Recently, we found an interesting advertising campaign project which was performed by Andrey, the former director of Taboola for Eastern Europe and Israel and the co-owner of the advertising agency Xevio and the educational project on native advertising, Native Hub. Here is what he has to say.
We started a campaign that promoted an orthopedic shoe brand with a spend of over $1 million on Taboola and Outbrain.
The reason for our active participation in the industry is very simple: we believe that native advertising is a high-quality alternative to Display, Social, and Search, i.e. a niche where you can consistently spin off colossal volumes while maintaining decent margins without black/grey hat practices. Our core business is agency, but having spun off more than $100 million promoting lead generation, D2C (aka E-Commerce) and search arbitrage, we began actively collaborating with Taboola, Outbrain, MGID and Baidu in the educational sphere. We help them educate solvent clients and provide all the necessary resources for this: our personal recorded trainings, tools for purchasing at a discount, group consultations and articles with data.
- Vertical: D2C (E-Commerce)
- Type: Shoes
- Source: Native networks Taboola & Outbrain
- GEOs: US, CA, AU, Germany
- Peak Period: 01/01/2024 – 31/01/2024
- Tracker: Voluum
- Spent: $935,080 on Taboola, $157,590 on Outbrain. Total: $1,092,670
- Earnings with Agency Model: $163,900 (15%)
- ROAS/ROI: 2.39 / 139% (profits if not using the agency model)
At AffDays, we have more case studies in different verticals, one of which is gambling. Have a look at the NineCasino offer from Chilli Partners case study.
Choosing the Source
There are several reasons why we chose the native advertising networks: stability, the least possible number of bans, and massive volumes.
By stability, we mean a range of factors, from the absence of fluctuations in spend based on the set daily budget to the minimal need for constantly changing approaches. If some white hat practices work in Taboola, they will most likely work in Outbrain.
When we talk about bans, we refer exclusively to white hat practices. The moderation teams of native networks are much more lenient than the soulless algorithms of other industry platforms. Bans are replaced by “rejections,” which are essentially soft denials. If you have a dedicated manager with whom you’ve built a relationship, they won’t just silently block your account. They will also provide detailed reasons for the rejection and recommendations on the necessary adjustments for approval. This is invaluable.
Speaking about volumes we find it important that the source can actually engage a large number of people across the Internet. Thus connecting businesses with engaged audiences can drive better marketer results.
Choosing the Vertical
Selecting an apropriate vertical was quite far from being a straightforward task. However, we decided to settle on Direct-to-consumer model that involves all possible niches, from gadgets and health and beauty to clothing and footwear and FMCG (Fast-Moving-Consumer-Goods). The search for a product led us to orthopedic footwear. A few months ago, this niche was almost nonexistent in native advertising (low competition), production and delivery fit our math, and the specifications matched the target audience. Native advertising caters to a financially capable 40+ audience that either has or is starting to experience health issues and, therefore, begins to actively monitor it.
Choosing the GEO
We often start our tests with Germany. Most of our team speaks German, making it much easier to dynamically test funnels without involving external translators. Alternatively, we can write in several other languages, including English. However, other factors come into play, such as market capacity and the availability of quality traffic. In Taboola and Outbrain, Germany is the largest European market with the biggest sites (comparable to the UK) for testing.
After testing Germany, we quickly expanded to the US market. Once we achieved solid results there, we decided to expand to the main English-speaking markets with minimal localization efforts.
If you work with affiliate networks, you probably won’t notice this due to payout equalization; we rarely encountered situations where an identical D2C product within the same CPA network had different payouts. Most often, the affiliate program absorbs all associated costs. However, if you work with direct advertisers, KPIs can vary greatly due to factors like local VAT, shipping costs, storage costs, customs duties, and so on. In such cases, upsells save the day by increasing the Average Order Value (AOV) and compensating for CPA costs.
Budgets
We don’t work with small brands. You can go for quantity, focusing on diversification, working with advances and a profit-linked commission rate. It’s a valid model, but we prioritize aggressive, targeted sales to brands that have both the funds and a high likelihood of success with native ads. The same logic applies to affiliates. You need to assess your capacity and either test many small offers or a few large ones, depending on your resources.
Working with American or Western European brands is much easier. They allocate higher test budgets due to their accustomed large internal markets. From the start, we agreed on a 4-week test with $20,000, split into $10,000 for two major GEOs.
Purchase Details
- Native ads generate cold traffic, making it pointless to work without a pre-landing page. Visitors coming from news websites are rarely ready to perform your target action. Most often, they click through to continue reading news. It’s the marketer’s job to craft a story that sparks interest and triggers a purchase or registration.
- Not all brands can be sold through a pre-landing page or even native ads in general. It’s not just about the difference between COD sales with one-pagers or SS sales with two-stage processes through Shopify. Given the nature of the traffic (cold), lack of precise targeting (absence of a closed ecosystem), and the high age of the audience (Boomers, or at best Gen X) with their corresponding financial situation and habits, you need to have a clear understanding of what these people need. Remember: no B2B and no products that require consulting a spouse. Only spontaneous decisions that can be justified as a fleeting existential indulgence.
- Decisions are made based on three metrics, in descending order of importance: LP CTR, CPM, CPA. If the click-through rate from the pre-landing page is below the desired 15-20%, there is no point in optimizing websites or creatives. To study on-page behavior, we recommend using the free tool Microsoft Clarity—it not only tracks LP CTR but also helps create heatmaps. Only after succeeding with the landing page should you consider optimizing the sales page. CPM (Cost per Mille) is a unique story. Unlike Facebook, where the goal is to lower the metric, in native ads, the opposite is true—you need to increase it. The higher your click or creative CTR, the higher the CPM, the more often you appear in the feed, and the higher the likelihood of a transition to an optimized pre-landing page. CPA (Cost per Acquisition): only when the first two elements are optimized should you start working with sites, forming blacklists and whitelists, and bringing your CPA to the desired value.
- On average, it takes a couple of days to make decisions on LP CTR, so in the first 2 weeks, we can make up to 15 iterations of pre-landing pages. ChatGPT helps, but it’s nowhere near as good as skilled copywriters. This doesn’t apply to designers, though: MidJourney works wonders, so you should keep only truly irreplaceable specialists in the team who have boundless creativity, rather than those who perform trivial, easily replaceable tasks. It takes a few more days to optimize the CPM, and after that, regularly monitor the sites to maintain the campaign at the desired level—without becoming obsessive.
- We used a variety of creatives, ranging from basic (foot and shoe) or hero-focused (a girl/guy holding a shoe and smiling) to problem-oriented (gifs with an orthopedic focus) or completely wild (a famous sneaker brand burning in flames, and as the gif ends, our shoes land on the ashes). In short, experiment, just keep your sets separate.
Conclusion
After testing many different products over the past few months, we were lucky to find one that went viral. We put in a lot of effort to find the perfect approach—we went through over 50 pre-landers and more than a few hundred creatives before reaching the volumes we’re currently maintaining.
Scaling is becoming harder. Over the past few months, quite a few competitors have emerged, each trying to stand out with their own strategy: either undercutting prices or attempting to create a cheaper product. Some have even managed to temporarily seize the initiative. I can’t say I recommend promoting shoes right now, at least not launching with a new product. The hot season is starting, so it’s time for fishing rods, mosquito repellents, air coolers, tents, portable fans, and other such nice but somewhat useless items. But if people are buying, then it’s worth selling.
