Testing is a must! A case with a profit of $43 962 on diabetes from an anonymous webmaster

I'm not seeking fame, so let's skip the names. I'll just say that for over four years now, I've been running nutra campaigns in the LatAm region. I won't be flaunting any astronomical ROI figures; I'll simply share my experience, which consistently brings me a profit. $44K in two months with a single offer – not too shabby.

This case will focus on how to run diabetes offers in Mexico. I started working on this almost from the very beginning of my career, and my approach is still effective. How is it possible? I've detailed everything below. I should also mention that while I usually work solo, from time to time, I request assistance from other buyers, and we run campaigns together. In this case, we were working with other folks, and they were able to replicate what I had done multiple times before. In summary, if you've gained experience, don't be afraid to deviate from the templates. The phrase “Testing is a must!” didn't become the motto of CPA for no reason.

Basic information:

• Offer: Insunol
• GEO: Mexico
• Campaign period: 17.03.23 — 20.05.23
• Total leads: 16 637
• Confirmed leads: 6525
• Bid: $18–22
• Approval rate: 39%
• Expenses: $88 552
• Revenue: $132 514
• Profit: $43 962
• ROI: ~50%
• Traffic source: Facebook

Features of the campaign setup

When running the campaign, we used a classic scheme: one farming account + 9 low-limit fresh accounts. We didn't use automated rules, all bids were set manually.

Before identifying the most converting combination, we conducted numerous tests. Each buyer was looking for their own approach, which was then scaled up. There were many accounts, and each was allowed to run for a minimum of 3 days. After all the tests, we turned off expensive campaigns and worked with the approaches that yielded good statistics. We used Keitaro as the tracker and AdsPower as the antidetect. We also used mobile proxies.

In general, if you want a high conversion rate, I advise you to look for something unusual—something highly interesting and unique. Try different approaches with the same combination—don't get fixated on just one thing. It may seem like a waste of time, but as soon as your combination consistently shows a good conversion rate over an extended period, the bids will undoubtedly skyrocket. However, if you keep testing, by that time you'll already have several more approaches to distribute traffic between, thereby reducing costs.

Moreover, different approaches elicit different responses from various types of audiences, allowing us to scale up volumes. For example, on our best days, we managed to achieve around $4K in revenue. And this wasn't a one-time occurrence.

Creatives

We also tested different creatives. We used both classic marketing stories and what the buyers themselves came up with. These stories were packaged into one-two minute videos and performed quite well. In the end, the best conversion rates came from creatives with a product-oriented approach—where the offer image was shown to the user right away.

I won't reveal the best creatives—they are still in use. Here are examples of regular creatives that also performed well.

Statistics

I didn't bother searching for screenshots from Facebook because a large number of accounts were used for the campaign, and screenshots from two or three wouldn't give a complete picture. Attaching 50 screenshots would be a bit excessive. There won't be any payment proofs either because the same cards were used for several offers, making it impossible to extract Insunol-specific data from the statistics. So, here are the screenshots from Keitaro and Leadbit.

In total, during the campaign period, $88 552 was allocated to all the buyers, and we managed to earn $132 514. The total profit for those running Insunol offers amounted to $43962.

Plans for the future

LatAm has become a bit cramped for me—I want to expand, so I'm actively testing different European GEOs now. I can't boast of significant successes in Europe yet, but there's a positive dynamic, which is good. Essentially, I've shared everything I wanted to. When there are results from Europe, I'll definitely share them. So, let's say goodbye for now! I hope my case will be helpful.