NativeWhich Taboola publishers send real traffic?
Not all Taboola traffic is equal. ValidVisit scores every visit 0–100 and pins it to the exact publisher that sent it — so you can tell real humans from bots and invalid clicks, worst publishers first.
site in TaboolaThe buyer takes the unique site names from their Realize Sites report and adds them to the campaign's Block Sites list (Advanced Options, or in bulk via Excel/Backstage API), capped at 1,500 blocked/targeted sites per campaign and 1,500 at account level.
ValidVisit reports the device, OS, browser — down to the version — plus the language and ISP behind every flagged visit, and Taboola supports OS version, browser, device type and connection type targeting. The segments we flag are segments you can exclude.
How invalid traffic slips in through Taboola publishers.
Taboola places your creative inside the "recommended content" widgets that appear at the foot of editorial pages across tens of thousands of publisher domains. The supply chain behind those placements is wide: direct publishers sit alongside resellers and content-farm operators, all surfacing clicks under the same sponsored-content wrapper. Taboola exposes a {site_id} on every click that maps to the specific publisher running the widget, and a {click_id} that ties the session to your campaign record. The challenge is that publisher quality across Taboola's open network varies sharply, and the widget format's visual context — credible editorial surroundings — offers cover for arbitrage operations that drive cheap, low-intent, or bot-assisted traffic onto pages and then monetize via the widget grid. ValidVisit takes a different approach to telling those apart: every inbound visit is measured against 100+ independent data points that span the network the click came from, the device sitting behind it, and the way the visitor actually behaves on the page, and all of that collapses into one 0–100 quality score for that single visit. Genuine readers clear the bar; automated and arbitraged sessions surface against it. Because each visit carries its own score, you can separate a publisher whose audience is merely disengaged from one whose click volume has a structural IVT problem.
Native widget traffic on Taboola carries IVT patterns that are shaped by how the supply chain actually operates, rather than by the surface format. The most persistent issue is arbitrage-driven inflation: certain publishers acquire visitors through cheap display or push buys, use those arrivals to build widget impression volume, and a share of the resulting sponsored-content clicks come from visitors who never had any interest in the ad — or from automated sessions that exist solely to generate click revenue. Because these sessions arrive through a real browser on a real publisher URL, crude IP-only filters rarely catch them. ValidVisit reads the network origin of each arriving click as part of its scoring, which is exactly where arbitrage shows its hand on Taboola: these publishers frequently route traffic through residential proxy pools to mask a datacenter or server-farm source, and that routing leaves a mark in the wider set of signals long before any single filter would flag it.
A second pattern distinct to the widget format is automation that doesn't move like a person: a meaningful slice of bot click traffic comes from tooling and stripped-down browser builds that behave nothing like a real reader once the click lands — the connection characteristics, the device profile, and the on-page activity simply don't line up with a human on the OS and browser they claim to be. On native networks, where the click path threads through several extra hops before landing, those inconsistencies are even easier to read, because the longer the journey the more places the mismatch has to show up.
A third, lower-volume pattern is low-intent human traffic misattributed as engaged: visitors bounced from pop or extra-hop buys onto arbitrage publisher pages, who interact with the widget reflexively. ValidVisit's scoring tells this apart from automation — the session looks human across the network and device signals, but its depth and the texture of its engagement read as involuntary arrival rather than genuine interest. That distinction matters because the remedy is different: a bot-heavy {site_id} warrants exclusion, while a low-intent human {site_id} may warrant a bid reduction rather than a full block.
What to watch on Taboola
Segment your ValidVisit report by Taboola's {site_id} token. Publishers driving a disproportionate share of your click volume alongside low quality scores are the primary signal of widget-level arbitrage. A single {site_id} whose score profile sits well below your campaign baseline warrants manual exclusion in Taboola's Site Breakdown report before that publisher's volume distorts your SmartBid signals.
For each publisher, look at whether a visit's poor score is driven mostly by where it came from on the network side or by how it behaved once it landed. A {site_id} whose low scores trace almost entirely to its traffic source points to a structural sourcing problem — the publisher is acquiring sessions through proxy or datacenter routes. One where the weakness shows up mainly in on-page behavior suggests a smaller bot operation rather than a wholesale arbitrage setup, and may be manageable with a lower bid rather than full exclusion.
If one ad variation on a given {site_id} draws a disproportionately low-quality click share while another variation on the same publisher draws clean traffic, the pattern is more likely publisher-side automation hitting the widget indiscriminately than anything about the creative itself. This comparison helps you rule out creative-level explanations before escalating a site exclusion.
Automated click activity on native networks often concentrates outside normal content-consumption hours, when real editorial audiences are not reading. If a high-volume {site_id} shows quality scores well below your campaign baseline in low-traffic windows but looks clean during peak hours, that time-pattern is itself a diagnostic signal worth including in your manual review before deciding on exclusion.
Pinpoint the bot publishers & placements in Taboola.
Taboola itself isn't the problem — bots and invalid traffic concentrate in a handful of its sub-sources: the publisher, site or zone, and the placement or widget within it. So we roll the score up by those Taboola tokens, not by creative (which says nothing about whether a click was human).
