Back to the on-screen lesson ·
What a channel cost divided by the customers it actually won, read beside what a customer is worth over the time they stay, with the people who were coming anyway taken out of the count.
Paper packet. Every task here also exists on screen, where it is checked automatically; answers written on paper are not assessed by Nydus. When you are back at a device, enter your answers there.
You will work out what one customer cost a channel to win, with the customers who were coming anyway taken out of the count, say what that channel earned once its own cost is paid, and set the cost beside what a customer is worth over the time they stay. You will also name what the count leaves out, and why the cost per customer alone settles nothing.
You can work out what one sale contributes — the price less what that sale costs in materials and fees — and you know how to count what customers actually do. This lesson puts a second number beside contribution: what one customer cost to win. Neither is worth much alone, and together they are how an owner chooses where to spend on reaching people.
| Term | What it means |
|---|---|
| Channel | A way of reaching people: a stall, a listing, leaflets, a sign, somebody else's counter. |
| Won | A customer the channel added, who would not otherwise have come. |
| Cost per customer | What a channel cost divided by the customers it won; also called acquisition cost. |
| Customer value | What one customer contributes over the time they keep buying. |
| Payback | The time until a customer's contribution covers what they cost to win. |
| Attribution | Deciding which channel won a customer, usually by asking them how they heard. |
Two channels cannot be compared on what they cost: one spend buys a weekend, another three months. The number that makes them comparable is what one customer cost:
$$\text{cost per customer} = \frac{\text{what the channel cost}}{\text{customers it actually won}}$$
A stall costing 240 dollars that brings twelve buyers cost 20 a customer.
On its own that decides nothing. A customer who costs twenty-five dollars is cheap if they book every week for a year and ruinous if they buy once. So read it in a pair, with what a customer contributes:
$$\text{earned} = (\text{customers won} \times \text{contribution each}) - \text{what the channel cost}$$
Then watch the denominator: people who were coming anyway are not customers the channel won.
This is how to work the number out and read it, not a recommendation about what to charge. What any particular business should do depends on a market nobody here can see, and on rules that differ from place to place.
Another way: table
| Channel | Cost | Won | Each |
|---|---|---|---|
| A weekend craft stall | 240 | 12 | 20 |
| Leaflets to two office blocks | 180 | 45 | 4 |
| A sign and samples | 60 | 30 | 2 |
| A paid listing, three months | 150 | 6 | 25 |
| Cards left with another shop | 200 | 25 | 8 |
Another way: steps
To judge a channel:
Add up the whole cost of the channel. The fee or the printing, plus the owner's hours at the costing rate: a weekend at a fair is two days of selling that could have been spent elsewhere. A channel costed without the hours cannot be compared with one somebody else ran.
Count the customers it won, honestly. Ask every new customer how they heard of the business, and write the answer in the daily record. Leave out inquiries that did not buy, and customers who were coming anyway.
Divide. Spend over customers won is the cost per customer.
Work out what a customer is worth. Contribution per purchase times the purchases they make while they stay. For a one-off buyer that is one contribution; for a regular it may be a year of them. The daily record and the owed list show how long customers from each channel really stay.
Compare the two. A channel whose customers are worth more than they cost is earning; one whose customers cost more than they are worth is losing, however cheap each customer looks.
Check the result by working it out in total as well as per customer: the customers won times their value, less the whole channel's cost, should give the same answer as (value less cost per customer) times the customers won. And check the count after a few months: a channel judged in its first month is judged before its customers have had time to come back.
The sign and samples cost 60 and won 30: 2 dollars each. The paid listing cost 150 and won 6: 25 each. On cost per customer the sign wins by twelve times. But a sample buyer contributes 3 dollars once, and a cleaning customer 40 a month for as long as they stay — so the sign earned 30 and the listing 90 in its first month, and keeps going. A cost per customer is a price, and a price means nothing until you say what was bought.
The denominator always flatters the channel. People already at the stall were not won by the sign, and somebody who heard of you from a neighbor was won by the neighbor. Inquiries are not customers. Ask every new customer how they heard of you, and write it down.
A customer worth far more than they cost to win can still strain the business, because the cost comes first and the value arrives month by month. The payback time is how long it takes a customer's contribution to cover what they cost to win: a customer who cost 60 and contributes 20 a month pays back in three months.
Payback matters for cash. A channel that wins twenty customers at 60 each costs 1,200 now and recovers it over three months; the cash forecast has to carry that gap. A channel with a quick payback — samples, a sign, a referral — can be grown without much cash; one with a slow payback — a listing, a large advertisement — needs the cash to be there first.
