佣金要按入学批次预测,而不是按月。
按全年平均,公司看起来没问题。十月的银行余额说的不一样,而余额才是对的。
Visatory · · 阅读 5 分钟
本文尚未翻译。下方正文为英文,上方的标题和摘要则不是。
In October the owner opens the year's plan and the plan says the agency is healthy. Commission for the year, divided by twelve, comfortably covers the salaries. The trouble is that the bank account does not divide by twelve. The September cohort has only just passed census, the money will not arrive until the new year, and payroll is on the twenty-fifth.
A monthly forecast is the wrong shape for this business. Costs run by the month. Income runs by the intake, arrives months after the students have enrolled, and depends on a chain of decisions made by families, institutions and visa officers. Most agencies respond in one of two ways. They add up everybody in the pipeline and call it the forecast, or they take last year's total and add a hopeful percentage. The first treats an application as money. The second has no idea which students it is counting.
Give each intake its own forecast.
One line per stage. Multiply the number of students at that stage by the probability that a student at that stage ends up as commission actually paid, then by the commission expected per student. The total says nothing about any individual student. It is an expected figure for the group, and it gets better every time a student moves stage.
A worked example, with invented numbers.
Every figure in this section is made up to show the arithmetic. None of it describes any real agency, rate or success level. Suppose it is May, the agency is looking at its September intake, and it expects commission of 1,000 per student in whatever currency it invoices.
- Application submitted: 40 students at 30 per cent, which is 12 expected enrolments.
- Offer received: 30 students at 50 per cent, which is 15.
- Deposit paid: 20 students at 80 per cent, which is 16.
- Visa granted: 10 students at 95 per cent, which is 9.5.
That is 52.5 expected enrolments from 100 students in the pipeline, so 52,500 of expected commission. An owner who counts the whole pipeline will say 100,000. An owner who counts only granted visas will say 10,000. Neither number is any use for deciding anything.
A more useful pair is the expected figure and a committed figure that counts only students at deposit or beyond: in this example 25.5 enrolments, or 25,500. The expected figure says what the intake is probably worth. The committed figure says how much of it can be spent against before the intake has happened.
Now add the timing, which is where the method earns its place. Suppose, still inventing, that the February intake forecasts 25,000 on the same method. Suppose salaries and rent come to 5,000 a month, or 60,000 a year. Suppose February's money finishes arriving in July, and September's arrives in January.
Viewed by the year, that is 77,500 of commission against 60,000 of costs: a comfortable surplus, and about 6,450 a month on average. Viewed by the intake, nothing arrives from August to December. Five months at 5,000 means 25,000 has to be in the bank at the end of July. An agency that starts August with 10,000 cannot pay its October salaries, in a year it will finish well ahead.
How to get your own probabilities.
The percentages above are placeholders, and borrowing somebody else's is worse than having none. Yours are in your own records, if the records have dates in them.
- Take the last two or three intakes and list every student who ever reached each stage, whatever happened to them afterwards.
- For each stage, count how many of those students went on to produce commission that was actually paid. Divide. That is the probability for that stage.
- Use the date each stage was reached, not the student's current stage. A record that holds only today's status cannot answer this, so if yours does not keep the dates, start keeping them this week.
- Separate the groups only where the difference is real: destinations with different visa outcomes, direct students and sub-agent students. A percentage based on six students is an anecdote.
- Take commission per student from the contract, on the tuition the rate applies to after scholarships, in the contract's currency, at a cautious exchange rate.
- Take the timing from what each institution actually did last time, rather than from the payment term written in the agreement.
The third step is the one that catches most offices. The history they need is the dates, and the dates were overwritten every time somebody updated a status. If that is the position, a rough reconstruction from invoices and email for one intake is still better than a guess, and a clean set arrives in twelve months.
Re-run it monthly and keep every version.
Each month, update the stage counts for every open intake, re-run the sum and keep the old version. After census, compare what the forecast said in May, July and September with what was invoiced and paid. That comparison shows whether the probabilities are too generous, and at what point in the year the forecast starts to be believable. For most owners it will be later than they assumed.
A forecast also has to change a decision, or it is decoration. The obvious ones are whether to hire before an intake or after it, how much the owner can safely draw, when to renew a lease, and how much pressure to put on an institution about an unpaid invoice from the intake before.
Find your own July number.
Before the end of this week, work out one figure: the number of months between the last lump of commission and the next, multiplied by monthly costs. That is what the business must hold on the day the last payment lands. Put it next to the current balance. Most forecasts are about how much will come in. This one is about whether the office is still open when it does.