The rise of programmatic advertising has shifted the focus towards accuracy and automation. It surged from $9.75 billion in 2023 to $12.46 billion in 2024, an annual growth rate of 27.8% and is expected to continue expanding, reaching $28.12 billion by 2028 at a compound annual growth rate (CAGR) of 22.6%. However, with AI coming into the picture, performance programmatic platforms are prone to ad fraud even more. The need for optimization of programmatic media buying with comprehensive ad fraud solution across the advertising funnel is the necessity to yield results.
More and more advertisers are pushing for a stronger and harder success KPI in programmatic campaigns. The shift from visibility only to performance-first is underway. With new and upcoming programmatic platforms selling inventory on impressions, it is today evident that impression fraud is 10-15% of campaign spends in the MENA region, as per mFilterIt reports. There is a ROI uplift of 7-10% when advertisers identify and block for Made-For-Ad sites and Ad Frequency cap violations.
For advertisers, an ad traffic validation tool is the need of the hour to weed out fraud, optimise programmatic traffic and improve the hard KPIs of their campaigns. Also, programmatic platforms & ad networks have started providing ‘Certificate of Verification’ to advertisers to ensure their ad inventories are validated.
Let’s dig deeper into the explore how programmatic ad fraud detection can help elevate performance of ad campaigns and what are the key challenges.
Why Programmatic Ad fraud prevention?
Protect your brand with programmatic ad fraud prevention. Ensuring the invalid traffic is blocked from malicious sources not only safeguards advertising budget but also protect brand reputation.
Here’s how mFilterIt guides with trust and transparency in programmatic advertising:
Impression fraud
Impression Fraud analysis is better at the post-bid stage than pre-bid, measuring performance beyond viewability metrics. In pre-bid analysis, i.e. before the ad is served, fraud can be identified based on only two parameters, IP and User Agents. Also, the time for analysis is limited to 10 milliseconds. This results in a meager 2% fraud identification.
This is where a post-bid analysis trumps a pre-bid impression validation. Now that we have several more parameters fraud detection is done on deterministic and heuristic measures as well. This results in the detection of higher invalid impressions of 15-20%. This results in improved ROI on Ad spending. Post-bid impression analysis is a more beneficial method for detecting ad fraud.
Made for Ad sites
Advertisers spend an average of 15% of their programmatic budget on MFA sites, but some may spend as much as 42%. While 35% of programmatic spending is wasted on low-value environments like MFA sites, according to a recent study by ANA (Association of National Advertisers). By focusing on robust ad fraud detection advertisers can combat the various forms of fraud that undermine their campaigns across digital advertising platforms. Prioritizing impression validation is essential for maximizing return on investment and maintaining trust in the advertising ecosystem.
MFA Sites not only drain budgets but also pose a challenge to a brand’s safety. Limited reach and exposure, no real user engagement misleading clicks, click fraud, artificially inflated metrics, poor conversion rates, low-quality/intent traffic and brand un-safe content tarnished brand image and lead to budget drainage.
mFilterIt identify ad placement on MFA sites with
- Deep Content Analytics: A multi-faceted analysis using NLP & image & video analysis to identify brand unsafe content.
- Advanced AI-ML Sophisticated Algos: AI –ML driven analytics for extraction of meaningful insights, patterns, and information.
- Regional & Contextual Understanding: Local language, cultural nuances and domestic norms lead to overall risk assessments.
- Extensive MFA Repository A collection of websites & metric measurement is gathered with regular updates & feedback loop
Fig. 1: The site has multiple ad-stacked ads with high refresh rates. It’s also brand-unsafe promoting gambling.
Ad Frequency Cap Violations
The most common and often neglected issue is Frequency Capping (F-Cap) violations along with bots spamming impressions for burning media budget. Brands need to be vigilant and identify F-cap violations to make sure their ad reaches the broader and relevant audience and is not seen by similar sets multiple times to generate impressions leading to ad fatigue, not conversions.
- A quick succession of impressions generated from the same google advertising ID. Distribution for a genuine user could be distributed throughout the day.
- These impressions were not only coming so excessively but were also being shown quickly.
- Multiple Impressions in a Short Period. A single GAID generates multiple impressions quickly.
- Impression Injection from subnets which reflect that the usage of device farm to fire multiple impressions. Subnets divide a larger network into smaller, more manageable sections.