So an owner choosing between channels asks three questions, in order: what does a customer cost to win, what is a customer worth over the time they stay, and how long until the cost is paid back? The first two decide whether the channel is worth using at all; the third decides how fast the business can afford to use it.
Many of the best channels for a small business cost nothing to buy and a good deal of the owner's time: talking to people at a market, posting pictures of the work, answering questions in a local online group, dropping in on shops that might stock the product. Because no invoice arrives, these channels are usually treated as free, and compared unfairly with the ones that cost money.
Cost them in hours at the owner's costing rate. Two hours a week posting and replying, at 30 an hour, is 60 a week or about 260 a month. If that effort wins five customers a month, each cost about 52 — more than many paid listings. If it wins twenty, each cost 13, and it may be the best channel the business has. Either way the comparison is now honest.
Hours also have a limit money does not. A channel that works but takes ten hours a week competes with the paid work for the same hours, so its real cost may be the jobs not done. An owner whose schedule is full should favor channels that cost money over channels that cost hours, even at a slightly higher cost per customer; one with spare time can do the opposite.
The simplest habit is to log the hours each channel takes in the same record as the customers it wins. After a month or two the numbers show which channels deserve more time and which are merely enjoyable, and the owner can move the hours accordingly, a few at a time, and count again after each move.
A mobile dog groomer tries three channels over a spring. A listing on a local pet-services site costs 45 a month; a stall at the town's dog show costs 120 plus a Saturday; and a referral card — a free nail trim for any customer who brings a friend — costs her about 8 in time per referral. She asks every new customer how they found her.
Over three months the listing wins 9 customers for 135: 15 each. The dog show wins 14 bookings, but 5 of those people already knew her from the park, so it won 9, for 120 plus a Saturday at her costing rate of 30 an hour for 8 hours — 360 in all, or 40 each. Referrals win 11 customers at 8 each.
Her records show that listing customers stay about 7 months, dog-show customers about 3, and referred customers about 10, each contributing about 28 a month. So a listing customer is worth 196 − 15 = 181; a show customer 84 − 40 = 44; a referred customer 280 − 8 = 272. The show looked busiest and was the weakest channel; the quiet referral card was the best. She keeps the listing and the card, and goes to next year's show only if she can make it a booking day rather than a selling day.
Subscription businesses and online shops track customer acquisition cost and customer lifetime value as a pair, and many aim for a lifetime value several times the acquisition cost, with a payback of a few months. The arithmetic is exactly this lesson's: what a customer cost to win, set beside what they are worth over the time they stay.
Choosing the cheapest per customer. That is the best channel only if its customers are worth the same as everybody else's, and the reason channels differ is that they are not.
Counting inquiries as customers. Forty asked and six bought: the cost per customer is the spend over six.
Counting people who were coming anyway. A channel is judged on the customers it added.
Leaving your own hours out. A weekend on a stall is two days of the owner's time; a channel costed without them is compared on a different basis from one somebody else ran.
Judging a channel on its first month. Regular customers pay back over months; a first-month verdict misses most of their value.
Comparing total spends. A channel you did not use cost nothing and won nobody, and by that measure it beats everything.
Write the spend and the customers won.
$\text{pitch } 240; \ 12 \text{ strangers bought}$
The channel's cost and what it produced.
Divide the spend by them.
$240 \div 12 = 20$
Nothing decided yet.
Set it beside a mug's contribution of 12.
$12 - 20 = -8$
A buyer who takes one mug and goes lost her 8.
Count repeat orders after a month: four buyers ordered again, contributing 24 each.
$4 \times 24 = 96$
The deciding figure took a second month to appear.
Find what the fair earned in total.
$12 \times 12 + 96 - 240 = 0$
Level so far; any further repeat orders are gain.
Write the count as first reported.
$180 \div 45 = 4 \text{ a customer}$
The flattering figure.
Ask each customer how they heard.
$27 \text{ walk past daily and came in anyway}$
They were coming anyway.
Take them out of the count.
$45 - 27 = 18$
The customers the leaflets won.
Divide the spend again.
$180 \div 18 = 10$
Two and a half times the first figure.
Set it beside a lunch's contribution of 3.
$10 \div 3 \approx 3.3 \text{ lunches to pay back}$
A customer must return about four times to cover their cost.
Check the returns after a month.
$\text{half the 18 returned weekly}$
Honest denominator, then the pairing, then the payback.
Write the listing: 150 for three months, 6 customers won.
$150 \div 6 = 25$
Cost per customer.
Write the sign and samples: 60, 30 customers won.
