Two hours a week sounds like nothing.
It's the amount of time your escrow coordinator spends chasing wire confirmations. It's the amount of time your listing coordinator spends re-typing property details into three different systems that don't talk. It's the amount of time your bookkeeper spends reconciling the earnest money account, your marketing person spends resizing images for four platforms, your transaction coordinator spends assembling closing packets.
Two hours. Nobody's going to bring that up in a staff meeting. Nobody's going to make a business case out of it. It's a rounding error.
Two hours a week is 104 hours a year. That's thirteen full working days — more than half a month of somebody's time, gone, on one task.
Now count how many people in your company have a two-hour task. And how many of them have three of them.
The arithmetic nobody does
A thirty-person real estate services firm. Not everyone, just the operations side — say twelve people who touch process work. Each of them has, on average, three recurring tasks that eat two hours a week apiece. That's six hours a week each, seventy-two hours a week across the group.
Seventy-two hours a week. At a blended loaded rate of $42, that's about $157,000 a year.
You do not have a $157,000 line item called "repetitive administrative work." You have twelve salaries, each of which looks reasonable, each of which is roughly fifteen percent consumed by work that a well-configured system would do in the background without complaint.
Fifteen percent of your operations payroll. That's the shape of it at most SMBs we look at, and the range runs from ten to thirty depending on how many systems you've bolted together over the years.
Why big tasks get automated and small ones don't
Because big tasks are visible and small ones are polite.
If a single task took twenty hours a week, somebody would have escalated it years ago. It would have a name. It would come up in the annual planning meeting. It would be an obvious candidate.
The two-hour tasks never escalate, because escalating them makes you sound like you're complaining about something trivial. So they sit there, absorbed, for years — and in aggregate they are far more expensive than the one big obvious thing, and far easier to fix, because each one is simple.
That inversion is the whole insight. The money is not in the one hard workflow. It's in the twenty easy ones.
And the twenty easy ones have a property the big one doesn't: they're low-risk. If the wire confirmation chaser misfires, somebody notices in an hour and fixes it. If your automated title production system misfires, you have a very bad month. The small stuff is where you should be starting anyway, on risk grounds alone.
How to actually find them
Not with a survey. Surveys produce diplomacy.
Do this instead, and it costs you one week and zero dollars. Ask everyone on the operations side to keep a simple log for five days: every time you do something you've done before, in roughly the same way, write down what it was and how long it took. Not a time sheet. A repetition log.
At the end of the week, you will have a list. It will be long. Sort it by (hours × 52 × their loaded rate) and look at the top ten.
Two things will happen. First, you'll find $100,000 you didn't know you were spending. Second — and this is the one owners always mention afterward — you will learn things about how your company actually operates that you did not know and would never have discovered any other way. The two systems that don't talk. The report that three people build separately because none of them knew the others were building it. The approval step that exists because of an incident in 2019 that nobody remembers.
That week of logging is the highest-ROI thing most owners can do this quarter, and it doesn't require hiring anyone.
What you do with the list
You don't automate all of it. You almost certainly shouldn't.
Some of those tasks will turn out to be worth $3,000 a year and cost $15,000 to automate, and the answer there is no. Some will turn out to be symptoms of a system you're about to replace anyway, and the answer is wait. Some will turn out to be work that shouldn't exist at all — a report nobody reads, a double-entry that only exists because of a process nobody has questioned since a merger — and the answer is delete, which is free and immediate.
But some of them — usually four or five — will be worth $20,000 or $40,000 a year each and will be genuinely mechanical. Those are your projects. Those clear a 2x return without any creative accounting, and they clear it in the first year.
The last part, which is the actual point
When you give twelve people six hours a week back, you don't get $157,000 in savings. You get seventy-two hours a week of human attention, released back into a business that has been starved of it.
That's a coordinator who can finally call every buyer the day before closing instead of the ones she gets to. That's a bookkeeper who catches the discrepancy in March instead of at year-end. That's a marketing person who makes something good instead of making something on time.
Two hours a week sounds like nothing. Multiply it by your whole team and it's the difference between a company that's holding on and a company that's growing.