Bought as one Taboola line, your spend is a single number. Scored per sub-source it runs from 93 down to 15 — the worst is nearly all bots. That’s the leak a blended average hides.
Taboola traffic scored 0–100 per sub-source, worst first — down to the placement you buy.
Bot / invalid-traffic score broken down by:
- Site (Publisher)
{site_id}
Per-click id: Taboola passes a unique click id, so we also run velocity, deduplication and repeat-source checks on every visit.
Compare bot & invalid-traffic breakdown across every ad network →See your own Taboola sub-sources scored this way.
How ValidVisit attributes Taboola traffic
Each Taboola macro maps to a normalized parameter, so every scored click is pinned to the right campaign, creative and publisher.
https://yoursite.com/landing?utm_source=taboola&utm_medium=native&vv_campaign_id={campaign_id}&vv_ad_id={campaign_item_id}&vv_publisher_id={site_id}&vv_click_id={click_id}| Token | Taboola macro | Maps to | Identifies |
|---|---|---|---|
| Campaign ID | {campaign_id} | campaign_id | campaign |
| Ad / Item ID | {campaign_item_id} | ad_id | ad |
| Site (Publisher) | {site_id} | publisher_id | publisher |
| Click ID | {click_id} | click_id | click |
- Taboola macro
{campaign_id}- Maps to
- campaign_id
- Identifies
- campaign
- Taboola macro
{campaign_item_id}- Maps to
- ad_id
- Identifies
- ad
- Taboola macro
{site_id}- Maps to
- publisher_id
- Identifies
- publisher
- Taboola macro
{click_id}- Maps to
- click_id
- Identifies
- click
How the detection works.
Data points → one score
Every visit is weighed against more than a hundred independent data points and reduced to a single, sortable 0–100 quality score.
Many angles, combined
Each data point is combined rather than checked in isolation, so a genuine human almost never trips enough of them to be flagged — and bots that beat one rarely beat the rest.
Proprietary, not a black box
The detection model is ours and stays that way. What you get is a clear verdict on every visit — not a single brittle rule you can game, and not an unexplained number you can't act on.
Pinned to the source
Every verdict maps to the campaign, publisher and placement that sent the click — so you know exactly which source to cut.
How ValidVisit helps you cut fraud and bad sites on Taboola.
Scoring and attribution are the means — the point is cutting the Taboola traffic that wastes your spend. Here's how ValidVisit gets you a list you can act on.
- Score
See what's actually landing
You buy Taboola clicks; what arrives are visits. ValidVisit scores each one 0–100 so real humans stand out from bots and invalid traffic — one script, no funnel hop, no fingerprinting.
- Attribute
Pin the fraud to its source
Every scored visit is tied to the exact Taboola site and zone via the network's own tokens — so the bad traffic has an address, not just a headline percentage.
- Cut
Take the sites off your buy
You get the worst offenders as a ready-to-use list plus postbacks to your tracker — so you can exclude them in Taboola and put your next dollar behind the traffic that converts.
Taboola traffic quality, answered.
Does ValidVisit integrate directly with Taboola's site exclusion tool, or is the process manual?+
The process is manual. ValidVisit scores every visit and surfaces the {site_id} values with weak quality scores in its reports and dashboard. You export those publisher IDs and paste them into Taboola's Site Exclusion list inside Campaign Manager. There is no automated push from ValidVisit to Taboola's platform — the workflow is: score in ValidVisit, identify the problem {site_id}, exclude it in Taboola. This keeps you in control of the exclusion decision and avoids unintended blocking of publishers that may have isolated IVT spikes rather than structural problems.
Taboola's SmartBid already learns from conversion data — won't it self-correct around bad publishers?+
SmartBid optimizes toward the conversion signals you send back to Taboola, so it can only self-correct if those signals are clean. When invalid clicks from a poor-quality publisher touch your conversion pixel or trigger your thank-you page, they register as conversion-adjacent events in Taboola's feedback loop, which may actually cause SmartBid to increase spend toward that publisher rather than reduce it. ValidVisit sits outside that loop entirely: it gives you a {click_id}-level quality score that you own independently, so you can identify which publishers are corrupting the optimization signal and remove them before the algorithm treats bad data as a positive learning.
Which Taboola tokens give the most useful segmentation in ValidVisit?+
The {site_id} is the highest-leverage token because it maps directly to the publisher generating widget traffic, and it is the same dimension Taboola surfaces in its own Site Breakdown report — meaning you can cross-reference ValidVisit quality scores against Taboola's impression and click data in one view without any translation step. The {click_id} provides session-level deduplication and ties each scored visit back to a specific Taboola click record. Adding {campaign_id} and {campaign_item_id} lets you confirm whether a low-quality pattern is site-wide or isolated to how a particular creative is being served on that publisher, which changes the exclusion logic meaningfully.
Detect fraud on other native networks
Find the bots in your Taboola spend.
See which publishers and placements send real buyers vs bots — every visit scored 0–100, worst first.
Free trial at launch · just your email
One script · no cookies · no fingerprinting · raw IP never stored