Why native is structurally different
Paid social is auction-based intent capture against an algorithm that can suspend you. Search is auction-based intent capture against an algorithm that mostly cannot. Native is something else entirely: it is editorial-style placement on premium publisher sites, served through DSPs (Taboola, Outbrain, RevContent, MGID, Yahoo Native, Yandex.Direct ad blocks, AdNow, propellerAds, RichAds), with conversion happening through a carefully engineered editorial-style landing page rather than a direct app store or registration page.
The structural difference matters because:
- Native networks have higher tolerance for regulated industries than Meta or TikTok at scale.
- Inventory is largely incremental to social — different audiences, different mindset.
- Account suspension risk is lower; the workflow assumption is different.
- CPM is structurally lower (often 5–10× cheaper than Meta), so the funnel can absorb longer pre-lander journeys profitably.
The editorial funnel
The reason native works for iGaming is the editorial funnel — the bridge between the placement and the operator. The user clicks a native unit ("Five things to know about online casinos in [country] 2026"), lands on an editorial-style article that genuinely informs the reader (5–8 minute read), and is presented with the operator as the recommended option at the end.
Done correctly, this is not "tricking" the user — it is the difference between intent capture (paid social) and intent generation (native). The user did not search for a casino; they read an article that gave them a reason to consider one. The economics are different because the context is different.
The pre-lander article carries the regulatory and brand-context burden, which is why two things matter:
- The article must be genuinely informative — not thin clickbait. Bounce rate is the canary.
- It must comply with the local regulator's content rules around responsible-gambling messaging, age-gating, and operator framing.
The nine networks that matter
For licensed iGaming, these are the native DSPs we run consistently across markets:
| Network | Strength | Region focus |
|---|---|---|
| Taboola | Quality publishers · scale | Global · strongest in Tier-1 |
| Outbrain | Quality publishers · slightly lower CPM | EU · LATAM |
| RevContent | Aggressive scale · less premium | NA · LATAM |
| MGID | Strong CIS & LATAM coverage | CIS · LATAM · Tier-2 EU |
| Yahoo Native | Premium · high CTR | NA · UK · APAC |
| PropellerAds | Push + native blend · scale | Tier-2 · Tier-3 |
| RichAds | Push-led with native overlap | Tier-2 · CIS |
| AdNow | Niche but valuable in pockets | EU Tier-2 |
| Yandex Native | CIS-only premium inventory | CIS |
Most operators run only Taboola and Outbrain. The 7 others are the diversification layer that makes native scale gracefully past €100K/month on the channel alone.
Pre-lander A/B as the core skill
The hardest part of native is not buying. It is engineering the pre-lander article so that bounce rate stays under 35%, time-on-page averages 90+ seconds, and the bottom-of-page CTA converts at 8–14%.
That is creative manufacturing of a different kind: long-form editorial copy, original-style imagery (or licensed editorial photography), local-language native writing, and a CTA flow that feels organic. Operators who write the pre-lander as if it were a Meta ad get 65% bounce and €70 CPA. Operators who treat it as a regulated-industry article get 28% bounce and €19 CPA. We have personally measured that exact range with the same operator switching framework.
We A/B test pre-landers on three dimensions, in order of impact:
- Headline framing — list, comparison, expert-recommendation, news-style.
- Article structure — sections, length, embedded data.
- CTA placement and language — single CTA at end vs distributed, copy variants.
CPA benchmarks per region
Approximate CPA ranges we see for licensed casino FTDs via native, post-pilot, optimized:
| Region | Typical CPA range | Notes |
|---|---|---|
| UK | £35–£60 | Premium inventory · regulated tone required |
| DE / AT | €22–€38 | Strong native inventory · responsible-gambling messaging |
| Nordics | €30–€50 | SE / DK heavy compliance · DK lower CPA |
| SE Tier-2 (CZ / SK / HR) | €10–€22 | Best CPA arbitrage in EU |
| LATAM (BR / MX / CL) | €8–€18 | BR scaling fast · MX premium · CL niche |
| NA (CA-ON) | CA$45–CA$70 | AGCO compliance heavy |
These are realistic operating ranges, not best-case. CPA below the low end of these ranges typically signals a tracking or attribution issue rather than genuine performance.
Why most operators get native wrong
The three failure modes we see most often:
- Treating native as direct-response. Skipping the pre-lander, sending users straight to register. CPA looks fine for two weeks, then conversions die.
- Using affiliate networks instead of running direct. Working through a managed-service affiliate that charges 30%+ on top of network rates. Margin destroyed; control lost.
- One network, no diversification. Same fragility as Meta-only.
Where native fits in the architecture
For our scale operators, native sits at 20–35% of total media spend, not the 5–10% most operators have it at. It is the channel that picks up volume when paid social is constrained, the channel with the longest account survival, and the most predictable CPA at €100K+/month. We treat it as a foundational pillar of scaling FTDs, not an experimental channel.
If your operator is running native at less than 15% of total spend, that is the cheapest experiment to run next quarter. The pre-lander work is what makes the difference — not the network, the budget, or the bidding strategy.