Native ads spy tool: what you can actually see in 2026

A native ads spy tool is a crawler-fed database of Taboola, Outbrain, MGID and Revcontent ad units, because none of those networks publishes a public ad archive. The dedicated tools run from $79.99 to $399 a month on vendor pages checked in September 2026. They show creative, publisher placement and run dates. They never show spend or conversion.

Why native ads need a spy tool at all

Meta, Google, TikTok, LinkedIn and Snapchat each publish a searchable library of the ads running on their inventory. Native does not. Taboola reported $1.9 billion of revenue for full year 2025, up 8.3%, with fourth quarter revenue of $522.3 million (Taboola fourth quarter and full year 2025 results, 25 February 2026). The largest native network in the market still has no equivalent of the Meta Ad Library.

The inventory explains the gap. Native units render inside a publisher page as a recommendation widget or an in-feed sponsored story, so the network has nothing to publish centrally. The widget exists only at the moment a reader loads that page. To see it you have to be on that page, in that country, with that device.

That is the product thesis for every tool in this category. A vendor runs crawler nodes across publisher sites known to carry native widgets, captures each unit as it renders, and indexes the creative, the landing page and the dates. It is the same mechanism paid tools use on social, pointed at a network that will not hand over a list.

What a native ad record actually contains

The useful fields are consistent across vendors, even where the interface is not. A record usually holds the creative image and headline exactly as rendered, the sponsored label, the publisher domain, the network that served it, a first-seen date, a last-seen date, and an inferred country and device.

Two columns carry the weight, and only one of them is measured. Observable: headline, thumbnail, advertiser, placement, disclosure wording, landing page URL, run duration. Inferred: whether the campaign is profitable, the budget, the bid strategy, the audience settings, the conversion rate.

Run duration is the most valuable observable in the set. An advertiser that keeps a native campaign live for two months across several publishers is not doing that at a loss. Treat the inference as a ranked signal rather than a measurement, and every honest tool in the category will agree with you.

For the general case across networks, our breakdown of what ad spy tools actually return sets out where each data route gets its records.

The landing page field is where native vendors separate. Native offers convert after the click, on an advertorial, a quiz or a pre-lander, so a tool that archives only the widget misses the half of the funnel that explains why an angle works. Anstrex and SpyOver both state that landing pages can be downloaded, which is why they keep appearing on affiliate shortlists (vendor pages, September 2026).

Which networks count as native in 2026

The core four are Taboola, Outbrain (now part of Teads), MGID and Revcontent. Around them sit Yahoo Gemini native, Nativo, TripleLift and Sharethrough on the premium side, with a long tail of Content.ad, Dianomi, Plista and ZergNet on the performance side.

Vendor network counts are not comparable, because each vendor counts differently. Anstrex lists 28 networks across 64 to 65 countries in its FAQ. Adbeat lists 29 native networks plus 145 display networks across 35 countries. SpyOver names 11 networks across 150 countries (vendor pages, September 2026). A larger total does not mean better coverage of the four networks you actually buy.

Check the named list, not the tally. If MGID is a real part of your competitive set and a vendor's product page does not name MGID, the tally is irrelevant to you.

What native ad spy tools cost in 2026

Prices below come from vendor pricing pages checked in September 2026, as recorded in AdMapix's review of native ad spy tools (updated 7 September 2026). Monthly billing unless stated otherwise.

Anstrex Native, $79.99 a month on the Solo plan, $139.99 for native plus push, $219.99 adding pops. The vendor states no free trial and no refunds for dissatisfaction. It is the lowest entry price among the dedicated native tools, and the landing page ripper is the reason buyers accept those terms.

AdPlexity Native, $249 a month, or $208 a month on annual billing. It adds detection of affiliate and arbitrage networks (ClickBank, BuyGoods, MediaForce, Tonic, Sedo) plus keyword search across landing page content. MGID's own roundup puts the AdPlexity native database at 20 million ads across 60 or more countries.

Adbeat, $249 a month on Standard and $399 on Advanced, with a free Basic tier. It is the only tool here that pairs native with display and exchange coverage across 145 display networks, which is what an agency buying both channels needs.

SpyOver, $149 a month on Professional and $299 on Corporate, priced by simultaneous sessions rather than by data volume.

BigSpy needs naming because it appears in almost every native roundup and does not belong there. Its platform list is Facebook, Instagram, Google, YouTube, TikTok, X, Pinterest, Unity and Yahoo. Taboola, Outbrain, MGID and Revcontent are absent, and its free tier covers Facebook only.

Why the roundups disagree on price

Two 2026 roundups, four months apart, quote different numbers for the same products. The August 2026 native tool comparison puts Anstrex at about $47 a month and AdPlexity at about $199. The September 2026 vendor check puts them at $79.99 and $249. Both cannot describe the same plan.

