Every new site faces the same wall: you need traffic to get traffic. Search does not rank you because you have no authority, social does not distribute you because you have no followers, and paid acquisition is unaffordable when you have no revenue.
Here are nine channels that work before you have an audience, ordered by the ratio of effort to result based on how they behaved for sites that actually used them.
1. Answer questions where they are already being asked
The highest-yield channel for most new sites, and the one people abandon too early.
Find the twenty questions your product answers and locate where they are currently being asked. That is usually a handful of forums and one or two Q&A communities, not the open web.
The mechanics matter:
- Answer the question completely, in the thread, without a link. If your answer is not useful without the link, the link will be removed and the answer will deserve it.
- Link only when it adds something the answer cannot - a calculator, a template, a dataset.
- Disclose. "I built this, and I am biased" converts better than pretending otherwise, and it survives moderation.
- Stay with the thread. Replying to follow-ups for a week is what turns one question into a referral source.
Realistic outcome: eight to twenty answered questions in the first month produce two to six hundred visits, plus a handful of permanent links that keep sending traffic for years.
2. Build one genuinely useful free tool
A calculator, a converter, a generator, a checker - something that produces a result on the same page. Tool pages accumulate links passively because people cite tools rather than articles.
The constraint is that it has to be the best thing available for the query, not a lead magnet wearing a tool's clothing. A gated calculator produces nothing. An ungated one that answers the question completely gets shared.
Time cost: one to three days. Typical twelve-month result: the highest-traffic page on the site.
3. Publish on a platform with existing distribution
Medium, dev.to, Hacker News, Reddit, Substack, LinkedIn - take one article and adapt it for a platform that already has readers.
Adapting means: change the length, change the framing, remove anything that only makes sense on your own site. A reposted identical article underperforms a native version by a wide margin.
A worked example from an engineering blog: a single post adapted for Hacker News produced 14,000 visits in 48 hours, 340 newsletter subscribers and eleven inbound links from other engineering blogs. The same article on the site's own domain produced 90 visits in the same period.
4. Get listed in directories - selectively
There are roughly four thousand general directories and a few hundred that still matter. The difference is measurable in the most direct way possible: whether they send referral traffic.
The screening process:
- Submit to directories in your specific niche first, general ones later.
- Wait thirty days, then check referral traffic per directory.
- Keep the ones that sent anyone at all. Delete the bookmarks for the rest.
Expect three to eight percent of submissions to send meaningful traffic. The value of the others is mostly an early link profile, which is worth having but not worth months of effort.
5. Join a listing platform or toplist
Toplists and promotional platforms solve the cold-start problem directly: they have readers who are already looking for what you have built.
The mechanics are worth understanding before you commit:
- Your listing is a permanent, indexable page describing your site with a link to it.
- A promotional link lets you earn placement without spending money.
- Clicks are deduplicated per visitor per day, so the numbers you are shown are close to real people.
The compounding property is what makes this different from paid ads. Your listing accumulates link equity and referral traffic while it sits there, and a well-written entry keeps converting months later. Budget one hour: write the entry properly rather than filling it in as quickly as possible.
6. Newsletter swaps with peers
Find ten sites of comparable size in adjacent niches and propose a swap: they mention you to their list, you mention them to yours.
The response rate is far higher than cold outreach in any other channel, because the exchange is symmetrical and costs neither party money. Specificity is what gets a yes - "I have 1,200 subscribers in indie game dev, you have 900 in game design tools, shall we each do a short section next month?" beats a vague proposal every time.
At 1,000 subscribers and a two percent click rate, one swap is twenty visits. Six swaps per quarter is a steady stream, and it starts working in week one.
7. Be genuinely useful in a community for ninety days
Pick one community - a Slack, a Discord, a subreddit - and become the person who answers questions well, with no promotion at all.
It takes three months to stop being the new account. After that, the referral traffic arrives without asking, because the community trusts the source. This is slower than everything above and it produces the highest-intent traffic of any channel on this list.
8. Take a public position
Publish a comparison, a benchmark, a teardown or a survey - something where you did the work and stated a conclusion others have to respond to.
Original data is the only content type that reliably attracts links without an existing audience to amplify it. A survey of eighty practitioners is enough. The finding needs to be specific enough to be quoted: not "teams struggle with onboarding" but "teams with more than four onboarding steps lost before activation convert at 40 percent of the rate of those with two."
9. Ship something worth mentioning
The channel that underlies all the others. Nothing on this list works for long if the destination is not good.
The narrow version of this: make one aspect of your product noticeably better than every alternative, and say clearly what it is. "The only tool that exports to [format]" is mentionable. "A better all-in-one platform" is not.
What the first month actually looks like
| Channel | Effort | Visits by day 30 |
|---|---|---|
| Answering questions | 6 hours | 200-600 |
| One free tool | 2 days | 100-400 (rising) |
| Platform adaptation | 4 hours | 0-14,000 (spiky) |
| Directory submissions | 3 hours | 20-100 |
| Toplist listing | 1 hour | 150-500 |
| Newsletter swaps | 2 hours | 40-120 |
Sum the middle column honestly and the total is roughly twelve days of work for somewhere between five hundred and a few thousand visits. That is the real cost of a first thousand visitors, and it is much cheaper than the paid alternative - at ten cents per click, 1,000 visits is a hundred dollars and it stops the moment you stop paying.
What to do on day thirty-one
Look at which pages converted. Not which pages got traffic - which pages turned visitors into subscribers or accounts. Then do more of the channel that fed those pages, and drop everything else.
Most people do the opposite: they chase whichever channel produced the biggest spike, regardless of whether the arrivals did anything once they landed.