3 client slots open —     
hello@native-advertising.net
native-advertising.net

6 min readBy Marcel Sattler

5 Native Ads Mistakes That Cost Me $1M on Taboola & Outbrain

Five native advertising mistakes on Taboola and Outbrain cost me over $1,000,000. Here is exactly what went wrong and how to avoid the same losses on your own accounts.

From the post

Five mistakes on Taboola and Outbrain cost me over $1,000,000.

— Marcel Sattler

↓ read on

Five mistakes on Taboola and Outbrain cost me over $1,000,000. Not in one campaign, not in one quarter, but across years of buying traffic on platforms where the entry barrier is brutal and there is almost zero legit material to learn from.

I'm telling you all five so you don't repeat them. The good news: every one of these is avoidable once you know what to look for. The bad news: most new media buyers walk straight into all five in their first few months because the playbooks they learned on Facebook and Google don't transfer to native.

I'm Marcel Sattler, founder of native-advertising.net, and since 2015 I've deployed more than $100M across Taboola, Outbrain, Newsbreak, MGID, Yahoo Native, Mediago, and RevContent on DTC, dropshipping, lead-gen, and affiliate offers. The mistakes below are not theory. I paid for each of them with real spend.

Why copying a competitor's native ads funnel doesn't work

Meta has the Facebook Ads Library. Native doesn't have an official equivalent, but spy tools exist for Taboola, Outbrain, and the rest. You type in a keyword, you see the running ads, you click through to the landing pages. So the temptation is obvious: find a funnel that looks profitable and duplicate it one-to-one.

People copy everything. The ads, the images, the headlines, the funnel flow, every landing page. Then they launch and wonder why the numbers don't match. The answer is simple. It's not the same funnel, even if every pixel matches. It's a different account with a different learning state, and the algorithm treats it as a brand-new entity.

Cloning the whole machine makes as much sense as a solopreneur copying Microsoft's or Oracle's infrastructure on day one. The procedures are public, but they're built for a company in a completely different state than you. In 2024, native rewards creativity, not duplication. Native is content marketing, and content has to be unique, especially now that AI can mass-produce the cheap, scammy-looking pages that no longer convert.

Use AI tools. Use new tech. But give people real value before you ask them to buy. If you want help building a funnel that fits your account instead of borrowing someone else's, that's exactly what we handle for DTC and dropshipping and affiliate clients.

Always split native campaigns by device on Taboola and Outbrain

I started as a Facebook media buyer at 17. Back then the algorithm was already smart, and today it's far smarter, so on Meta you set up a campaign, most of your traffic is mobile, and the algorithm does the rest. That habit cost me on native.

Taboola and Outbrain have an algorithm too, but it's years behind Meta and Google. So you can't lean on it to sort device performance for you. The fix is structural: split every campaign by device. Run one campaign for desktop, a duplicate for mobile, and optionally a third for tablet.

Splitting by device matters for two reasons:

  • You can optimize each device independently instead of letting one drag the other down.
  • You set separate bids. This is a bidding platform, and a desktop click is worth a different amount than a mobile click, so you pay accordingly.

If you're asking what to bid per click, there's no single answer. It depends on the country, the device, the niche, and the traffic you're shown. There is no one-size-fits-all bid in native. What is non-negotiable is the device split. Skip it and you're optimizing blind, which is how my early Taboola and Outbrain budgets bled out. We rebuild account structures like this for lead-gen and ecommerce advertisers every week, and the device split is always step one.

Stop being emotional about your media buying campaigns

This one isn't native-specific, it's media buying in general, and I'm guilty of it. You build a campaign, you pour creativity and hours into it, you hit activate, and you wait. And wait. And the results never match the excitement.

The problem is the emotional attachment. The more effort and creativity you put into a campaign, the more connected you feel to it, and the harder it becomes to turn off. Killing a campaign you've barely touched is easy. Killing one you sweated over is painful, so people let it run far past the point where the math says stop.

As a media buyer, your job is KPIs, not feelings. Look at the numbers. If the campaign doesn't make sense on the KPIs, shut it down. Don't ride a dead horse because you're proud of the creative. The emotional connection forms fast, often within the first hours of a launch, so build the discipline early.

Don't overkill the campaign or cut it off too soon

This is the flip side of the last mistake, and it's where I see new media buyers go wrong constantly. They hit F5 every five minutes, refresh the ads manager, and tweak something on every refresh. Native is not day trading. Constant adjustments don't help, they sabotage.

The algorithm needs time. Your funnel needs impressions and visibility before it can show what it's worth. So don't overkill the campaign with changes, and don't cut it off prematurely. The most common version of this: a media buyer who came from Facebook spends $200, sees zero conversions, and kills the campaign. In most cases that's too soon.

Before you cut, optimize. Find the error inside the campaign first:

  1. Check where your traffic is coming from. If you're getting low-quality push traffic, that's a structural problem, not a creative one.
  2. Dig into the funnel statistics and find the step where people drop.
  3. Fix the leak, give it time, and only then decide whether to kill it.

Avoid too many changes in a short window. The two failure modes, refreshing every five minutes and quitting at $200, are both impatience. Native pays patience.

Choose the right Taboola account type: self-service vs. rep

The fifth mistake cost me real money: running on the wrong account type. On Taboola, Outbrain, MGID, and most networks there are self-service accounts and rep-managed accounts where a platform representative is assigned to you. The difference is not cosmetic. It directly affects how profitable you can be.

Here's the concrete example. On Taboola, you often need a rep account to use a block list. A block list lets you cut off the scammy push traffic that, in most cases, converts far worse than expected. Premium publishers don't run on push traffic, so without a block list you keep paying for garbage placements that drain budget.

A rep is also your direct line to the platform. Questions come up sooner or later, and a good rep sees data you can't see in the dashboard. Build a strong relationship and they'll share insights and tips that move your numbers. If you're stuck on a self-service account with no block list, you're fighting with one hand tied. We operate as the rep-level layer for clients across Taboola, Outbrain, and MGID, which is how we get block lists and premium inventory working from day one.

Watch the full breakdown

Is your account a fit for the same play?

If any of these five mistakes sounds familiar, the cloned funnel, the single-device campaign, the dead horse you can't put down, the itchy F5 finger, or the self-service account with no block list, you're not alone. These are the default failure patterns for everyone learning native, and they're all fixable with the right structure and the right rep-level access.

The fastest way to skip the seven-figure learning curve is to have someone who's already deployed $100M+ audit your setup. Book a strategy call and we'll look at your account structure, your device splits, and your traffic quality. You can also browse our case studies to see what these fixes look like in real accounts, or work through our free resources if you'd rather start on your own.

▸ Keep reading

Three more on the same topic.