Smith Jones, Performance Media Buyer

Who I am

I have spent nine years buying performance traffic, and almost all of that time in the unglamorous part of the job. Not the campaign launch, but the fourth day of it: pulling source-level reports, finding the six placements consuming a third of the budget at zero conversions, and cutting them before the week closes.

I came into this the way most media buyers do, from the offer side rather than the media side. Learning why a landing page fails teaches you more about traffic quality than any network dashboard, because it forces you to distinguish between traffic that never arrived, traffic that arrived and bounced, and traffic that was never human to begin with.

I write here because the published material in this niche is mostly produced by people selling the product, and I read it as a buyer long before I wrote a word of it. What came out of that is the reference I wanted when I started: how to buy website traffic without paying for sessions no human ever loaded.

What I actually know

Formats and where each one belongs

Push and in-page push carry short direct-response offers, and in-page reaches iOS because it needs no subscription. Popunder buys a full page load, which suits a landing page that needs room to do its work. Native fits content funnels, banner does retargeting, and video buys attention inside a player. Choosing the wrong one is the most common and most expensive mistake I see from new buyers.

Source-level optimisation

The single skill that separates a profitable campaign from a losing one is knowing what to do with source identifiers, and it is the part of a decision to buy website traffic that nobody sells you upfront. Not just blacklisting: bidding differently on lists that convert differently, so volume survives while cost per conversion falls. Most networks let you blacklist; fewer let you bid per source, and the ones that do are worth paying attention to.

Invalid traffic

I read the bot-traffic research each year because it explains what I see in reports. When a seller advertises residential IP addresses as proof that traffic is genuine, they are describing the exact technique the security literature lists as an evasion method. Knowing that saves a buyer a great deal of money.

Tracking and attribution

A campaign you cannot attribute is a campaign you cannot optimise. I work with server-to-server postbacks, click identifiers and the macro sets that networks expose, because a report that cannot tell you which source produced a conversion is decoration rather than data.

How I evaluate a platform, and what I do not claim

This is where I differ from most people writing in this niche, so I want to be exact about it.

What I verify directly: what a platform documents about its formats, targeting, thresholds and integrations; what the interface exposes without a deposit; what the published policies of major advertising programmes actually say, read in the original rather than in someone's summary; what named industry research reports, with the date attached.

What I report as evidence rather than fact: dated feedback from independent affiliate forums, presented as what those people said and when they said it, including the negative entries about platforms this site earns from.

What I will not do is invent a measurement. If a page says a network processes a certain number of bid requests a day, that number is the company's own claim and the page says so. I will not write that I deposited a specific sum and received a specific result unless that is exactly what happened, because a fabricated figure is the one thing a reader can check and the one thing that would justify never trusting this site again.

Where a live campaign would settle a question that documentation cannot, the page says which question that is. The full process is set out in the editorial policy.

What I look for before spending money

• Whether rates are visible before deposit. A platform that shows bid levels by country and format before taking money is telling you it expects the numbers to survive scrutiny.
• Whether source-level control exists. Blacklists alone leave you choosing between a bad placement and no placement. Per-source bidding gives you a third option.
• What the filters actually remove. Proxy exclusion and IPv6 exclusion are not standard across the industry, and both cut a slice of automated activity before it is billed.
• Where the inventory comes from. Directly contracted publishers and open exchange inventory behave differently, and platforms that mix both rarely publish the split.
• What the complaints say. A single angry post proves nothing. The same complaint from different people across several months is a pattern worth writing down.

Get in touch

If you spot an error in something I wrote, send it over and I will check it against the source. If you work at a platform I have covered and believe the description is unfair, send documentation and I will correct anything I got wrong.

Reach me through the contact page or on LinkedIn. More about the site itself is on the about page.