours-saved number understates it every time.
Captured proprietary data. When your team works a real task with AI, you begin to see which of your workflows are standard and cleanly automatable and which are judgment-heavy and document-heavy, where a build might pay off later. That map — of where your hours actually disappear — is proprietary, and it is what lets you scope a real automation with precision instead of guessing. A firm that has drawn it can commission a build that pays back; a firm that has not is gambling. It also tells you where the cheap route runs out, which is the subject of our look at where no-code and low-code tools stop working.
Neither return shows up in a payback figure, which is exactly why a firm that optimizes only for the visible number tends to buy narrow point tools and miss the compounding assets. The full case for treating these as first-class returns runs through the small-firm operating manifesto that frames how a lean shop out-operates larger rivals.
From a Number to a Decision: Subscribe, Train, or Build
Here is the part the calculators leave out. The payback figure is not the answer — it is the input to a decision, and for a small CRE firm there are three moves on the menu, ordered by cost and reversibility.
Subscribe. For a task an off-the-shelf tool already handles well — a CRM, a listing platform, a document tool — a subscription is the cheapest, most reversible move. If the payback math works, subscribe and move on. You can cancel next month if it does not stick, so the downside is small and the ROI bar is low.
Train the team. For the broad band of tasks that general tools plus a skilled operator can compress — drafting, summarizing, first-pass analysis — the highest-ROI move is usually not more software. It is fluency. As the worked example showed, the training is the cost and the return is large and durable, and it improves every later decision you make about tools and builds.
Build. A custom automation for a proprietary, high-volume workflow can deliver the biggest return of the three — and it is the most expensive and least reversible. Market rates for custom automation run roughly $25,000 to $150,000, and once commissioned the spend is largely sunk. The ROI bar here is high, and you should only clear it once your own newly fluent team can tell you exactly what to build.
The discipline is to run the payback math and ask which move it argues for, cheapest and most reversible first. A short payback on a subscription is a clean yes. A short payback that would require a six-figure build deserves more scrutiny, because the risk profile is entirely different even when the number looks identical. That reversibility axis — cancelable subscription versus sunk build — is the spine of the buy-versus-build playbook, and the three-year subscription-versus-build math shows how the same task can flip from “subscribe” to “build” as volume grows.
How Vendor ROI Calculators Get Gamed
Most ROI calculators online are built by firms selling the cure, and they are gamed in two predictable directions. Knowing the moves lets you read any number a vendor hands you.
They inflate the numerator. The hours-saved figure assumes a compression you will not hit for months, if ever, and often folds in speculative new revenue — “deals you’ll close because you moved faster” — which is real but unmeasurable in advance. Strip it out and model a compression you would stake your own money on.
They shrink the denominator. The sticker price is quoted; the setup, integration, training, and maintenance are not. For a small firm without an IT department, those hidden lines are frequently larger than the subscription itself, and they are what turns a promised two-month payback into a two-year one.
One honest discipline fixes both: measure a baseline first. Time the task for a week before you automate it. Firms that document the starting point reach real returns faster than firms that deploy on a vendor’s assumed numbers, because they can see whether the compression actually happened instead of trusting a projection. If you take one habit from this piece, take that one — it is free, it takes a week, and it converts ROI on automation from a marketing figure into a number you can stand behind in front of your partners.
Frequently Asked Questions
What is ROI on automation, in one sentence? It is the value an automation returns to your firm — almost always reclaimed hours priced at what your people cost — minus the full price of the automation, expressed against the time it takes to pay back. For a small firm, lead with the payback period (months until it has returned more than you spent) rather than the ratio, because that is the question an owner actually asks.
How do I calculate the ROI of an automation for my firm? Use four inputs: weekly hours the task consumes across everyone who touches it, the loaded hourly cost of those people (salary plus taxes, benefits, and overhead), a realistic time compression you can defend, and the total cost including setup, training, and maintenance. Weekly hours reclaimed times the loaded rate times fifty-two gives the annual return; total first-year cost divided by the monthly return gives the payback period in months.
What is a good payback period for a small-firm automation? It depends on reversibility more than on a fixed benchmark. Cancelable subscriptions and team-training moves often pay back inside a few months, and because you can reverse them the bar is low. A custom build that runs into six figures and cannot be undone deserves a much higher bar even if the raw payback looks similar.
Why should I use payback period instead of a percentage return? The percentage answers “is this a good ratio”; payback answers “how long until this stops costing me,” which is the decision a small firm is actually making. A high percentage return on a slow, irreversible project can still be the wrong move for a lean firm that needs to preserve optionality.
What costs do firms forget when calculating automation ROI? The ones no vendor quotes: setup and integration, the hours your team spends learning the tool, ongoing maintenance, and the cost of handling confidential deal data securely. For a firm without an IT department these hidden lines are often larger than the subscription itself, and leaving them out is what turns a promised short payback into a long one.
Does ROI on automation include things other than hours saved? The honest hard-dollar core is hours saved, and you should build the number on that. But two returns do not fit the calculator and often matter more for a small firm: compounding team fluency, which pays off on tasks you have not automated yet, and captured proprietary data about where your hours actually disappear. Model the hours, but do not treat them as the whole return.
Should a small CRE firm build custom automation to get the best ROI? Usually not first. The highest-ROI opening move for most 4–20 person firms is training the team on general tools they already have, because the cost is a workshop and the return is large and durable. Custom builds, at roughly $25,000 to $150,000, can deliver the biggest return but are the least reversible, so clear that bar only after a fluent team can specify exactly what to build.
How do I know if a vendor’s ROI calculator is trustworthy? Check whether it inflates the numerator with a compression you will not hit or with speculative new revenue, and whether it shrinks the denominator by hiding setup, training, and maintenance. Then ignore its number and measure your own baseline: time the task for a week before automating it.
Why do so many automation projects fail to deliver the ROI that was promised? Most failures cluster where a firm bought software before building the fluency to use it, or deployed against a vendor’s assumed numbers without measuring a real baseline. Broad industry surveys have found that while most real estate firms have started AI pilots, only a small fraction report hitting most of their goals. Starting with fluency and a measured baseline, and deferring irreversible builds until the evidence is in, is the pattern that avoids that column.
Where to Start
Do not start by buying a platform, and do not start by trusting a vendor’s calculator. Start with a baseline: pick the single task that eats the most repetitive hours at your firm, time it honestly for a week, and run the four inputs on your own numbers. The payback figure will tell you whether to act — and, just as important, whether the right move is to subscribe, train the team, or eventually build.
If you want help finding the task and pricing the return, that is exactly what a free AI-readiness assessment does: a working session that inventories where your team’s hours disappear, estimates the real ROI on automating each candidate, and returns a plain-language plan — including the honest calls about where the answer is a subscription, where it is a workshop, and where a build should wait until you have the evidence to scope it. Book a free AI-readiness assessment if you want that map for your firm. You will leave with a defensible number and a ranked plan, whether or not any of it turns into a project.
Arthur Wandzel
