Case: two channels for Dana

A one-person business usually has almost no money for marketing. What it spends instead is hours, and hours are just as limited: every evening on posts is an evening not spent teaching or resting.

Dana, a Klasna English tutor, started her own practice for adults in September. Her first three students came from friends. Since then she has tried everything at once: posts, short videos, a messenger channel, asking for introductions, a talk in a community chat. In about eight weeks that took 35 hours. From November she can give marketing 3 hours a week, 24 hours to the end of December.

So the same channel thinking applies as at Klasna, with one swap: where Klasna divides money by families still paying, Dana divides hours by students who stay. A channel that eats evenings and keeps no one is expensive, even if it costs nothing.

One rule frames every choice: her contract with Klasna says she never approaches Klasna's families or learners.

Dana's log from 1 September to 25 October: posts 14 hours, short videos 10, messenger channel 4, asking friends and students for introductions 2, a talk in a job-hunting community chat 5. Total 35 hours, 5 enquiries, 2 students who started and are still studying.

The average across all channels: 35 ÷ 2 = 17.5 hours per student still studying.

Is that good enough? She has room for three more students: five take about 30 of her 48 lessons a month. At 3 marketing hours a week, 24 hours to the end of December, the average would bring little more than one new student. So the average is not good enough.

But an average hides how different the channels are. Some of those 35 hours brought students quickly; some brought none. The average is where the question starts: the next step is to split it by channel and see which hours did the work.

Divide each channel's hours by the students it brought who are still studying:

• Referrals (asking friends and her students for introductions): 2 hours, 1 student → 2 hours per student.
• The community talk, agreed with the chat's admin: 5 hours, 1 student → 5 hours.
• Posts: 14 hours, 2 enquiries, no student. Short videos: 10 hours, nothing. Messenger channel: 4 hours, nothing. None of these has a finite figure at all: dividing by zero students has no answer.

Dana felt her posts were the most productive: two enquiries, as many as referrals. But hours per enquiry is the wrong measure, for the same reason cost per enquiry misled Klasna: the posts' two enquiries never became students (one couldn't fit her evenings, one wanted a group).

So the ranking, fewest hours first: referrals, then the community talk; posts, videos and the channel kept no one in these eight weeks.

Ranking channels is not the end. Each one gets a role that fits what it can do.

• Referrals are the cheapest, but the pool is small: five students and a few friends, and she can't ask them every week. So ask each student once, with a short note they can forward.
• Community talks are the channel that can grow: there are other relocation and job-hunting communities, and a talk can be reused. This is where most of the 24 hours should go.
• The profile stays as a shop window, not a channel: a clear profile line and a pinned post, about an hour a month. People others sent her look at it before they write, so it serves "compare and check".
• The messenger channel is a place to follow up with people who chose to subscribe, not a way to find new ones.
• Short videos stop for now.

Two channels get the hours; the others either support them cheaply or wait.

Not every channel idea is open to Dana. Two rules sort them.

Her contract with Klasna: she never approaches Klasna's families or learners. That rules out offering private lessons to the adults she teaches at Klasna, and posting her offer in a Klasna parents' chat she belongs to. Being in a chat doesn't make its members her audience; approaching them indirectly is still approaching them.

Consent: people join her messenger channel because they chose to. Adding everyone from a community chat without asking breaks that, and it also burns her standing with the community whose trust she was borrowing.

What stays open: asking each of her own private students once for an introduction, and giving a talk in another job-hunting community, agreed with its admin first. Both use trust she has earned, in the open.

A channel plan that breaks a contract or ignores consent is not cheaper; it moves the cost to her reputation and her job.

The trap in Dana's log is the same one Klasna met with the marketplace, at a smaller size. Her posts brought two enquiries, as many as referrals. Counted by enquiries, they look equally good.

Counted by what it took and what stayed, they are not. The posts took 14 hours for those two enquiries; referrals took 2. And neither person from the posts started lessons: one couldn't fit her evenings, one wanted a group. Referrals brought one student who is still studying.

So the measure that matters is students who stay per hour, just as Klasna judges channels by families still paying after 90 days rather than by enquiries. An enquiry is a step on the path, not the result.

The reasons the post enquiries didn't start are useful too: they say something about fit (evening times, groups), which may matter for a future offer. But they don't make the posts a good channel for the next eight weeks.

"Everyone says short videos are how you grow." In Dana's log, videos took 10 hours and brought no one. Both statements can be true at once.

What her log shows is a fact about her eight weeks: for her, with her niche (adults in IT who need English for work calls), at her current skill with videos, ten hours bought nothing. It is not a law that videos don't work for tutors, and it isn't proof they would never work for her.

That difference changes the decision. Because it is a fact about recent weeks, the right move is to stop videos for now and give the hours to channels that kept students. Because it is not a law, the right move is also to look again in spring, perhaps with a clearer idea of what a useful video for her niche would show.

The same caution applies to the winners: 2 and 5 hours per student come from one student each. Small numbers support a choice for the next eight weeks, not a rule forever.

Dana hesitates to stop anything: "I hate wasting what I already made." The 10 hours of videos and 14 of posts are spent whatever she chooses now. Keeping a channel because of past effort only adds new hours to the old ones. Economists call these sunk costs; the practical rule is simpler: decide on the next 24 hours only.

Two habits make the decision stick:

• Write down what stops and why, using the log: "videos: 10 hours, no one; posts: 14 hours, no student." In a busy week it is easy to drift back to what feels productive.
• Pick one number to watch with a rule, so in four weeks she knows whether it works, for example new students who start per hour of talks, and a line for what she does if it stays at zero.

How it fits together: when the budget is hours, judge channels by hours per student who stays, give each channel a role, keep within contract and consent, and decide forward, not backward. The next module, back at Klasna, asks what to say in the channels and with what budget.