Industry analysis

The 2026 state of native ads — a traffic-source stack rank

An annual stack-rank of the major native-ad and content-recommendation networks heading into mid-2026 — Taboola, Outbrain, RevContent, MGID, Yahoo Native, Bing Audience, Google Discovery — with what's working, what's broken, who's buying, who's selling, and how much each network's payout windows actually drift.

By Eyal RosenthalMay 6, 202617 min readAI-assisted research

This is the annual state-of-the-native-ads piece, the kind of thing AdExchanger writes once a year and AdWeek writes twice a year and every operator-side blog tries to do once and abandons by the second year. I'll try to make this one not abandon. The structure: an honest read on each major traffic source — what's working, what's broken, who's buying, who's selling, payout windows, account-management experience, supply quality. Sourced from the public IR materials of the publicly-traded networks (Taboola, Outbrain, System1), from official platform documentation, from earnings-call transcripts Q4 2025 and Q1 2026, and from operator-side reporting where I can cite credible primary sources.

The state of the industry, top-down

A few macro framings before the per-network breakdown:

The duopoly is more entrenched. Taboola and Outbrain together account for the majority of high-quality (premium-publisher) native impression supply globally. Their 2024 merger discussions ultimately did not consummate, but the competitive dynamic continues to favor their combined dominance over the second tier (RevContent, MGID, Sharethrough, Nativo). Read Taboola's 2024 10-K and Outbrain's 2024 10-K for the supply-side concentration.

The search-arbitrage tail is being structurally compressed. System1's 2025 earnings calls have repeatedly described "RPM compression" in tier-1 verticals as the search partners (Google, Yahoo, Bing) tighten their click-quality acceptance. The System1 IR archive has the recent transcripts. The implication for the buy-side is that the easy-money search-arb opportunities of 2018-2022 are mostly gone; what remains rewards operational excellence, not "find the right offer."

AI-creative supply is reshaping the demand side. The cost of producing 100 creative variants has fallen by an order of magnitude over 18 months. The networks' compliance teams are racing to detect AI-generated faces and AI-generated income claims. The buyers who can produce-fast-and-still-pass-compliance are gaining share over the legacy "one creative, one campaign" operators.

Programmatic native via DSPs is gaining share at the expense of the closed networks. The Trade Desk's native solution, DV360's native inventory, and Sharethrough's open programmatic native have grown materially as a percentage of total native spend. The IAB's State of Programmatic report tracks this. The closed networks (Taboola, Outbrain) still dominate premium publisher supply but the competitive frontier is increasingly programmatic.

Now, network-by-network.

Taboola (NASDAQ: TBLA)

The shape of the company. Public, ~$1.5B+ in 2024 revenue per their 10-K, majority-mobile inventory. Yahoo deal (the multi-year exclusive partnership announced 2022 and rolled out from 2023) gave them a structural revenue lift that has shaped 2024-2025 financials.

What's working. Strong tier-1 publisher relationships in the US, EU, and APAC. The Yahoo deal materially expanded supply. Their Open Web programmatic positioning is the most aggressive of the closed-network incumbents — they are pushing harder than Outbrain to become a destination for non-Taboola-direct demand. Their advertiser tooling (Taboola Ads, the native bidder, the conversion-bid-strategy product) is mature and competitive with Outbrain.

What's broken. Margins remain compressed relative to where they were pre-Yahoo-deal — the deal traded margin for revenue scale, and that's reflected in the Q4 2025 earnings call. Account-manager turnover at the mid-tier has been noticeable per operator reports; the team experienced is variable. Compliance-side responsiveness on borderline creative is slower than Outbrain in 2026 by my own observation.

Who's buying. The advertiser mix is still heavily content-arbitrage and search-feed buyers, with a meaningful and growing share of branded direct-response (DTC, fintech, insurance). The push from Taboola's commercial team has been to diversify away from arb dependence; this has proceeded, but slowly.

Who's selling (publisher side). Tier-1 supply (CNN, USA Today, NBC properties, regional newspaper groups) is the core. Yahoo properties are the largest single supply contributor post-deal.

Payout windows. Publishers paid net-45 to net-60 from month-end. Advertisers charged daily/weekly. Standard for the industry.

Supply quality. Generally good. IVT below 5% on tier-1 by DoubleVerify's industry reporting. Some long-tail publisher inventory has higher IVT but is filterable in the bidder.

Stack-rank position: #1 for scale, #2 for margin per click after Outbrain, #1 for search-arb friendliness at present.

Outbrain (NASDAQ: OB)

The shape of the company. Public, ~$900M+ in 2024 revenue per their 10-K. More premium-skewed in publisher mix than Taboola, with stronger relative position in the EU.