The likeliest explanation is plan drift rather than error: a lower tier, an annual rate, or a legacy price carried forward in a post nobody re-checked. The $47 figure also matches the price MGID's roundup records for Native Ad Buzz, a different tool entirely, which is how conflation starts.

Treat any price without a plan name and a check date as marketing. The number that matters is the one on the vendor's pricing page on the day your card is charged, and the only reliable way to get it is one click from the vendor's own domain.

How to test a native ads spy tool in twenty minutes

Trials in this category are capped on purpose. A short structured test finds the gaps faster than a week of casual browsing.

Search a campaign you saw yesterday. Pick a native ad you personally saw on a publisher site in the last 48 hours and search for the advertiser or the offer. If the tool cannot find something you saw yesterday, its crawl frequency is too slow for live research.

Click through to the landing page. Confirm the tool captured the actual page rather than a dead link. Native landing pages rotate faster than the ad units themselves, so this is the field that breaks first.

Check your own networks against your mix. A search for Taboola and a search for MGID should both return volume, not a handful of records.

Filter by run duration. If the interface has no first-seen and last-seen dates, the tool cannot support the longevity signal, which is the main reason to pay for any of them.

Export one record. If the data cannot leave the interface as a URL or a file, it cannot join your reporting, and somebody will retype it by hand every week.

Can you research native ads without paying?

Partly, and for a first testing backlog it is enough. The manual route has three steps: start from a small set of publishers in your category, browse with a clean profile in the target country, and log every sponsored unit with its headline, thumbnail, disclosure text and destination page.

MGID's own guidance is useful here on two points. It recommends testing multiple headline and image combinations rather than a single creative, and its policy requires advertorial pages to be identified as advertising. The label an advertiser chooses is itself part of the creative, and it tells you how that advertiser reads the compliance line.

A manual pass costs an afternoon and returns maybe twenty examples. That is enough to answer whether native is active in your vertical. It is not enough to answer what changed since last quarter.

If you are starting a free routine, our guide to analyzing competitor ad performance without expensive tools sets out the weekly cadence that does not collapse after two weeks.

Is a native spy tool worth it for a brand team?

For affiliates and arbitrage buyers the answer is settled. They need volume, landing page rips and geo filters, and the $249 AdPlexity plan is a normal cost of doing business.

For an in-house brand or agency team the maths is different. Native research answers a narrow question, which is what angle a competitor runs on one network. It cannot answer the broader question that usually matters more: whether the same advertiser runs that angle on Meta, TikTok or search in the same month.

That gap is where brand teams stop getting value from a single-channel subscription. Paying $249 a month to watch one network in isolation, while the social side of the same competitor stays invisible, is the most common bad purchase in this category.

The honest rule: buy a dedicated native tool when native is more than about ten percent of your paid mix, or when a competitor's native presence is the one thing you genuinely cannot see. Otherwise put that budget into the channel where you already compete.

Where native spy tools stop, and what to change

No tool in this category shows real Taboola or Outbrain spend, and none has access to a competitor's account. What you get are proxies: ad counts, days running, publisher counts, and on Adbeat's top tier an estimated spend figure. Read those as ranked signals of scale, never as budgets.

Our guide to tracking competitor ad spend sets out how far those proxies stretch before the estimate stops being defensible.

The second limit is structural. A crawler-fed database shows you what the crawler caught. Coverage of the four big networks is good. Coverage of a smaller regional network depends entirely on how many crawler nodes the vendor points at it, so ask for a demo against your own market before you commit.

That is also why API access matters more than interface polish once a team runs agents. A record that can be queried programmatically joins native and social data in one table. A record locked inside a vendor interface does not, and the manual retyping that follows is where most research programmes quietly die.

If the output has to feed a model rather than a person, our comparison of an ad intelligence API and the Meta Ad Library API covers what each route can support.

Nothing in the 2026 vendor landscape suggests a public native ad archive is coming. The closed inventory that made these tools necessary is the same thing that keeps the subscription price up, so the crawler model is the market for the foreseeable future.

What to do next

If you want the dedicated-vendor comparison, we broke down AdPlexity alternatives by buyer type after the price check.

For the tool that appears in most native conversations but is built around social, AdSpy alternatives maps four routes and who each one fits.

Shortlist in one line: Anstrex Native for the cheapest entry with landing page capture, AdPlexity Native for affiliate and arbitrage network detection, Adbeat when you also buy display, SpyOver when country breadth matters more than network depth.

Then run the twenty minute test on a trial and buy the plan whose named network list matches the networks on your own insertion orders. Nothing else in the comparison changes the outcome.

Frequently asked questions