- IP Repetition with Same IP, different users. It reflects high chances of fake impressions being injected with different GAIDs.
- Same IP, Different Impressions. This issue is not limited to IP repetition, but it extends further with the same IP generating multiple unique GAIDs and different impressions.
Viewability & Attention metrics
Instead of focusing on a single data signal, check on attention metrics along with viewability encompassing a range of data points. These are processed by a machine-learning model to estimate the probability that a specific media environment and ad creative will capture the attention of a hypothetical audience member.
However, Viewability only itself does not help in taking decisions when it comes to effectiveness or attention. Multiple factors need to be measured, monitored and acted upon swiftly. The Viewability and Attention Model encompasses several key factors that determine the effectiveness of an ad in capturing audience attention. Viewability refers to the percentage of an ad that is actually visible to users and the duration it remains in view.
It must also include:
Viewability Metrices
- % of ad viewability and number of second viewed based on IAB standards
- Display ads should be at least 50% of the ad’s pixels are visible in the browser window for at least one second
- Video ads must be at least 50% of the ad unit is in view for two consecutive seconds
- Larger ads at least 30% of the ad’s pixels are visible in the browser window for at least one second
Attention Metrics
- The position of the ad on the page, whether it’s above or below the fold, and its visibility when the user scrolls or navigates away from the page. Ad is severed on positive coordinates are on page while served on negative coordinate are displayed off page with no engagement or attention.
- Time off the page reflects the amount of time the ad is out of sight, affecting how much attention it can capture.
- Engagement is another critical factor, measured through clicks and impressions, which directly indicate how much interaction the ad receives.
- Device signals, such as screen orientation and whether audio is muted or unmuted, provide additional context to understanding how users interact with the ad.
- Quartile progression tracks how far users progress through video ads, offering insights into how effectively the content maintains attention throughout the entire duration from start to completon.
These factors together form a comprehensive model for evaluating ad performance and optimizing for higher engagement and viewability.
Case: How a Global Media Agency optimized OTT ad Campaigns with mFilterIt viewability attention metrices monitoring
Problem Statement: The increasing concern over performance inconsistencies and ROI left a global media agency running ad campaigns on an OTT platform uncertain about the effectiveness of their campaigns. The agency needed the campaign metrics monitoring and measurement that truly reflect the real audience engagement.
Objective: The key objective was to boost visibility and audience reach. They wanted to gain clarity on the engagement metrics and improve ad campaigns’ overall effectiveness.
mFilterIt Solution: We help them with viewability & attention metrics the help accurately measure brand performance and optimize campaigns. This is done via measuring viewability and attention with quartile progression on multiple factors from start to completion.
Fig. 1: Quartile completion rates progression from start to completion viewability
AI-ML powered ad fraud detection solution and ad traffic validation solution mFilterIt Valid8 took following steps to ensure ad campaign optimization:
- Ensured seamless brand Integrations with the OTT platform to begin analyzing campaign traffic.
- Leveraged AI-ML rule engines to analyze traffic attributes and detect invalid traffic sources, such as bots, fake impressions, and click farms.
- Real-Time Traffic Analysis ensured continuous monitored traffic patterns with deterministic, heuristic and behavioral checks to identify anomalies and prevent wasteful Ad spend on invalid or fraudulent impressions.
- Checks on viewability and attention metrics for accuracy in performance measurement.
Impact ad fraud prevention with Viewability attention metrics monitoring
With mFilterIt, the brand received accurate insights into the campaign’s engagement metrics. This resulted in:
- Weeding out invalid traffic and ensuring that all engagement metrics were based on genuine interactions.
- Insights from engagement data allowed the brand to refine their targeting strategy for more effective reach and higher ROI.
- With clearer metrics and more accurate targeting, it helps optimize its OTT campaigns and enhance visibility and engagement.
Conclusion
AI in programmatic advertising is evolving but it needs to be coupled with advanced AI-ML based programmatic ad fraud detection for effective and efficient results. Our ad traffic validation and fraud prevention tool, mFilterIt has advanced capabilities powered with AI-ML tech to provide comprehensive fraud detection across the programmatic advertising ecosystem. It ensure massive savings on advertising budget by weeding out fraudulent or invalid traffic and active blocking & blacklisting.
