What's the cheapest way to get traffic to a website?

Search for cheap website traffic and you land in two markets that share a name. One sells visitors by the thousand for the price of a sandwich; the other is the slow, unglamorous set of channels that send real people for close to nothing. Price every source per visitor and the market gets legible fast: paid search runs roughly $5 and up a visitor, social somewhere between fifty cents and two dollars, sponsorships in between, and a handful of channels genuinely approach zero. But "cheapest" hides a second variable — some cheap traffic is worth nothing at all — so this list prices both dollars and quality. (Disclosure: published by BidSurvivor, one of the near-zero rows, so weigh that entry accordingly.)

How to price a channel per visitor

The arithmetic is the same for every row: what it cost, divided by the visitors it sent (the ad budget calculator does this for any channel). The trap is forgetting that time is a cost. If an afternoon writing an article is four hours and your hour is worth $50, that article cost $200 before it earned a single click. If it goes on to send thirty visitors a month for two years, that is 720 visitors and a cost of about 28 cents each — cheap, but only in hindsight, and only if it ranks. Ahrefs' study of roughly fourteen billion pages found that 96.55% of them get no traffic from Google at all, so the honest per-visitor price of SEO is the price of the articles that worked plus the price of the ones that did not.

Paid channels are easier to price and harder to make cheap. Google's own all-industry benchmark puts the average search click around $5.42 in 2026 (WordStream); $100 is eighteen visitors, before you have learnt anything. That is expensive per visitor and, as we will get to, sometimes the best deal on the list.

Near-zero cost per visitor

SEO. The famous one: expensive in time, near-free per visitor once it works, and the visitors arrive pre-qualified by their own search. The catch is the delay — months, not days — and the failure rate in the Ahrefs number above. Start now, expect nothing this quarter, and write for the question a person actually types rather than the phrase you wish they typed. Search sends fewer clicks than it used to, and where small sites find visitors after AI Overviews is its own question.

Communities. A genuinely useful post in the right subreddit or forum sends dozens to thousands of visitors at $0, with intent attached, because the reader was already asking the question you answered. It does not scale on demand and it punishes anything that smells like a campaign. One honest post a week beats seven mentions a day.

Launch spikes. One-time bursts of hundreds to thousands, free, from launch platforms. High curiosity, mixed intent — good for finding out whether anything on your page converts, useless for finding out whether anyone will come back. Smaller platforms are often kinder to a first outing than the biggest one; the comparison is here.

A free auction slot — ours. BidSurvivor sells two twelve-hour slots a day on one page. An empty slot is free to the first brand that claims it, no account or card needed; a held slot can be taken from a $1.00 floor, never above $5,000, and every visitor starts with $100 of house credit before signing up, so bids inside that credit cost nothing out of pocket. Each slot's card opens and click-throughs are published on the front page and on the brand's own page. Cost per visitor at current clearing prices: $0. Volume: honestly small, and publicly so — treat it as free measured traffic, the kind you use to test a pitch line before buying volume elsewhere. One rule to know before you rely on it: if somebody outbids you, you lose the slot and the money, with no refund, though a bid that fails to take a slot costs nothing.

Cheap per click, expensive per truth

Discount traffic sources — low-CPC display networks, traffic-exchange schemes, "10,000 visitors for $5" packages — are the worthless-cheap quadrant. The economics explain themselves: real attention costs money to produce, so anything priced far below the market is either bots or people paid to click, and both poison your analytics into the bargain. Juniper Research expects advertisers to lose roughly $100 billion to fraud in 2026 (Juniper); the discount tier is where a lot of it lives.

Broad social at low bids sits one notch up: real humans, weak intent, a feed they were scrolling for a different reason. Fine for retargeting fuel, poor for direct response. Banner-style placements sit lower still; the Nielsen Norman Group has documented banner blindness since 1997, and no creative brief has cured it.

A test that separates cheap from worthless

Run this before you believe any per-visitor price, including ours.

1. Pick one action that a real visitor would take and a bot would not — scroll past the fold, click the pricing tab, start the form. One event, not a dashboard.
2. Get a baseline from a source you trust to be human: twenty visitors from a community post or from people you emailed. Note the share who did the action.
3. Run the cheap source and compare. Visitors who never scroll, never click and leave inside two seconds are not traffic, whatever the invoice says.
4. Price per action, not per visit. A $5 search click that converts one time in twenty is $100 per action; a "free" visitor who converts one time in five hundred, at ten minutes of your time per hundred visitors, is not free either.

Step four is where the counter-argument to this whole article lives. Expensive traffic is often the cheapest per outcome, because intent is the thing you are paying for and intent is what converts. What $200 buys on Google Ads is a small number of visitors and a large amount of information per visitor; a free launch spike is the reverse.

The honest ranking, per useful visitor

1. SEO (eventually) and communities (now) — free and qualified.
2. Launch spikes — free, one-time, curious rather than committed.
3. Free slot tests — $0, small, fully measured, cold.
4. Newsletter sponsorships — a dollar or so per genuinely warm visitor (the $50 version).
5. Paid search — expensive, but the intent ceiling of the whole list.
6. Discount traffic — free at any price is still overpriced.

Cheapest and best are not the same list, but they overlap at the top, which is the actual answer: the cheapest traffic worth having comes from being findable and useful, and the paid channels exist to amplify what those already proved. The full zero-budget sequence is in marketing with no budget at all, and the arithmetic for the moment a budget appears is in how much it costs to advertise a startup.