Case Study: KetoBalance Top Weight Loss Offer – $11,235 in 12 Days

Hello, everyone! The TriTe team is here with one of its team leaders – ArtemKO. Today, we will study the weight loss offer from the M1 Direct Advertiser as in the case of TikTok Ads. But first, let me tell you a background story…

It’s no secret that since 2022, the TikTok Advertising Platform has taken up advertising optimization in earnest. In the way Google or FB does, they began to clamp down on nutra offers and other ads with “black” approaches.

Nutra has always been our team’s main focus in different GEOs. We have already succeeded with the offer in question before, but this March, the guys from M1 offered us excellent partnership terms. We couldn’t refuse so I decided to resume the tests. Read on to find out how that turned out…

Offer Breakdown

In Keitaro, you can see the total leads volume, but, unfortunately, we didn’t pull the expense in this case. For this, the team uses third-party services. See the second screenshot.

A couple of indicators are shaded, as this is confidential information about taxes and team policy.

Traffic Arbitrage Approaches

To run ads, I used an agent’s business center, where I obtained ad cabinets. Cabinets did not require farming, being on hold, or else. I drove traffic using different methods. When new creos were tested, I drove traffic manually in a small volume according to the one campaign/ten adsets/two creos principle. There were three or four such campaigns per cabinet and up to ten cabinets per test.

Yes, there are strict rules in our team: 60+ creos (minimum) per buyer daily. I duplicated the best creos five to seven times. After the third cycle, I picked top creos and used them in other cabinets as tables (TT feature), making many copies of the best creos. Then, I multiplied them in cabinets according to the scheme. Even though I was near the computer all the time, to play it safe, I used auto-rules with the necessary settings, which TT also provides.

Screenshots from Ad Cabinets

P.S. I’m sorry to say this, but TikTok scaling depends on cabinets, and since they were relatively quickly banned, I can’t provide the screenshots of all cabinets here.

Campaign Settings + Targeting

Let’s start with the age settings. This is important! I started with 25-55+ and received feedback from the M1 that there are a lot of deviations from the required audience of the age of 25-28. There were several reasons for that: they couldn’t purchase the product, because it’s expensive or they had no money at all, etc. Another major reason for cancellations was that there was no delivery option. There are islands in Spain where they can’t deliver goods, hence many order cancellations. We solved this problem quickly by excluding it in the GEO settings. Fortunately, the TT settings allowed this (but there are other restrictions in TT).

When I decided to target a 35+ audience, in a couple of days the traffic’s daily dynamics dropped. This is because the majority of the TT audience are people of the age of 25-35+.

After three days, I decided to restore the age settings. There was a deviation from the age of 25-28, but the lower threshold in TT after 25 is 35+, which means that the audience from 28 to 35 years old was lost, which is a lot of people. When I restored the age settings (25-35+), useful leads increased noticeably, but the deviations remained. But it was worth it!

Internal Adjustments

  • Event: Completed payment, Add to cart
  • Placement: TikTok
  • Gender: M, F
  • Age: 25-55+
  • Interests: None
  • Bidding strategy: Standard
  • Advertising objective: Conversions
  • Budget: Any
  • Automatic creatives optimization +

Ad Settings

  • 2 creos,
  • 1 icon (any),
  • 5 descriptions,
  • Buttons: dynamic or standard which includes two actions.

Creative Approaches




While ads were running, I did not use pre-landing pages or cloaks. Although there is one campaign with a cloak in Keitaro, it did not perform well – the traffic was cut a great deal.

The very approach was about “whitening” the landing page + creative with no “before and after” pictures. There was no product text box on creative as well.

In case creative didn’t pass the review after some time, then there were two solutions to opt for:

  1. Use the creative uniquelization service or render it again in the video editor to obtain new metadata;
  2. In the video editor, overlay frames, fog, or rain effects on the video, and if they are too bright, then lower the transparency level so that the creative is visible. It also helps to make creatives in b/w.

Complications We Faced

A large number of creos in a short time

Do daily tests with the best creative, even if you already have a working scheme. I scaled campaigns but kept on making new creos.

What to select: “Completed payment” or “Add to cart”?

Both events work. However, “Completed payment” drives traffic, steadily picking an audience, while “Add to cart” provides a very quick turnover and a lot of traffic, but you must make lower campaign bids in this case and keep a closer eye on it! I’d go so far as to say that you must feel it, otherwise, the turnover will be excessive and you won’t get any valid indicators.

Age settings

As I mentioned above, there were deviations, so I had to give up some part of the audience and drive traffic as is.

Many variables tested in the TT cabinet

I should have tested which button was better, whether I should have driven traffic only on TT or on other platforms that had cabinets (like Pangle), whether I should have opted for a standard bidding strategy or the lowest bidding price, and whether or not I should have opted for targeting, etc.

The only difference between button options is that in case you choose a dynamic button, you won’t be able to upload the campaign to the table after the campaign is finished. You also can’t upload a table in a campaign with a dynamic button, since there is no such parameter in the table or it will throw an error message. In terms of the best indicators, the buttons do not matter. They perform equally well. When I tested the static button, I opted for “in store/get quote”. The “More Details” button works just as fine.


We would like to thank the guys from M1, the entire M1 affiliate network, and our account manager Dmitry in particular. Our special thanks to manager Irina for assistance in this case and with other partnership issues.

I was very surprised by the way the guys did their work: the call center specialists made calls to confirm orders ten times a day, or even more. Despite the short promotion period, in case the quality of the traffic dropped, the guys immediately gave feedback. This was not just brief feedback, but a full breakdown of each lead which made it clear what was wrong and what we should fix.

Even with the deviations (there are many factors like the age of the audience, delivery, etc.), the approved leads count was huge! Given the low price per lead, ROI was very pleasing. The payout per lead was $35 with great bumps. Cutting to the chase, it was great!

Thank you!

TriTe Team

Like it? Share with your friends!