The commission rollercoaster: how agents build income they can predict

The cycle, and why it repeats

Every agent knows the shape of it. February is terrifying — nothing under contract, the phone quiet, a mortgage payment due. So you prospect like your life depends on it, because it does. In March and April the work lands: three buyers, two listings, everything at once. You disappear into transactions. May and June you close, the account fills, and for about six weeks you feel like a professional.

Then July is quiet. And August is worse. And you cannot understand it, because you have never worked harder in your life.

Here is the mechanism, stated plainly: the famine you are experiencing today was created ninety days ago, during a month when you were too busy to prospect. Real estate has a long production lag. What you do this week does not pay this week; it pays in one to four months. So when you stop prospecting during a busy stretch, nothing bad happens immediately — the consequence arrives a full quarter later, by which time you have forgotten the cause and are looking for a different explanation. The market. The season. Rates. Bad luck.

This is not a discipline flaw. It is a structural feature of any business where production and fulfillment are done by the same person and separated by a long delay. Which means the fix is structural too: prospecting has to become a fixed cost of the week, like rent, rather than a response to how the month is going.

The 90-day lag nobody plans for

Write down the real timeline for a single deal from your own business, and the fix becomes obvious.

Stage	Typical elapsed time
First contact to real conversation	0–30 days
Conversation to appointment	3–45 days
Appointment to signed client	0–30 days
Signed buyer to accepted offer	14–90 days
Under contract to closing	21–45 days
Closing to commission in your account	0–14 days

Even at the fast end, that is roughly two months. In the middle of the range it is four to six. Which means the honest way to read your income statement is this: this month's deposits are a report card on the prospecting you did last quarter.

Two consequences follow, and both are worth internalizing.

You cannot fix a bad month in that month. By the time the account looks alarming, the only levers you have are on deals already in flight. Panic-prospecting in a famine month is still correct — it fixes the month after next — but it will not save the one you are in.

Your busiest weeks are your most dangerous. The month with five closings feels like proof of success. It is actually the moment of maximum risk, because it is when the pipeline behind those closings is most likely to be starving.

The backwards math: from income to conversations

Most agents set goals at the top — "I want to do 24 deals" — and then hope. The useful direction is the opposite: start with the income you need and work backwards through your own ratios until you arrive at a weekly activity number that does not change with the weather.

Here is the full calculation with representative numbers. Substitute your own at every line.

1. Personal income needed: $96,000 after business expenses and taxes.
2. Gross commission income required: assume roughly 35–40% goes to taxes, brokerage splits, and business costs, so target about $155,000 GCI.
3. Average net commission per closing: say $8,500 after the split.
4. Closings needed: $155,000 ÷ $8,500 ≈ 18 closings.
5. Signed clients needed: if 80% of signed clients close, you need about 23.
6. Appointments needed: if you convert 45% of appointments to signed clients, that is about 51 appointments.
7. Real conversations needed: if one in six conversations becomes an appointment, that is about 306 conversations.
8. Weekly target: 306 ÷ 48 working weeks ≈ 6–7 real conversations per week.

Six or seven conversations a week. That is the whole business, reduced to a number you can hit before lunch on Tuesday. It is also a number you can miss for three straight weeks without noticing — which is precisely how the rollercoaster starts.

Two notes on doing this honestly. First, use your ratios, not benchmarks from a coaching program; if you do not know them, track for one quarter and you will. Second, a "real conversation" means a two-way exchange about their situation and timeline — not a voicemail, not a mass email, not a like on a post.

Pipeline stages that tell you the truth

A pipeline you cannot see is a pipeline you will overestimate. Every agent in a slow month has the same conversation with themselves: "I've got a lot cooking." Written down, "a lot" is usually four names, two of which went quiet in April.

Six stages are enough, and the discipline is that a contact only advances on an event, never on a feeling:

• New — inquiry received, not yet reached.
• Contacted — you have had a two-way conversation. Motivation and timeline captured.
• Appointment set — a specific date and time exists.
• Signed — a buyer agreement or listing agreement is executed.
• Active — actively touring or listed, offers in play.
• Under contract — executed contract with a closing date.

Then attach two numbers to each stage: how many people are in it, and the expected value of the stage (count × average commission × your historical conversion rate from that stage). The result is a forecast that is boring and accurate, rather than optimistic and wrong.

The most useful thing this exposes is not the total. It is the shape. A pipeline with eight under contract and two contacted is a pipeline in trouble, and it will feel wonderful for another sixty days before the trouble becomes visible in your bank account. Conversely, twenty conversations and one closing is a business that is about to work.

Balancing your lead mix

Consistency also depends on where the business comes from, because different sources have completely different lag profiles and cost structures.

• Sphere and past clients. Highest conversion, lowest cost, slowest to build, and the most durable. This is the foundation, and it is systematically underworked because it feels like it should happen on its own. It does not — see the repeat and referral guide.
• Referrals from other agents and professionals. Excellent conversion, requires deliberate relationship maintenance with lenders, attorneys, contractors and out-of-area agents.
• Portal and paid online leads. Low conversion, immediate volume, expensive, and completely dependent on speed to lead and follow-up stamina. Useful for filling gaps, dangerous as a foundation.
• Open houses and sign calls. Moderate conversion, cheap, and heavily dependent on capture — most of the value leaks at the sign-in sheet.
• Geographic or niche farming. Long lag, compounding returns, only works with consistency measured in years rather than months.

