"I want to make $3,600 a month" is a clear goal and an empty instruction.
It is empty because the same $3,600 comes from selling a $35 subscription to 103 people, or a $2,084 service to two clients. Those are not two versions of one job. One is a traffic-and-automation problem. The other is a sales-and-trust problem. You will be good at one of them before you are good at the other.
This playbook is the arithmetic that turns a target into a decision, and the checks a Korea-based operator has to clear before selling at either end.
The arithmetic
There is only one equation, and it has no opinion about you.
Step 1. Cheap and many
103 customersat $35 / month to reach $3,600
Self-serve signup, no sales calls. You need people to find you and buy without talking to you — which means traffic, onboarding, and support that runs without you.
Step 2. Mid
10 customersat $365 / month
Somewhere between. Usually a demo or a trial, then a decision. You can hold ten relationships in your head; you cannot hold a hundred.
Step 3. Expensive and few
2 clientsat $2,084 / month
Every sale is a conversation, a proposal, and a reference. Losing one client is losing half the business. You need trust before you need traffic.
Nothing above is a projection. It is division. target ÷ price = customers, and
the only reason to write it out is that people skip it — they set the target,
skip the split, and then work on whichever task is nearest.
Three businesses, actually verified
Here are three AI businesses whose revenue we could confirm through a third party, placed on that same map.
These four are not solo side businesses. Checking each official site (2026-08-14; FrameLiq on 08-15): two are companies with staff and investors, and **two have unverified operators.** They are used here only to show the distribution of price points — not as things to copy. Both unverified ones are **listed for sale** ($12,000 and $17,000), which says something about what ends up on a list like this.
| Case | Computed (MRR÷subs) | Published price | Customers |
|---|---|---|---|
| FrameLiq | $19 | $19 / $49 | 12 |
| UGC Copilot | $35 | $29 / $79 / $149 | 9 |
| RankAI | $365 | not published | 155 |
| AEO Engine | $2,084 | $1,597 / $2,997 | 38 |
Read the dashed line first: every point on it produces the same monthly revenue. Then look at how far apart the three businesses sit along it. UGC Copilot and AEO Engine are both real, both AI, both selling subscriptions — and one charges roughly sixty times what the other does.
That spread is the reason a revenue goal is not a plan.
What each end actually demands
Pick a target
The easy part. Everyone gets here.
Divide by a price
Now you have a customer count, which is a real thing you can count.
Look at that count
Two clients and a hundred subscribers are different companies.
Pick the job you can do
Traffic and automation, or sales and trust.
Flow: Pick a target, then Divide by a price, then Look at that count, then Pick the job you can do
✓ Low price, many customers
- You need distribution before you need a product tier
- Support has to survive volume — templates, docs, self-serve
- One unhappy customer costs you little; churn is a rate, not an event
- Pricing changes are cheap to test
✕ High price, few customers
- You need references and a track record before anyone signs
- Every customer expects to reach a person — you
- Losing one client can halve revenue; concentration is the risk
- Pricing is negotiated, so each deal takes real time
Neither column is better. The left column is a worse fit if you hate writing and marketing. The right column is a worse fit if you cannot yet point at work you have done for someone else.
If you are selling from Korea
The band you pick changes what you have to set up. This is the part that stops people after they have already built the product.
| What to check | Low price, many customers | High price, few clients |
|---|---|---|
| Who can buy | Anyone with a card, worldwide | Usually businesses that need an invoice and a contract |
| Taking payment | A Merchant of Record (Lemon Squeezy, Paddle) collects and remits sales tax in each jurisdiction for you | Direct bank transfer or invoicing is common; the MoR route often does not fit contract work |
| Sales tax / VAT | The MoR handles registration and filing across jurisdictions. Without one, the obligation is yours in every country you sell into | Depends on the client's country and whether the service is B2B; usually invoice-level, not automated |
| Getting paid out | MoR pays out to a bank account on a schedule; PayPal in some regions | Wire transfer, per invoice |
| Minimum setup | A business registration is normally required to open a payment account, plus identity verification | Same, plus the ability to issue an invoice the client's accounting will accept |
| Ongoing work | Mostly automated once configured | Per-client paperwork every time |
We are not giving tax advice, and this table is not a substitute for it. Whether a given sale is zero-rated for Korean VAT, and what records you have to keep, depends on details we cannot see from here. Take the table as the list of questions to bring to an accountant, not the answers.
The one durable point: at the low-price end, a Merchant of Record removes an obligation you genuinely cannot meet alone — collecting and remitting sales tax in dozens of countries. That is why it is worth its fee, and why the fee is not the number to optimise first.
Do this today
Step 1. Write your target as a division
1 linetarget ÷ price = customers
Not a range. One price, one customer count. If the count surprises you, that is the point of doing it.
Step 2. Say the customer count out loud
“I need 103 paying customers” and “I need 2 clients” are sentences you react to. The revenue figure is not — that is why it feels comfortable.
Step 3. Name the skill that band needs
Traffic and automation, or sales and trust. Then ask which one you have evidence of — work you have already done, not work you intend to do.
Step 4. Check the payment path before you build
Find out what taking money in that band requires from where you live. People discover this after the product is finished, and it changes the product.
What we do not know
We do not know the conversion rates behind any of the three businesses on the map — none of them publish those. So this piece cannot tell you how much traffic 103 subscribers costs, or how many conversations produce two clients. Those are the numbers that decide whether a band is reachable for you, and we do not have them yet.
We also do not claim any of these numbers are reachable by you. Two of the three businesses have staff and outside investors; the third has an unverified operator. They are on the map to show the spread of price points, and for nothing else.
What the arithmetic does say is narrower and still useful: a revenue target you have not divided is a target you have not chosen.