What's working. Premium publisher relationships in EU, UK, US (CNN, BBC, Bloomberg-tier). Conversion Bid Strategy automation is competitive with anything Taboola or any DSP offers. The advertiser tooling is mature, the account-management at scale is generally competent. Compliance posture is the strictest among the major networks, which is good for advertiser trust signals once you're in.

What's broken. Slower scale growth than Taboola — the Yahoo deal asymmetry left Outbrain growing more slowly in absolute revenue terms. Their Q4 2025 earnings call (transcript via IR site) emphasized cost discipline and margin protection over top-line, which is the right play but not the exciting one. The supply-side concentration in EU exposes them to regional regulatory risk (GDPR enforcement, Digital Services Act compliance investments).

Who's buying. A more mature DTC/branded mix than Taboola. Search-arb buyers do operate on Outbrain but Outbrain's compliance has historically been more aggressive at filtering low-quality arb traffic, which means the operators who survive there are running cleaner offers.

Who's selling. Premium publishers; significant EU-newspaper-group supply.

Payout windows. Similar to Taboola — net-30 to net-60 publisher side.

Supply quality. Best-in-class IVT rates among the closed networks. Trustworthy Accountability Group certification and DoubleVerify third-party verification both confirm.

Stack-rank position: #2 by scale, #1 by margin-per-click and quality. The default network for advertisers who can clear compliance.

RevContent

The shape of the company. Private, US-based, smaller than the public incumbents. Estimated mid-nine-figures in revenue per trade press reporting, but no audited public number.

What's working. Cheaper inventory than Outbrain or Taboola for similar geos. Faster creative approval. More permissive on borderline verticals. Healthy long-tail publisher mix. Genuine product differentiation as the "more permissive native" choice.

What's broken. Account-manager experience is inconsistent. Reporting and tooling are noticeably less mature than Outbrain or Taboola. Click quality on the lowest-tier publishers is meaningfully worse — search-feed reversal rates I see operators report on RevContent traffic typically run 1.5x-3x the rate on Outbrain. The pricing advantage doesn't fully survive once that is netted out.

Who's buying. Search arbitrage operators, weight-loss / supplement vertical, certain financial-offer affiliates whose creative wouldn't clear Outbrain. A real and durable buyer base, but skewed toward verticals that Outbrain compliance has filtered out.

Who's selling. Long-tail and mid-tier publishers. Some reasonably well-known regional publishers; many unknown sites.

Payout windows. Net-30 to net-45 publisher side. Comparable to Taboola/Outbrain.

Supply quality. Variable. Best filtered carefully.

Stack-rank position: #3 for scale, #4 for quality, #1 for "test bed before paying Outbrain prices."

MGID

The shape of the company. Private, originally Ukrainian-rooted, US/EU operations, similar scale band to RevContent.

What's working. Strong presence in LATAM, EMEA, and certain APAC markets where Outbrain and Taboola have thinner publisher relationships. Reasonable advertiser tooling. Pricing comparable to RevContent.

What's broken. Geopolitical context (their historical Ukrainian roots) has created some operator-side wariness. Account-manager experience focused on non-US markets means US buyers can find them less responsive than Outbrain or Taboola. Fraud-rate signals are similar to RevContent in tier-1; sometimes better in tier-2/3 markets.

Who's buying. Heavy LATAM, EMEA, APAC affiliate operators. US presence smaller.

Who's selling. Diverse publisher mix skewed away from US tier-1.

Payout windows. Standard.

Supply quality. Variable. Strong in non-tier-1 geos.

Stack-rank position: #1 for non-US, non-EU geos. #4 for US tier-1.

Yahoo Native (post-Taboola deal)

The shape. Yahoo's owned-and-operated native inventory has been routed primarily through Taboola since the 2022 deal announcement and rolled out beginning 2023. From a buyer perspective, "Yahoo Native" is now substantially indistinguishable from "Taboola with Yahoo properties."

What's working. Significant supply at premium quality from Yahoo, Engadget, AOL, and related properties — accessible through Taboola.

What's broken. Less direct buyer relationship than the pre-deal era; the auction dynamics flow through Taboola's bidder.

Implication for buyers. Bid via Taboola, target Yahoo properties via the publisher list. There isn't a separate "Yahoo Native" buy.

Microsoft / Bing Audience

The shape. Microsoft Audience Solutions is Microsoft's native advertising product, served on Microsoft properties (Outlook, MSN), partner properties, and through the Microsoft Advertising platform. Distinct from Bing Search, though they share the Microsoft Ads platform.

