How to choose a marketing channel when you have one hour a day

Pick two, not five: one that pays this month and one that compounds, scored against your product rather than against what you enjoy doing

The most common early mistake is not choosing the wrong channel. It is choosing five, and underfunding all of them. Two channels done weekly for three months beats five channels done twice, because most channels pay nothing at all until you cross a threshold of consistency, and a channel abandoned before that threshold produces exactly zero rather than a small amount.

Pair a fast channel with a slow one

Channels differ enormously in how long they take to tell you anything.

Fast, days to weeks: community participation, direct outreach, launch platforms, short form video. You post, and you learn something within days. These keep you sane and pay the rent early, but they stop the moment you stop.

Slow, months: search, content, app store optimisation, partnerships. These do nothing for weeks and then keep working while you sleep. Starting only these means three months of silence, which almost nobody survives psychologically.

Run one of each. The fast channel gives you conversations and evidence about who your buyer really is; the slow channel is where that evidence gets invested.

The ten channels and who each actually suits

• Founder led outbound. Direct messages and emails to specific people who have described your problem. Suits B2B, higher price points, and anyone whose buyer is findable by name. Fastest feedback of anything on this list, and the least popular because it involves being ignored personally rather than statistically.
• Community distribution. Answering questions where your buyers already are. Suits developer tools, niche B2B, and anything with an obvious existing forum. Free in money, expensive in judgement, and the failure mode is permanent if you get it wrong.
• Search and landing pages. Suits products with existing demand and competitor names to compete against. Slow, compounding, and the only channel that keeps returning after you stop. Useless for a product nobody knows to search for.
• App store optimisation. Only relevant if you are in a store, in which case it is usually the highest leverage thing available and most developers ignore it.
• Launch platforms. One day of traffic, useful mostly as a forcing function and a source of links. A launch is not a channel; it is an event that a channel can use.
• Creator and short form video. Suits visual products with an obvious before and after. Brutal if your product is a dashboard.
• Artifact led growth. A free tool, calculator or dataset that is easier to share than the product. Excellent when the artifact is genuinely useful, and a large amount of work to find out.
• Partnerships. Suits products with a clear complement. Slow, relationship dependent, and usually premature before you have proof the product works.
• Trust packaging. Security pages, documentation, and public proof. Not a growth channel but a conversion unlock for anything touching business data. Ignoring it silently caps every other channel.
• Paid ads. Suits products with a known conversion rate and a known customer value. Before you have those numbers, paid ads are a way of paying to learn something a hundred conversations would have taught you for free.

Score them in fifteen minutes

Rate each candidate from 1 to 5 on four questions, and be honest rather than hopeful.

1. Are your buyers demonstrably there? Not plausibly, demonstrably. Can you name the specific subreddit, forum, search query or shelf, and paste a link to a conversation from the last month?
2. Can you produce the native format weekly? Each channel has a format, and if you dread it you will stop in week three. This is a real constraint, not a character flaw. Choose the channel you will still be doing in March.
3. How long until the channel tells you something? Shorter is better while you are still learning who your buyer is.
4. Does the effort accumulate? A post that dies in a day is worth less than a page that ranks, at equal effort.

Take the top scorer that is fast and the top scorer that compounds. Write down the seven you are not doing, where you will see the list. That second part matters more than it looks; without it you will quietly restart an abandoned channel in six weeks because somebody on the internet said it worked for them.

How to know it is working before it works

Every channel has a signal that arrives long before revenue. Community gives you replies and profile visits. Outbound gives you response rate. Search gives you impressions in Search Console weeks before clicks. Watch the early signal, not the sale, or you will kill a working channel in week two.

Set a review date twelve weeks out and leave the choice alone until then, unless the early signal is zero. Zero replies from forty good faith comments is information. Three replies and no signups is a message problem, not a channel problem.

Getting this decided without doing it manually

This is exactly what OctoLoops does first. Paste your URL and it reads the site, works out who your buyer is, and scores all ten channels for your product, with reasoning, an expected time to signal, and an explicit list of what to avoid for now. Free, no account, no card, because the decision is worthless if you cannot check it against your own judgement.

Then the loops run the channel you picked. The Community Loop finds three live threads this week, drafts a comment for each, and tracks which ones you posted. The SEO Loop picks three concrete link, on page or content actions. The point is the weekly repetition, since the channel only pays if you are still doing it in month three.

If you have not got as far as choosing yet because nothing is happening at all, the diagnosis in why an app has no users comes first.