Most brokers we talk to look at their numbers on Friday, when nothing can be done with them, and they look at whatever their software puts in front of them rather than the ones that decide something. This guide is the opposite: seven numbers, on Monday, with the formula to get each by hand and the concrete decision it triggers.
The manual formula is deliberate. If you need software to calculate them, they are not your numbers — they are your vendor’s. Everything below comes out of a spreadsheet in fifteen minutes, and works just as well if you never buy anything.
1. Pipeline by stage
How many deals sit in each stage — new, contacted, appointment, offer, under contract — and what they add up to in potential commission. You get it by counting, literally. The decision it triggers: if one stage swells while the next empties, that is the bottleneck and that is where Monday’s meeting goes. A pipeline that grows at the top and does not move at the bottom is not growth, it is accumulation.
2. Lead-to-appointment conversion
Appointments booked divided by new leads in the same period, as a percentage. It is the number that fastest reveals whether your problem is quantity or quality. If a hundred leads produce two appointments, buying more leads is burning money: what is failing is the first call or the response time. Look at it per agent and per source, never only in total.
3. Projected GCI against target
Add up the gross commission of each deal under contract multiplied by its probability of closing — use an honest three-value scale, not invented decimals — and compare it against the quarter’s target. The decision: if the projection falls short, the answer is not to motivate, it is to count how many appointments close the gap and share them out. A goal without arithmetic is a poster.
4. Active listings and their days on market
How many listings you have alive and how many days each has been sitting. Sort them longest first. The top three on that list are this week’s price conversation — not next month’s, when the seller is already angry. How to read the market around those days is in the Florida housing market snapshot, which is refreshed every quarter.
5. The source of every new lead
Where each lead that came in this week came from: portal, referral, open house, farming, social, your own site. It gets recorded in the moment or it never gets recorded. Six weeks of this tells you which spend to keep and which to cut, and that single decision usually pays for the year. Without it, the marketing budget gets allocated by habit.
6. Leads untouched for seven days or more
The most uncomfortable number and the most profitable one. Count assigned leads with no activity at all — call, message, note — in the last week. In almost every brokerage that looks at this for the first time, the number has two digits. The decision is not to scold: it is to reassign. A lead untouched for a week belongs to nobody.
7. Activity per agent
Calls, messages, appointments and stage advances per person. Not to police: to tell apart the person with a method problem from the person with a volume problem. Those are opposite conversations, and without this number they get confused constantly. And watch the reverse too: plenty of activity with little conversion is also a problem, just a friendlier one.
The Monday routine, in twenty minutes
- Five minutes: write the seven numbers on the same sheet as always, under last week’s. The series is worth more than the single figure.
- Five minutes: mark the two that moved most, up or down. Only two.
- Five minutes: one decision per number, with a name and a date attached. "Talk to the team" is not a decision; "Ana reassigns the eleven cold leads today" is.
- Five minutes: check whether last Monday’s decisions actually happened. This is the step nearly everyone skips, and the one that turns a routine into results.
What the sheet looks like
One row per week and seven columns, plus an eighth for the decisions. No per-agent tabs and no charts: the point is that it fits on one screen and that the series can be read at a glance. Eight weeks in, the trends are visible, and trends are what actually inform — a single figure rarely says anything, while two in a row already tell a story.
Two details that look minor and are not. First: the numbers get written down on Monday even when they are ugly, especially when they are ugly; a series with gaps on the bad weeks lies by omission. Second: always the same cut-off. If one week you count through Sunday and the next through Tuesday, the comparison stops working and the whole effort of gathering them is wasted.
The numbers that get watched too much
Social followers, website visits and the all-time lead count in the database are the three stars of this category. It is not that they mean nothing: it is that they trigger no decision on a Monday morning. A number that does not change what you will do this week is decoration, however good it looks on a dashboard.
Watching fifteen indicators does not help either. Seven survive; fifteen get abandoned in three weeks — and an abandoned routine informs worse than none at all, because it leaves the feeling that measuring was already happening.
All of this runs on a spreadsheet, which is why it is written this way. If gathering it by hand eventually wears you down, Central Command holds them on one screen and exports them to PDF; but the routine produces the result, not the tool. A broker with a spreadsheet and Monday discipline beats one with a beautiful dashboard opened on Friday.