What's working. Strong in B2B and certain professional verticals. Microsoft's first-party data (LinkedIn, Outlook signals, Edge browser data) gives differentiated audience targeting. Inventory is high-quality.

What's broken. Smaller scale than Taboola/Outbrain. Less affiliate-friendly historically — Microsoft's compliance has been stricter than the closed networks.

Stack-rank position: Worth a slot for branded direct-response, especially in finance and B2B. Less obvious for arbitrage.

Google Discovery / Google Ads native

The shape. Google's native-equivalent is Discovery Ads (now folded into Demand Gen), serving on Google's owned properties (Discover feed, Gmail, YouTube home) and partner inventory.

What's working. Massive scale. Google's first-party signals. Strong machine-learning bid optimization.

What's broken. Google Ads' Misrepresentation policy is the most aggressive of any major network. Search-arb to a Google-partner search feed is not really viable (you can't arb Google's search to Google's search). Compliance is unforgiving. Account suspension is final and fingerprint-aware.

Stack-rank position: Real for branded direct-response. Hostile to traditional affiliate.

Sharethrough and other programmatic native

The shape. Sharethrough and similar programmatic native players (Nativo, TripleLift native, Connatix) operate as DSP-accessible native inventory.

What's working. Open-programmatic accessibility means buyers can route through The Trade Desk, DV360, etc. and access native inventory alongside display.

What's broken. Quality varies. The closed networks remain ahead on premium-publisher relationships.

Stack-rank position: Worth a programmatic test if you're already running a DSP.

Push-notification networks (PropellerAds, RichAds, Adsterra, etc.)

The shape. Push and pop networks are a category of their own — PropellerAds, RichAds, Adsterra, and others operate at much lower CPC ($0.001-$0.02) but with correspondingly lower click intent.

What's working. Cheap. Can drive testing volume. Some real ROI in low-intent verticals (sweepstakes, certain dating, tier-3 ecommerce).

What's broken. Click quality is bottom-of-the-barrel. Most buyers running search-arb-style funnels do not see meaningful conversion from push traffic.

Stack-rank position: Niche. Not a primary native traffic source for most operators.

Skimlinks, AdMarketplace, and the link-monetization adjacent

Skimlinks (corporate site) is a publisher-side affiliate-link-monetization service. Not a buy-side traffic source but worth knowing for the supply context.

AdMarketplace is a search-monetization partner similar to Tonic — relevant to buyers running search-feed arbitrage. They have been quieter publicly than System1 but operate in the same category.

Bodis and Domain Holdings are domain-monetization partners that some search-arb operators use as alternatives or supplements to Tonic.

These are not "native networks" in the strict sense but are part of the operator's full traffic-source stack.

Account-manager experience, ranked

Subjective, based on personal and operator-network reporting:

Outbrain — most consistently competent at $5K+/mo spend tiers
Taboola — generally good at scale, variable mid-tier
Microsoft — competent for branded buyers, less affiliate-fluent
MGID — strong in their core geos, weaker elsewhere
RevContent — variable, often a routing function rather than a partner

Payment terms, summarized

| Network | Advertiser charges | Publisher pays | |---|---|---| | Outbrain | Daily prepaid / net-7-30 managed | Net-45 to net-60 from month-end | | Taboola | Daily prepaid / net-7-30 managed | Net-45 to net-60 from month-end | | RevContent | Prepaid | Net-30 to net-45 | | MGID | Prepaid | Net-30 to net-45 | | Microsoft | Prepaid / managed credit | n/a (closed system) |

Supply quality, ranked

By third-party measurement (DoubleVerify Global Insights, IAS, TAG-certified IVT levels):

Outbrain
Taboola
Microsoft Audience
RevContent (high variance — premium-tier comparable, long-tail much worse)
MGID (similar variance)

What's coming for the industry in the next 12 months

A few predictions, with the appropriate hedge that I am wrong about a non-trivial fraction of these:

Continued consolidation pressure between Taboola and Outbrain. The 2024 merger talks were public. Whether they revive or whether one acquires the other through a smaller transaction, the structural pressure remains.
AI-creative compliance arms race accelerates. Both networks will invest heavily in detecting AI-generated faces, AI-generated income claims, and AI-generated reviews. Expect 2026-2027 to bring meaningfully tighter creative review.
Programmatic native gains another 5-10 share points. The closed networks' dominance of premium supply is structural; the programmatic adjacency for non-premium inventory keeps growing.
Search-arb continues structural compression. RPM tightens. The bar for operating profitably rises. Operators with operational excellence keep working; the long tail keeps being squeezed out.
State-AG enforcement on affiliate native ads expands. California, Washington, New York have all increased consumer-protection activity against affiliate-style advertising. Expect more cases like Roomster and BetterHelp at the state level.
Account-warm-up timelines lengthen on Outbrain and Taboola as platforms invest in trust-graph signals. New accounts will be slower to scale than they were in 2023-2024.

