Buy Website Traffic? What a $0.001 Click Is Worth

Type "buy website traffic" into a search box and the first price you see reads like a typo.

$0.001 a click.

That's not a rounding error. PropellerAds prices push notification traffic at a $0.001 minimum CPC and a $0.01 minimum CPM, on the network's own blog. A hundredth of what a single Google Search click costs on an average day.

So here's the answer to the question you typed. The cheapest website traffic you can buy right now doesn't come from Meta or Google.

It comes from networks that sell push notifications and pop-unders instead. Some sell native placements, ads dressed up as an article recommendation, which cost a little more but read as content instead of an interruption.

"Cheap" here measures one thing only: the price of the click.

It says nothing about who clicked.

Every list ranking for this search hands you the same five or six network names and stops there. One 2026 roundup leads with RichAds and PropellerAds, and recommends MGID right behind them, because clicks on these networks start at $0.001 to $0.01, which it calls ideal for a small business testing a tight budget.

None of those lists price what happens after you spend the money the networks themselves recommend to find out.

Here's the honest version: what buying traffic from these networks actually costs right now, and what a $0.001 click is actually worth once you count who's on the other end of it.

Then the two-hour setup that tells you within 48 hours whether your money reached a person.

What Bought Traffic Actually Is

Push and pop ads aren't the same thing as native ads, and paying for one doesn't buy the same attention as paying for the other.

Push notification and popunder ads pay to interrupt. A fake system alert, or a tab that opens behind the one you're already reading.

Native ads work differently. They pay to blend in at the bottom of a news page, formatted to look like the article above it.

Read the fourth column, not the first.

The headline minimum ($100, twice over) is what gets your card charged today. It isn't what the network privately expects a real test to cost.

MGID's own help page states the $100 minimum outright, then recommends $650 or more, in its own words, "to generate enough data for testing." That gap between the sign-up price and the honest price sits on the same page, printed by the company selling both numbers.

"Tier 1" in that table means the countries every network prices highest: the US, the UK, Canada, Australia. Tier 2 and 3 cover everywhere else, at a fraction of the rate, which is the first clue that a lot of "cheap" traffic is cheap because of where the click came from, not because of how well it's targeted.

There's a reason the same handful of network names show up in almost every roundup, and it isn't coincidence. MGID runs a partner program that pays affiliate sites a 25% bonus when a referred advertiser deposits $1,000 to $5,000. The blog recommending MGID to you may earn more the more you deposit, which is exactly backward from the advice a small budget needs.

The Number the Cheap Lists Don't Print

Buy website traffic at $0.001 a click and it sounds like two thousand visitors for two dollars.

Run the math exactly as PropellerAds runs it. A $10 test budget at a $0.005 CPC buys 2,000 visitors, by the network's own example.

Two thousand visitors for ten dollars is real. What it doesn't say is how many of those two thousand were a person deciding whether to buy something, against a script clicking through a push queue on a device nobody's holding.

Two independent 2026 reports measured that gap, and their numbers don't agree with each other.

Lunio's Global Invalid Traffic Report, published January 20, 2026 off an analysis of 2.7 billion paid clicks, put the average invalid traffic rate across Google, Meta, TikTok, LinkedIn, X and Bing at 8.51%. Spider AF's first-half 2026 fraud report, built from 4.006 billion measured clicks, put the narrower ad fraud rate at 5.58%.

The two disagree because they're measuring different things. Lunio counts invalid traffic broadly, bots and accidental non-human hits, whether anyone meant to fake them or not. Spider AF counts deliberate fraud specifically, a smaller and harder-to-prove slice of the same problem.

For a small budget, trust the broader number. A click your analytics never converts costs the same whether the bot behind it was malicious or just badly built.

Lunio's own published math shows why the rate matters more than it sounds like it should. Of 50,000 measured clicks at that 8.51% average, 4,255 are invalid, leaving 45,745 genuine ones. The true cost per valid click swings by more than 9% once the fake ones are subtracted out.

Run that same subtraction on your own push or pop test. At 2,000 clicks and even a conservative 15% junk rate, three hundred of them never happened at all, and your $10 test just became a $10 test of 1,700 real people instead of 2,000.

Neither report covers PropellerAds. Neither one covers MGID or RichAds either. Both measure platforms with a public ad exchange, the same six named above.

Business of Apps' 2026 breakdown of the popunder format names the risk instead of publishing a rate for it: "masked traffic and geo mismatch," the two most common ways a popunder click gets faked. A competing network, Adcash, markets anti-fraud filtering as a headline feature, and claims by its own account that the feature saved advertisers $32.8 million in 2025. Nobody builds and sells a fix for a problem that isn't happening.

