The eight numbers a Dubai brokerage owner should read on the first of every month

A thinner market shows up the reporting first

Volume has thinned. CBRE's UAE real estate market review puts the second quarter of 2026 at fewer than 37,000 residential sales worth AED 88 billion, against more than 51,000 sales worth nearly AED 154 billion a year earlier. The ValuStrat Price Index stood at 219.2 points in July 2026, down 1.6% year on year. Fewer enquiries reach each desk, so each one has to be worked properly.

In a rising market a brokerage can be badly run and still profitable: new leads arrive faster than the mistakes compound. That cover has gone. The eight numbers below take under an hour to read, and they show where money is leaking long before the bank balance does.

Read them in the same order every month, using the same definitions, and give each one a named owner. A number nobody owns is a number nobody fixes. Print the report before the first team meeting, so the conversation starts from the same page rather than from whoever remembers last month best.

Number one: median first-response time, by agent

Measure the minutes between a lead landing in your system and the first genuine attempt to reach that person: a call placed, or a WhatsApp sent from the agent's own number. Use the median per agent, not the team average, and exclude nothing. A lead that arrived at 21:40 and was first called at 09:15 counts as eleven and a half hours, not as a fast morning response.

No official benchmark is worth quoting here, so read the shape instead of chasing a target. A bad reading is a median sitting hours away from the enquiry, or a respectable median hiding a tail of leads answered the next day. The cause is rarely laziness. It is leads sitting unassigned overnight and over the weekend, or an agent working from personal WhatsApp while the system says otherwise.

One action moves it. Assign every lead the moment it lands, and name a duty agent for evenings and Fridays so nothing waits for an owner. Raabta reports first-response time per agent, and a manager can split a batch of overnight enquiries evenly across the desk in one tap.

Numbers two and three: cost per lead, and source-to-viewing rate

Count leads by source for the month: Bayut, Property Finder, dubizzle, each social ad account, the website form, WhatsApp, walk-ins and referrals. Divide the full cost of each source by its lead count. Full cost means subscription, ad spend, photography and video, and advertising permits. Dubai Land Department lists AED 1,000 for a standard real estate advertising permit plus an AED 20 knowledge and innovation fee.

Source-to-viewing rate is the share of a source's leads that reached a viewing the client attended. Fix the window once, write it down, and never move it, because a window that quietly stretches turns a weak month into a strong one. A source can look cheap per lead and still be the most expensive thing you buy.

The usual cause of a low rate is a mismatch between the advert and the stock: a Damac Hills 2 budget answering a Dubai Marina listing. Re-brief the adverts for one month. If the rate does not move, cut the source and put the money where viewings already happen.

• ✦Count the whole cost. Subscriptions, boosts, media production and permits all belong in the numerator. Leaving them out makes portals look cheaper than they are.
• ✦Attended viewings only. A booked viewing the client did not attend is a cancellation, not a conversion. Counting it flatters every number downstream.
• ✦Keep walk-ins separate. Walk-ins and referrals convert on a different curve from paid leads. Blending them into one average hides both.
• ✦Read small sources carefully. A source with twelve leads and four viewings is telling you something. Do not dismiss it for low volume.

Numbers four and five: viewing-to-offer rate, and closed deals with gross commission

Viewing-to-offer rate is offers submitted divided by viewings attended in the same month. It is the cleanest test of whether your agents qualify. Judge it against your own trailing three months and against the gap between your strongest and weakest agent. Plenty of viewings producing few offers usually means budget, payment method and timeline were never established before the keys came out of the drawer.

Closed deals and gross commission travel as a pair, never one alone. Count deals that actually completed and the commission invoiced against them, not projections sitting in someone's head. Deal count without commission rewards volume. Commission without deal count hides a team carried by one transaction.

The action for a weak offer rate is unglamorous. Sit with each agent and read the last ten viewings back: what was known about the buyer beforehand, what stock was shown, and what was asked at the end. Two hours per agent, once a month, changes more than any script.

