Which marketing channels actually work for indie hackers?

Indie hacker marketing runs under two constraints that reshape all the standard advice: one person's hours, and a budget that competes with rent. Under those constraints, channels don't just have costs — they have shapes: some need daily feeding, some compound unattended. This ranking weights shape as heavily as payoff. (Disclosure: published by BidSurvivor, which appears in the low-effort tier, so weigh that entry accordingly.)

Compounding, low-maintenance — the indie sweet spot

SEO. The consensus indie channel for a reason: write once, rank for years, no algorithm to feed. The tax is patience. One useful article a week is a complete strategy — with a caveat from this year that gets its own section below.

Directories and integrations. Listings (the free ones) and marketplace presence (Chrome store, Slack directory, plugin registries) sit there working. An evening each, returns that keep arriving.

Programmatic pages — comparison pages, template galleries, calculators. Effort once, long-tail search for as long as the long tail exists. The indie-scale version of a content team.

Spiky, occasional — schedule them

Launches (PH and the gentler alternatives) — a spike per polish cycle; how to stage one. Show HN — the biggest spike available to a technical product with an honest story. "Building in public" posts — the one social behaviour with indie-sized returns, because the artefact (a revenue graph, a lesson) does the work, not the posting cadence.

Continuous, human — budget the hours honestly

Community participation produces the stickiest users and eats the most calendar. Cold outreach works for B2B indie products at ~10 emails a day — see how we disclose our own seeded placements for how we do it with consent and honesty baked in.

Paid, indie-sized

Most paid channels aren't — the minimum useful budgets exceed indie testing money. The exceptions: newsletter one-offs from ~$50, and free-to-$1 auction slots — ours: BidSurvivor sells two twelve-hour slots a day, first claim free, $100 house credit for every visitor, account or not, every result published on the front page and the brand's own page. Zero-dollar message testing with published numbers; small board, and the numbers say so.

What changed this year: evergreen is losing

The SEO entry above used to be the safe one. It is now the one with an asterisk. Search Engine Land's analysis of AI Overviews put search clicks down 42% from baseline, with click-through rates falling around 61% on queries where an Overview appears, and publishers reporting that evergreen explainers are the content losing fastest while real-time pieces and Discover traffic hold up. Two adjustments for an indie hacker follow from that.

First, the programmatic-page strategy still works, but for queries with a specific answer — a comparison, a price, a calculator result — rather than for the "what is X" explainer that an Overview now answers in place. Second, "write once, rank for years" has become "write once, be the source the Overview cites", which rewards plain claims, checkable numbers and a page that is easy for a machine to read. Evergreen content is losing goes through what is winning instead.

The two-channel rule

Every indie playbook published this year converges on the same shape — build in public, community listening, SEO, intent outreach — and on the same warning: pick two, not seven. The buildinpublic.so playbook puts it as a constraint on attention rather than money: a channel you run at half attention produces a quarter of the result, because the half you skipped was the replying, the following up and the measuring.

The practical version. Pick one compounding channel and one human channel. Run them for a quarter. Change nothing else. If one of them is visibly working by the end, that is the channel your next quarter doubles down on; if neither is, swap one, not both. The two-channel rule is the longer argument, including the case against the "try everything for a month" approach that most makers start with and most regret.

A worked week, five hours

For a maker with five marketing hours a week, the portfolio below is the one we'd actually run.

Two hours — one piece of writing that answers a question your users ask in their own words. Specific, with a number in it, easy to cite. One hour — the community you live in. Answer the threads where your problem comes up; mention the product only when it is the answer. One hour — measurement. Search impressions, referral visits, sign-ups by source, in one spreadsheet, once a week. One hour — a paid test, most weeks $0. An empty board slot, or every couple of months a $50 newsletter placement, with the click-throughs written next to the writing's numbers so you can see which sent more.

Nothing on that list needs a posting schedule, and the only line that needs money is the one you can run for free until it stops being free.

The honest portfolio

One compounding channel (SEO, fed weekly and written for the queries an Overview can't answer), one spike on the calendar per quarter (launch or Show HN), one community lived in genuinely, and paid tests only from the indie-sized menu. Four lines, one person, no treadmill — the whole system in marketing without social media if you want the treadmill formally cancelled.