Testing native ads for Nutra: $3,547 in Mexico (Case Study)

Currently, within the arbitrage media space, there is high demand for alternative traffic sources. For instance, certain affiliate programs, seller agents, and influencers are actively promoting native ad networks. So today we will present a case study from one of our media buyers who was running campaigns with native traffic for the first time and immediately turned it into great profit!

Introductory information about the campaign

• Offer: Diapsula
• Vertical: Diabetes
• Affiliate Network: OMNI CPA
• Traffic Source: AdsKeeper native network
• Campaign Period: 04/12/2025 – 05/12/2025
• GEO: Mexico
• Amount Spent: $2661.66
• Revenue: $3547.00
• Profit: $885.34
• ROI: 33%

The task was simple — to test native traffic for the diabetes vertical targeting Mexico. We worked with the Diapsula offer from OMNI CPA; the funnel was already checked out and proven to work well with another source, so the question was to understand how relevant AdsKeeper is for such cases. I will note separately that our work together was extremely pleasant: a manager was assigned immediately and then provided useful advice throughout the entire campaign. The landing page was pre-moderated instantly without any requests for changes, without additional conditions or cloaking. This saved a significant amount of time at the start.

The entire setup boiled down to generating a tracking link in the affiliate network panel (affiliate network), connecting our tracker, and launching. There was no headache with creatives, no "optimization" to fit their moderation policy — in general, they allowed us to enter the source like normal people.

Creatives and Pre-landings​

As mentioned above, the funnel was already prepared, but following the manager's advice, we added another pre-landing page to the split test. One featured a classic format with a Mexican endocrinologist, Dr. Carlos Ramirez, known to many from other sources, and the other featured an interview with a famous Japanese scientist who explains that he was threatened by the Mexican mafia after launching his new method that eliminates type 2 diabetes from the body in a few weeks.



Spoiler: the pre-landing with the Japanese specialist had an average conversion rate about 20% higher.

Different giveaway formats with doors and first-aid kits were also tested. On AdsKeeper, the "doors" performed better, unlike with another source where first-aid kits achieved a higher CTR.

For the teasers (creatives) — a standard teaser approach was used, appealing to an unexpected cause of the disease. The texts were as follows:

• "Endocrinólogo: 'Bebe esto para eliminar la diabetes en 7 días.'", which translates to: "Endocrinologist: 'Drink this to get rid of diabetes in 7 days.'"
• "¡La diabetes no es por lo dulce! Lee aquí la principal causa", which translates to "Diabetes is not from sugar! Read about the main cause here". This phrase was used on teasers with cashews and with root/grater.

Below are the teasers that showed the best results. Images of leaves, nuts, horseradish, and test strips were used. AI generators were not used — archived creatives that had worked previously were utilized. The effectiveness of the creatives was tracked exclusively by EPC.



Here you can see the breakdown by impressions and clicks:

Campaign Setup​

The campaign was launched targeting Mexico exclusively on mobile devices — desktop was cut off immediately to avoid spreading the budget too thin. The language, of course, was Spanish. The minimum test budget was $100, with a click bid set starting from $0.025. I tried not to overcomplicate things: the task was simply to understand how the source itself works under standard conditions.

The AdsKeeper manager provided a whitelist of placements for the offer, so the launch was faster than expected. This is also a plus — there was no need to build zones from scratch or burn budget on random placements. A landing page with a fixed price — 590 Mexican pesos — was used; it didn't need to be altered for the source, everything went through as is.

Campaign Launch, Optimization, and Scaling​

Optimization was done at the source level — through auto-rules: if the conversion cost was > $6 over 7 days, the placement was blocked.


Bid coefficients (CPC bids) were also manually adjusted for different placements. Those that provided cheap leads without approval were cut off. Placements were tracked in the affiliate network panel for analysis; the focus was not only on lead cost but also on approval rate and EPC, as leads from some placements could be cheap but have no approval.

Regarding scaling — the entire campaign was mostly created to get a feel for the source and understand how everything works here, so the case study remained within the framework of one "lazy" test.

During the campaign, $2,661.66 was spent, 94,240 clicks were received, total conversions according to AdsKeeper statistics — 674 at $3.95 each (394 conversions reached the affiliate network panel, of which 248 were valid, and 128 were approved). EPC — $0.06. Well, $2,661.66 was spent on traffic, earned (looking at the affiliate network panel stats) — $3,547, and net profit was $885. ROI — 33%. We consider the test successful.

Conclusions about working with native traffic​

The experience showed that native traffic can work even with a "lazy" launch without fine-tuning. The main thing is to start with a proven funnel and not neglect basic things like whitelists/blacklists. It is also important to understand that pre-landings can perform completely differently depending on the source: what works well on FB might not convert on native traffic at all. Therefore, it's important to be ready for tests and replacing the pre-landing.

Regarding bids — it's better to set a bid at around 80–90% of the recommended one, and then adjust manually. Furthermore, it's important not only to lower bids on expensive placements but also to raise them on those that yield cheap and high-quality conversions. Native traffic operates by auction rules, and if you set your bid too low — you buy residual, often lower-quality traffic. If you set it too high — you risk buying more traffic than necessary and going into the red. Therefore, it is critical to monitor the balance of CPC + CTR: a teaser with a low CTR might have a cheap conversion, but it needs a higher bid to win impressions. Hence the conclusion — bid management for placements and teasers must be dynamic, and manual adjustments yield better results than automated rules.