Numbers six and seven: average commission per deal, and pipeline value by stage

Average commission per deal is gross commission divided by closed deals. It tells you what one unit of your team's effort is worth. Falling while the deal count holds steady means the desk has drifted into smaller units, or is quietly discounting to keep deals alive. Both are fixable, but only once you can see which one is happening.

Read it against the market rather than in isolation. Cushman and Wakefield Core reported Dubai sales prices down 4% quarter on quarter in the second quarter of 2026, with Palm Jumeirah apartments down 9% and Downtown Dubai and Business Bay apartments down 7%. In those communities the same unit earns the desk less than last year, a mix question rather than a performance one.

Pipeline value by stage is the commission you would earn if every live deal closed at its current value, with a column showing days since the last logged contact. Without that age column the total is decoration. Raabta's pipeline carries deal value and commission on each card, so the monthly figure falls out of daily work instead of a spreadsheet exercise.

Number eight: the reason each dead lead died

Every dead lead should carry a reason chosen from a short fixed list. Free text is not a reason. This is the one number here that points at the top of the funnel rather than the bottom, which is why it is usually the first thing missing from a brokerage report.

The action is to keep the list short enough to pick from at the end of a call, and to set the closing rule in writing so every agent applies it the same way. Five or six options is plenty. The five below cover most of what a Dubai desk actually sees.

• ✦Wrong number. The contact does not belong to the enquirer, or does not exist. A spike almost always means someone has been buying data.
• ✦Budget mismatch. The client's budget cannot reach the stock advertised. That is an advertising brief problem, not an agent problem.
• ✦Bought or rented elsewhere. The client transacted with someone else. Note how many days passed between the enquiry and the loss.
• ✦No answer after five attempts. Five logged attempts across different times of day and at least two channels. Fewer than that is not a dead lead, it is an unworked one.
• ✦Outside our area or product. A Creek Harbour enquiry on a JVC-only desk. Track it, because a rising count is a listings decision waiting to be made.

Four traps hiding inside an honest report

The report can be accurate and still mislead you. Four traps account for most of it, and each is easier to catch with the full month in front of you than in the middle of a busy week, reacting to whichever number someone put in a message.

A spike in wrong numbers deserves separate attention, because it is a compliance question as well as a data one. RERA fined one brokerage office AED 50,000 and suspended nine of its brokers for three months over direct telemarketing, and its practice guide states that brokers are not allowed to contact owners who are not registered on the Green List in the Dubai REST app.

• ✦Vanity lead counts. Four hundred leads that produced two viewings is not a strong month. Put the viewing count directly beside the lead count so the two are never read apart.
• ✦One outlier carrying the average. A single large Business Bay deal can lift the whole desk's average commission. Show the median beside the average and list the largest deal separately.
• ✦The agent with no problems. An agent whose leads show no bad outcomes usually is not logging anything. Check their logged activity count before you praise a clean sheet.
• ✦Deals frozen in one stage. Set a stale threshold, say forty-five days with no logged contact, and pull anything past it out of the pipeline total for review by name.

None of this works without same-day logging

Outcomes logged three days late are guesses, and a report built on guesses is worse than no report at all, because it carries authority. Same-day logging, on every lead, by every agent, is the entire foundation. Everything above is downstream of that one habit.

So make the habit small enough to be automatic. Logging should take one tap at the end of a call, from the phone, standing in the car park outside the building. If an agent has to open a laptop to record a no-answer, the no-answers never get recorded, and your dead-lead reasons quietly become fiction.

Raabta was built around exactly that. Logging a call outcome sets the lead status, tags the lead, schedules the next follow-up and offers a pre-written WhatsApp message from the agent's own number, so the eight numbers assemble themselves out of work the team was already doing. It is built for Dubai brokerages by Bedflow, with a free trial and no card required.