Validated Actor Inventory (VAI)
Measure whether AI traffic is dragging down CPMs.
VAI gives publishers and ad ops teams a lightweight way to classify traffic and review CPM performance by actor type.
Start with measurement, then act with confidence
Some AI traffic looks enough like normal browsing to trigger ads, but it doesn’t monetize like a real visitor.
VAI helps your team measure whether that traffic is showing up in your inventory and whether it is associated with weaker CPMs or lower-value impressions.
The first step is diagnostic:
- classify likely human and AI-driven traffic
- send those signals into your existing reporting workflow
- review CPM performance by actor type
- decide whether any routing or suppression changes are warranted
This is a measurement-first pilot, not a platform migration.
What VAI does
VAI tags pageviews using two dimensions:
1. Actor type
Who or what is most likely behind the traffic?
- HUMAN — likely a real human visitor
- AI_AGENT — an AI system acting on behalf of a person
- SHARING — preview and sharing systems such as link preview bots
- OTHER — non-human automation such as crawlers, renderers, monitoring, or QA
2. Confidence tier
How confident is the classification?
- ACT-1 — high confidence
- ACT-2 — medium confidence
- ACT-3 — low confidence
These signals let your team compare monetization performance across traffic segments instead of treating all traffic the same.
What a pilot looks like
A VAI pilot is designed to be lightweight and low risk.
Step 1 — Add a tag
Deploy VAI through GTM or a page-level script.
Step 2 — Pass signals into reporting
Send actor-type and confidence values into your ad server, analytics, or reporting workflow.
Step 3 — Review the results
Compare CPMs and related monetization signals by segment.
Step 4 — Decide what to do next
If the data shows meaningful differences, you can test routing, suppression, or deeper integration.
No demand-side participation is required for the initial measurement phase.
What your team gets
With VAI, your team can:
- see whether AI traffic is present in monetizing inventory
- compare CPM performance by actor type
- separate high-confidence human traffic from lower-confidence automation
- make better decisions about whether traffic should be monitored, routed differently, or suppressed
VAI is useful even before any monetization policy changes are made.
Why this matters
Many publishers assume the demand side will filter out poor-quality automation.
But if automation is mixed into monetizing traffic, it can still affect pricing and inventory quality before you have clear visibility into what is happening.
VAI gives your team a way to measure that directly using your own reporting systems.
VAI and IVT tools
VAI does not replace IVT detection or fraud controls.
IVT tools are designed to identify invalid traffic and reduce buyer risk. VAI gives publishers a separate measurement and decision layer that helps them understand how different traffic segments perform and whether they should be treated differently.
In short:
- IVT tools identify risk
- VAI helps publishers measure and act on traffic differences
Lightweight implementation
You can start with a simple script-based deployment:
<script src="https://paywalls.net/pw/vai.js"></script>
From there, actor-type and confidence signals can be passed into your reporting workflow for analysis.
Deeper edge or CDN integration can come later, if needed.
What VAI is not
- Not a promise of immediate revenue lift
- Not a replacement for IVT or fraud systems
- Not a user identity system
- Not an anti-AI blocking product
- Not a requirement to change routing on day one
It is a measurement layer that helps publishers understand whether AI traffic is affecting monetization.
Request a VAI measurement pilot
Start with a lightweight pilot to see whether AI traffic is dragging down CPMs on your inventory.
VAI complements IVT detection and other traffic-quality controls by giving publishers additional visibility into how traffic segments perform.