I talk to a lot of estate brokers. Mostly in Dubai and the UK, some in the US. The conversation usually starts the same way.
They tell me they're selling. Or showing. Or negotiating. Or networking. Whatever the word is, it's the part of the job that makes the commission land.
Then I ask them what they did yesterday.
What they list, almost without exception, is admin. Writing vendor update emails. Chasing a solicitor on a commission that should have cleared three weeks ago. Replacing a Trakheesi permit the developer forgot to extend. Booking a viewing. Confirming a viewing. Confirming a viewing again because the buyer changed their mind about the time.
Not one of those tasks earned a penny yesterday. But they collectively ate the day.
If you sit with a broker and walk through a real week hour by hour, the admin total usually lands between twelve and fifteen hours. Some weeks it's eighteen. I haven't met one under ten.
Call it thirteen. A third of the working week.
Here's what that looks like broken out:
| Category | Typical hours/week |
|---|---|
| Email drafting (replies, not new threads) | 4–6 |
| Vendor updates (weekly per listing) | 2–3 |
| Commission chasing (solicitors, developers, accounts) | 1–2 |
| Compliance tracking (permits, renewals, logs) | 1–2 |
| Calendar coordination (viewings, reschedules, RSVPs) | 1–2 |
| CRM updating (after every call, every viewing) | 1–2 |
| Listing copy, social posts, newsletter | 2–3 |
| Total | 12–20 |
The numbers are wide because brokers differ. A solo owner in Richmond doesn't have the same compliance load as a mid-agency in Business Bay. But the shape is consistent: the admin half is real, and it's enormous.
Here's the maths most brokers don't run on themselves.
You close, let's say, eight deals a year. Your average commission is around £6,000 per deal. That's £48,000 in gross commission income.
Your effective working hours, the hours where you're actually able to move a deal forward, are about 32 per week if thirteen of your forty go to admin. Over 48 working weeks that's 1,536 hours.
Divide GCI by effective hours. You're earning roughly £31 per productive hour.
Now divide GCI by total hours in the week, including admin. You drop to about £25. Six pounds an hour, just gone.
Or frame it differently: you're doing £6 an hour work for thirteen hours a week. That's £78 a week, £4,056 a year, in pure-admin output. If you could hire that time back from someone else at a lower rate you would, but the work is sensitive, personal, and in your voice. You can't hand it to a VA.
"But the admin has to get done. Someone has to write the vendor update. Someone has to chase the solicitor."
Correct. But ask a different question: does it have to be you writing it from scratch?
Here's the thought experiment.
Imagine you get three hours back a week. Not all thirteen. Just three.
Those three hours, redirected into the selling half (a second viewing you wouldn't have had time for, a prospecting call you kept putting off, one extra instruction conversation with a vendor who was wavering), add something to the top of the funnel.
At your current conversion rate, three more productive hours a week over 48 weeks is 144 hours. If every eight hours of prospecting turns into one instruction (a generous but not unreasonable broker conversion), that's eighteen extra instructions.
At one closed deal for every four instructions, that's four extra closings. At £6,000 each, that's £24,000 in GCI you didn't have before.
Three hours a week. £24,000 a year.
Even if you halve every assumption (conversion, hours, close rate) you're still finding £6,000 a year that wasn't there. For three hours of admin clipped back.
Every broker I've met knows this maths. Most of them haven't clipped three hours off in years. Why?
Because the admin is atomised. It's not one big block you can delegate. It's ninety small interruptions spread across forty hours. A reply here. A reschedule there. A CRM note between viewings. Each one is a two-minute task. None of them is worth outsourcing in isolation.
You can't hire someone to do your ninety two-minute tasks. You can hire someone to do a ten-hour-per-week chunk of admin, but that person has to be in your voice, understand your contacts, know your preferences, and fit into your existing tools. That's a role that costs at least £30,000 a year once you add overhead. You've just eaten most of the £24,000 you saved.
The only way the maths works is if the helper is cheap, in your voice, and triggered by the same events you are; an email landing, a viewing finishing, a compliance deadline approaching.
That's what NORĴI is. An AI helper that watches the same stream of events you do and drafts the two-minute response before you get to your laptop. You approve with one tap. It sends. The three hours a week come back.
Not all admin goes away. You can't get NORĴI to have the instruction conversation. It can't take the buyer to the property. It can't negotiate. The work that requires your judgement stays with you, because that's what you're good at.
What it takes off is the boilerplate. The weekly vendor update that follows the same shape every week. The commission chase that's been sent fifty times before. The compliance reminder you already know how to handle but that eats twenty minutes every Friday afternoon.
Roughly: NORĴI clips about half the admin block, not all of it. The other half is irreducibly yours. But half of thirteen is six. You aim for three, you often find five or six.
This is where it matters. Getting three hours back doesn't help if you fill them with more admin from a different bucket.
The brokers I know who make this work are specific about what fills the hole. They book a standing Friday-afternoon prospecting block. They take the 10 a.m. Tuesday slot and turn it into a second-viewing window. They start one hour earlier on Mondays with a proper week-ahead plan instead of reacting to the inbox from 8:30.
The time re-fills with high-value work only if you fence it. Otherwise it refills with low-value work because low-value work is cheaper to start.
Two places this logic doesn't work.
The first is brokers who love the admin. I've met two. They find the weekly vendor update therapeutic. They enjoy the back-and-forth of chasing a slow solicitor. If that's you, this post isn't for you. Your commission per productive hour is already where it should be because you count the admin as productive.
The second is brokers where the admin IS the deal. A very complex off-plan handover with six moving parts genuinely requires that level of focus. AI can draft, but it can't replace the integration-level thinking. Those brokers benefit from NORĴI clipping the mundane admin around the margins, not from trying to automate the complex stuff.
Everywhere else, the maths is the maths. Thirteen hours a week, six quid an hour of pure-admin output, three-hour clip worth roughly £24,000 a year. Run your own numbers. I'd be surprised if they came out differently by more than ten percent.