$60 \div 30 = 2$
Twelve times cheaper per customer.
Value a listing customer: 40 a month for 10 months on average.
$40 \times 10 = 400$
A regular's worth over the time they stay.
Value a sample customer: one purchase contributing 3.
$\text{value} = 3$
A one-off buyer's worth.
Find each channel's earnings.
$6 \times 400 - 150 = 2250; \quad 30 \times 3 - 60 = 30$
Customer value decides the channel, not cost per customer.
Find the listing's payback per customer.
$25 \div 40 \approx 0.6 \text{ months}$
Paid back within the first month.
Decide which channels to keep.
$\text{keep the listing; use samples only for one-off lines}$
Each channel judged on what its customers are worth.
Find what one customer cost to win.
$200 \div 25 = 8$
Spend divided by customers won.
Find the contribution of the customers won.
$25 \times 15 = 375$
What they brought in.
Find what the channel earned.
Four channels, and four things a cost per customer cannot see. Match each channel to what its own number leaves out.
| Whether a fair buyer ever comes back, or buys once and is gone | How many of them were already walking past the counter | That one customer books every week for a year and another leaves in a month | Whether the shop keeps sending people once the cards run out | |
|---|---|---|---|---|
| A stall at a craft fair for the weekend | ||||
| Leaflets through the doors of the two nearest office blocks | ||||
| A paid listing on a local services site for three months | ||||
| Discount cards left with the phone shop on the main street |
Complete the worked solution: a channel cost $209$ dollars and $36$ new customers were counted, but $17$ of them said they were coming anyway. Each customer won contributes $27$ dollars. What did one customer cost, and what did the channel earn?
Take out the customers who were coming anyway.
$36 - 17 =$ n
A channel is judged on the customers it added.
Divide the spend by the customers won.
$209 \div (\text{won}) =$ p
What one customer really cost.
Find the contribution of the customers won.
$(\text{won}) \times 27 =$ t
What they brought in.
Take off the channel's cost.
$(\text{contribution}) - 209 =$ e
What the channel earned after paying for itself.
Compare with the flattering figure.
$209 \div 36 < (\text{cost per customer})$
Counting the walkers-by would have made the channel look cheaper.
Bright Home Cleaning spent $150$ dollars on a paid listing on a local services site for three months, and it won $6$ customers. What did one customer cost, in dollars?
Answer:
Neighborhood Kitchen tried two channels last month. Cards handed out at the office park cost $168$ dollars and won $42$ customers, almost all of whom bought one lunch and were never seen again. A listing on a lunch app cost $110$ dollars and won $5$ customers, and those customers order most weeks. Which channel was better?
Neighborhood Kitchen spent $154$ dollars on leaflets and counted $25$ new lunch customers. Asked how they had heard of the kitchen, $11$ said they walk past every day and had meant to try it anyway. What did one customer actually won by the leaflets cost, in dollars?
| Amount | |
|---|---|
| Customers the leaflets actually won | |
| Cost of one customer won, dollars |
A cleaning listing costs Bright Home Cleaning $46$ dollars for each customer it wins. A customer won that way contributes $30$ dollars a month and, on her records, stays for $7$ months on average. How many dollars does one such customer leave the business, once the cost of winning them is taken off?
Answer:
A dog-walking round pays $30$ dollars a month for a listing on a local site. Each customer the listing wins contributes $15$ dollars a month, and the round has room for at most $6$ new customers a month. For how many customers won in a month is the listing worth paying for — leaving the round better off than not having it — and still inside what the round can take?
This task has no paper form; do it on a device.
Lesson test: one question per skill, one attempt each, no hints. Your answers are checked when you submit.
Neighborhood Kitchen tried one way of reaching people: leaflets through the doors of the two nearest office blocks. It cost $180$ dollars and won $45$ customers, and one customer contributes $16$ dollars. Fill in what one customer cost to win, and what the channel earned once it had paid for itself.
| Amount | |
|---|---|
| What the channel cost, dollars | 180 |
| Customers it won | 45 |
| What one customer cost to win, dollars | |
| What one customer contributes, dollars | 16 |
| What the channel earned after its own cost, dollars |
You can divide a channel's spend by the customers it honestly won, read the result beside what one of those customers is worth over time, and say who should never have been in the count. Tell someone why a customer who costs twenty-five dollars can be cheaper than one who costs two. Next: keeping the customers a channel has already won.
15. Your turn: 200 dollars of discount cards, 25 customers, each contributing 15 dollars once, step 3
$375 - 200 = 175$
Provided all twenty-five were won by the cards, which is what to check next.