A business whose income depends on a single source is fragile in a specific way: when that source changes — a portal's pricing, an algorithm, a referral partner retiring — income does not dip, it disappears. Aim for no single source producing more than about half your closings, and treat sphere as the base layer that every other source sits on top of.

The non-negotiable prospecting hour

Everything above resolves into one habit. If you take nothing else from this guide, take this.

One hour a day, same time, every working day, spent on new and existing relationships — regardless of how busy you are.

Three rules make it stick.

It happens first. Not after email, not after the transaction fires are out. The hour that follows other work never happens, because other work is infinite. Put it at the start of your working day and treat it exactly like a listing appointment: it is on the calendar, and you do not cancel appointments.

The list is prepared the night before. The most common failure mode is spending the hour deciding who to call. Twenty names, ready. Past clients due a touch, leads with a next step falling this week, sphere members you have not spoken to in ninety days, and anyone whose stated timeline arrives this month. If your database can produce that list in two minutes, the habit survives; if the list takes twenty minutes to assemble, the habit dies. This is the practical reason database hygiene matters — a point covered in the CRM guide.

It counts conversations, not attempts. Six to seven two-way conversations a week is the target from the math above. Dials do not count. Voicemails do not count. This keeps the hour honest.

An assistant that keeps the list current and surfaces the right twenty names each morning is the difference between a habit that survives a busy month and one that quietly stops. That is exactly the job Heykeyper does: every lead has a next step with a date, and the morning view tells you who needs attention today — including during the weeks you would otherwise have skipped.

Managing cash on commission income

Even a well-run pipeline produces lumpy deposits. A closing calendar is not a payroll schedule, and treating it like one is what turns a normal three-week gap into a crisis. The financial structure below is standard practice among agents who have stopped feeling the rollercoaster even though their income still arrives unevenly.

Separate accounts. Business operating, tax reserve, and personal. Commissions land in business. Nothing moves to personal except a scheduled draw.

Take taxes off the top, immediately. Self-employment tax plus income tax commonly lands somewhere between a quarter and a third of gross for U.S. agents, depending on bracket, entity and deductions. Move that percentage to the tax account the day the commission arrives, and make quarterly estimated payments from it. A CPA who works with commissioned real estate professionals is worth more than any software.

Pay yourself a salary. Decide a fixed monthly personal draw you can sustain across a normal year, and pay it on the same date every month regardless of what closed. The buffer account absorbs the variance. This one change does more for an agent's stress level than any productivity system, because household stability stops tracking the closing calendar.

Build a runway before you build anything else. Three to six months of personal and business expenses in reserve. Below that, every slow month becomes an emergency, and agents who are in emergencies make bad decisions — they take unsuitable clients, discount their fee reflexively, and abandon the long-horizon activities that would have fixed the problem.

Know your true cost per deal. Dues, MLS, lockboxes, signage, photography, marketing, CRM, insurance, mileage, education. Divide by deals closed. Most agents discover their real net per transaction is meaningfully below what they assumed, which changes how they think about both fee conversations and lead spend.

What to do in an actual slow month

You will still get them, particularly early on. The difference between a slow month and a spiral is what you do with the hours the market just handed you.

1. Double the prospecting hour, do not eliminate it. The instinct in a slow month is to work on marketing projects, because prospecting is uncomfortable and a website redesign feels like progress. It is not.
2. Work the database before buying leads. The cheapest business available is the people who already know you. Every past client, every lead with a timeline that has now arrived, every conversation that went quiet in the spring. Reactivation messages routinely produce a live conversation for every four or five sent.
3. Fix one system. One. The follow-up cadence, the transaction checklist, the database cleanup. Slow months are the only time these get built, and each one raises conversion permanently.
4. Have the honest conversation about your ratios. If conversations are high and appointments are low, the problem is your consultation. If appointments are high and signings are low, the problem is your value conversation — see the buyer agreement guide. Slow months are diagnostic gifts if you use them.
5. Do not discount reflexively. Fee cuts made from fear rarely win the business you were losing and permanently reset what you can charge the people who would have paid.

The weekly review that keeps the pipeline honest

Thirty minutes, same time every week — Friday afternoon works well because it lets you set up Monday. Five questions:

1. How many real conversations did I have this week? Against your target from the backwards math.
2. What moved stages, and why? Every advance should trace to an event. If nothing moved, that is a finding, not a mood.
3. Does every person in the pipeline have a next step with a date? The single strongest predictor of whether a pipeline produces.
4. What does the next 90 days look like based on current stages? Not hope — the count times your conversion rates.
5. What is the one thing I will do differently next week? One. Not a new system; a single adjustment.

Agents who run this review for two quarters stop being surprised by their income. They can tell you in July what October looks like, and if October looks thin they still have time to change it — which is the entire difference between a business and a series of hopeful months.

The commission rollercoaster is not the price of the profession. It is the predictable output of prospecting only when you are not busy. Fix the input — one protected hour, a written cadence, a visible pipeline, a salary paid from a buffer — and the output flattens out. Not immediately: the lag cuts both ways, and it takes about a quarter to feel it. But it does flatten, and it stays flat as long as the hour stays protected.

Frequently asked questions