None of this makes the traffic worthless.

It means the price on the sign-up page prices attention. The customer is a separate question the sign-up page never answers.

A Fifty-Dollar Test, Priced Honestly

Take MGID's own recommended entry point: a $650 deposit, spent at roughly $0.12 a click, to reach the volume MGID's own bidding tool needs before there's anything to optimize.

That's real money for a business testing a channel for the first time, and MGID says so itself.

Cut it down to a number a small business can actually afford to lose finding out. Fifty dollars, at that same $0.12 rate, buys around 415 clicks. Not enough for MGID's own tool to kick in.

It's plenty to answer a narrower question. Does anyone on the other end of that click stay on the page.

Cheap advertising for a small business is a different purchase, and that post already prices the same fifty dollars through a Nextdoor evergreen ad instead, at roughly two to four dollars a click. That buys somewhere between twelve and twenty-five clicks, from an audience that lives near your storefront and clicked because a neighbor's app showed them something local.

Compare that to a push click landing on a phone three time zones away, mid-game, with no idea what your storefront even sells.

Neither number is wrong. They answer different questions.

The native network answers whether the creative gets clicked at all. The Nextdoor test answers whether a real, local person responds to it. A small business chasing the second question with the first channel's traffic reads a full inbox of clicks and an empty one of calls, and blames the offer instead of the source.

The Two-Hour Setup That Keeps You Honest

Run this before the money leaves your card, not after.

1. Open the account at the network's stated minimum: $100 for MGID or PropellerAds. Match the format to your product. Native works for a considered purchase; push or pop suits an impulse one better.
2. Set a hard daily spend cap at $10 to $15. Ignore the platform's "recommended" starting deposit until step 6 tells you to raise it.
3. Build one landing page URL with a UTM source unique to that network. Its traffic then lands as its own row in your analytics, separate from everything else you run.
4. Let the campaign run for 48 hours, or until it clears 200 clicks, whichever comes first. Don't touch the bid inside that window.
5. Open the referral report for that UTM source. Read two numbers: average session duration and bounce rate.
6. A passing result looks ordinary: something like a 35 to 60 second average session and a bounce rate in the 50s or 60s, the normal range for a stranger clicking an unfamiliar ad. A failing one looks nothing like that: sessions under 5 seconds and bounce rates past 90%. Clicks arriving in tight bursts within the same minute, rather than spread across the day, are a second red flag on top of the first. That's a pattern, not a slow week. Kill the campaign and keep the receipt.
7. If the numbers look like real visitors, raise the daily cap toward the network's own recommended figure in stages, not in one jump.

Don't count on getting the ten dollars back even if the network agrees the traffic was fake. Improvado's 2026 ad fraud guide puts the realistic refund approval rate at 30% to 50%, even with strong evidence attached to the claim.

That's the real reason step six matters more than a support ticket. Catching bad traffic inside 48 hours costs you the original ten dollars. Fighting for a refund after the fact costs you the ten dollars and the week you spent writing the complaint.

You paid ten or fifteen dollars to learn what the network's own recommended deposit would have cost $650 to find out.

That gap is the entire argument for testing small before testing at the number the platform suggests.

Where a Dollar Buys Something Verifiable

Every channel above prices a click and asks you to trust what's behind it.

The Board prices something else: a listing, sorted only by how much has been paid for it, with every placement disclosed as paid. A spot starts at a dollar. Taking the top one costs a dollar more than whoever holds it now, in the open, with no bid to place and no exchange deciding which bot filter ran on your impression this week.

That's not a replacement for testing a native network or running a Nextdoor ad. It answers a narrower question than either one.

Where a dollar spent is verifiably a dollar spent, because there's no click being billed in between the payment and the placement.

The rest of what a small budget can test this month, priced the same honest way channel by channel, sits on the advertising shelf.

What to Open This Afternoon

Skip the network with the practical multi-thousand-dollar floor if fifty dollars is what you actually have to lose this month. Where to advertise a startup already prices that tier, and it's a different budget conversation than this one.

Open MGID or PropellerAds at the stated $100 minimum. Cap it at $10 a day.

Give it 48 hours before you touch the bid.

Then read the bounce rate before you read the click count. The click count is the number the network wants you to see first.

The bounce rate is the one that tells you whether you bought website traffic, or rented a script for an afternoon.