Admetrics: Bid Intelligence

admetrics observes vast amounts of signals and metrics of all kinds of ad traffic. A recent analysis of tens of billions of impressions concluded that at least 30% of ad traffic, even from from well-known networks and SSPs, is actually not worth buying.

Modifying bids based on admetrics’ Bid Intelligence data constitutes a vast potential to increase profitability, especially for businesses that buy impressions and bill clicks, conversions or installs.

Bid Intelligence

How it works

Our technology measures and analyzes hundreds of signals per impression to detect non human traffic, viewability and attention, brand safety, context and app quality. By detecting traffic anomalies and fraudulent usage patterns in traffic, frequency and engagement signals across impressions, devices and networks the coverage and accuracy of the data set gets constantly increased.

Custom data, like MMP data from Adjust, Appsflyer, Tune, Kochava, etc., can be ingested to improve the efficiency of our algorithms.

Through admetrics’ low latency, co-hostable S2S service the solution can be applied pre-bid to qualify bid requests.

Background

Available Bid Intelligence Data

+

−

Non-Human Traffic Assessment

ARPA, admetrics’ fraud risk assessment algorithm, scores each measured impression/device by confidence level. Scored results are available for IPv4/v6 addresses, IDFA’s, Apps and Domains.

+

−

App Quality Data

There is a close relation between user perception of an app and ad performance - users are less likely to engage if an app is cluttered with ads. Admetrics app quality scores are based on app reviews, app store ratings and do consider recency. App data is continuously updated and re-assessed to reflect the latest changes in user perception.

+

−

Contextual Data and Brand Safety

Contextual relevance and brand safety ensures the ads are displayed in relevant and brand-safe environments. Leveraging admetrics' natural language processing and image classification services ensures that brands will not appear in the context of violent, adult or otherwise inappropriate content.

+

−

Viewability and Attention Data

Viewability, view time, user engagement and other attention based metrics do impact ad performance. Leveraging attention data when placing a bid can drive superior performance, especially in a world where advertisers are moving from “last click” attribution to more sophisticated “multi-touch” models.

Two parts which, in combination, complete the feedback, training and validation loop

Leverage Bid Intelligence Data in Real-Time

Allows modification of bids based on insights provided by a pre-bid S2S reputation service including:

Non-human Scores for IP’s, IDFA’s, Domains and Apps
Brand Safety Scores for Domains, URL and Apps
Viewability, Attention and Engagement Scores for Domains and Apps
Quality Scores for Apps

Post Impression Measurement and Validation

To continuously classify and validate traffic the implementation of a post impression measurement script is required. Results will be available near real-time in the S2S service to apply the bid intelligence data (1.).

Sophisticated Non Human Detection
Measurement of Viewability, Attention and Engagement metrics
Assessment of Context and Brand safety

Try Out Bid Intelligence

Take part in a free trial with no commitment and find out how much saving potential there is.

Get in touch

Removing programmatic spend on fraudulent or low quality traffic