What's quietly changed since 2024

A few structural shifts that don't fit cleanly into the network-by-network rundown but matter for the operator stack:

Cookieless attribution is finally biting. Apple's 2021 App Tracking Transparency and the long-running deprecation of third-party cookies in Chrome have, by 2026, materially degraded the attribution signals available to native networks. Outbrain and Taboola have invested in first-party-data integrations and modeled attribution to compensate. The result is that the conversion-tracking accuracy on these networks is now meaningfully worse than it was three years ago for a class of advertiser without their own first-party-data infrastructure. Affiliate operators who route conversions through their own tracker (Voluum, RedTrack) are largely insulated from this; advertisers relying on the network's own conversion pixel are more exposed.

Server-to-server (S2S) tracking has become standard. Five years ago, S2S was an advanced setup. Today it's a default expectation among scaled buyers. The networks have built better S2S documentation; trackers have built better default S2S templates. The implication for new operators is that the technical bar for "knowing your numbers" is higher than it used to be.

Privacy-coalition signals (TCF v2.2, GPP) are now table-stakes for EU operations. The IAB Europe TCF framework v2.2 is now the de-facto standard for managing user-consent signals, and operators routing EU traffic without proper TCF integration are seeing meaningfully degraded performance. This is not optional anymore.

The DSP-side tooling has matured. The Trade Desk, DV360, Amazon DSP, and several smaller players have all built more native-aware bidding tools over the past 18 months. The "DSP for native" use case is genuine for branded advertisers in 2026 in a way it wasn't in 2022.

LinkedIn and Quora have become real native-adjacent traffic sources. LinkedIn's native ad product and Quora's promoted answers have evolved into legitimate options for B2B-targeted native. They're not Outbrain or Taboola scale, but they're worth a slot in a B2B-oriented stack.

A note on the "next big thing" speculation

Every state-of-industry piece is obligated to speculate on the next big thing. I'll do it briefly and with appropriate humility:

Connected TV (CTV) native. Companies are aggressively pitching "native" formats inside CTV, but the actual operational economics are still poor for affiliate operators. The CPMs are too high, the attribution is too lossy, and the format constraints make creative iteration expensive. Worth watching for 2027-2028 readiness; not yet a viable operator stack slot.

On-platform AI search products. OpenAI's, Anthropic's, Perplexity's, and Google's AI-search products have started monetizing in various forms. Whether they evolve into a meaningful native-equivalent traffic source over the next 12-24 months is genuinely uncertain. The structural fit is plausible — answer-engine-adjacent advertising could work like search-feed monetization — but no clean monetization product has shipped at scale yet.

Direct-to-publisher native. Some publishers have begun selling native inventory directly via their own ad-ops teams, bypassing the closed networks. This is a small share of supply but growing. For high-budget direct-response advertisers it can offer better economics than going through Outbrain or Taboola; for affiliate operators the economics rarely work because publishers prefer the higher-quality direct-brand demand.

I am quite uncertain about all three of these and am happy to be wrong.

Where my read is most uncertain

For honesty: a few areas where I genuinely don't have great data:

Net advertiser margins by network. No one publishes this. Operator surveys are self-reported and biased toward winners.
The direction of CPM evolution at the placement level. The networks publish blended numbers; the underlying distribution is more important than the average and isn't transparent.
AI-creative effectiveness vs human-creative effectiveness, controlled for everything else. Operator reports are mixed. The aggregate creative trends I documented in What 7 million scraped ads tell us suggest AI is at least non-inferior in many categories but the data doesn't conclusively show it's better.

What to actually do with this stack rank

For a buyer reading this and trying to decide where to spend:

If you have a clean offer and budget, start on Outbrain. Higher CPC, higher trust-graph compounding, fewer compliance surprises.
If you have a borderline offer, start on RevContent or MGID to test economics, then graduate clean creatives to Outbrain or Taboola.
If your geos are non-US-tier-1, start on MGID for LATAM/EMEA/APAC.
If you're branded direct-response with a real budget, add Microsoft Audience to your stack alongside the closed networks.
If you're optimizing programmatically across many channels, add Sharethrough or Nativo via your DSP.
Don't run only one network at scale. Cross-network data is the cheapest edge.