I Ran PropellerAds This Way, and the Results Surprised Me

Lately, the first question many people ask after adding me on WeChat is: how do I get PropellerAds campaigns to run profitably?

They bought a tracker, they’re using a landing page, they added back-button traffic, and CTR looks fine. But ROI still hovers around -30%.

My answer might disappoint you: direct link. No tracker, no lander. Sometimes it actually works better.

I wrote earlier about a real test where I gave OpenClaw $100 to run ads. Those campaigns were direct-linked on PropellerAds.

What I did on PropellerAds

I use the most basic approach there is:

Direct-link the offer. No landing page, no multi-hop redirects.

Later I dropped the tracker altogether and now make decisions straight from PropellerAds’ own dashboard data, combined with data analysis through the MCP interface: which zones convert, which GEOs perform, and when to auto-pause or restart campaigns.

Many people subconsciously assume you can’t run without a tracker. But look at it from the other side: adding a tracker adds one more redirect hop, and you may lose some converting traffic along the way. And most of what you look at is zone, GEO and time of day anyway.

So why not just use the data the traffic source already gives you?

Of course, this doesn’t mean trackers are useless. With big budgets, multiple traffic sources and complex funnels, you need a tracker. But if you’re in the early testing phase, with a small budget and a single traffic source, direct linking can be an efficient way to get started.

One less step is one less place for things to go wrong. Every redirect you remove buys back a little conversion rate.

Spend tens of thousands and lose 80%: the problem isn’t the offer

A few days ago, a reader of my WeChat account got in touch.

He said he’d been testing mVAS offers in Nigeria (NG) on another traffic source. He tested for a while, spent tens of thousands in total, and his ROI stayed between -70% and -80%.

I asked how he was running it. He said he used an age lander plus a tracker, and put all his budget into NG.

That’s the problem. That might have worked a few years ago.

The mVAS vertical has changed a lot over the past two years. Lately most of the people getting volume have been running it via PIN API, while you’re still on a single network, a single offer, a single lander, on a market as competitive as NG. You picked a tough fight from the start without preparing for it.

It’s not that the offer doesn’t work. It’s that the methodology is outdated.

This has nothing to do with tools. You can use the most expensive tracker, the best server and a proven lander, but if the buying logic itself is behind the times, more budget only means validating an outdated assumption more expensively.

Don’t be held hostage by the “standard process”

There’s a standard process everyone in the industry agrees on:

Pick an offer → set up a tracker → build a lander → launch → check the data → optimize → scale.

Nothing wrong with the process. The problem is that many people treat it as the only correct way.

As if not adding a tracker makes you unprofessional, and not building a lander means wasting traffic.

I’ve been running ads for ten years, and here’s what experience tells me: the value of a methodology isn’t in how complete it is, but in whether you’ve verified it yourself.

People say direct links don’t work. Have you tested it?

People say you must add a lander. Have you compared the data with and without one?

There’s no permanently correct method in affiliate marketing. What worked last year may stop working this year, or in a few months. A tutorial from five years ago may now be a cautionary example.

To confirm that direct linking still works, I deliberately created new campaigns last night. It still works.

One last thing

The plainest ingredients often make the best dish.

I’m not saying tools don’t matter. But before you’ve tested the simplest approach, don’t rush to pile on complexity.

Can direct links run profitably? It’s not up to me. Test it.

Is the data enough without a tracker? Try it.

Your own test results are worth more than anyone’s advice. Note: this case isn’t necessarily replicable. It’s for reference only, so use your own judgment.