Fighting on the Ad Fraud Frontline: Detecting and Preventing Ad Fraud on CTV and Online

There is arguably nothing more vexing for CTV and online marketers than ad fraud, which can take on many forms. According to Kim Norris, Group Vice President, Advanced Advertising Sales at Spectrum Reach, ad fraud at her company is a high priority problem to solve.

Spectrum recently released a Brand Safety Guide which underscored the degree to which advertisers need to prevent ad fraud and protect their ad spend. According to their guide, on average in 2020, (a year where consumers were spending more time online and at home), small and medium-sized businesses lost almost $15,000 annually on ad impressions that were never seen by actual people. For agencies and their clients during the same period, ad spend losses reached $207,000 annually which was 10 times the amount of Server Message Blocks (SMB) according to PPC Global Click Fraud Report of 2020-2-21.

The types of ad fraud are staggering. Most are primarily IVT (Invalid traffic) which can include bots and other non-human interaction. There are bots which are viruses that drive up useless traffic, domain spoofing where lower quality inventory can be passed off to unsafe or low-quality websites, pixel stuffing and ad stacking which makes ads unviewable, location fraud which feeds ads in undesired regions, cookie stuffing and user-agent spoofing.

For Norris, one of the most pernicious is, “the misrepresentation of inventory. Low quality or short-form inventory is passed as high quality or long-form inventory, and mobile experience is passed as CTV because they are both OTT channels.” Premium inventory is easy to define. “Premium inventory is inventory that is sourced from Spectrum Reach’s brand strategy and marketing,” Norris explained, and added, “It is verified by a third party on the back end to ensure that it ran where they say it ran, that it was seen by a real person, and is optimized towards supply quality metrics such as measurability, viewability, and runs in a brand-safe environment.” This inventory is culled from Spectrum’s premium pool of inventory consisting of owned and operated properties, direct deals, carriage agreements, and partner relationships. This ensures no inventory is being aggregated blindly from indirect marketplaces or open exchanges.

To fully address ad fraud, Spectrum partnered with Moat which is a neutral third-party verification company who, Norris noted, “looks ‘under the hood’ at 100% of the inventory flowing through our system. They will confirm that it was seen by real people, that it’s not bot traffic, that there is no malware and it’s in brand safe content.” In this way, Spectrum does not grade its own homework and relies on a trusted, MRC accredited partner for an external audient of their campaigns.

The decision to go with Moat was a careful process. Bill Sheahan, Vice President of Product management for Spectum, noted that his company, “was in search of a media verification partner with specific capabilities that aligned with its overall goal of providing a brand-safe environment for its advertisers; one that verified that ads indeed reached their intended audience on trusted programming.” Several vendors were vetted to insure that vetting could be conducted across all platforms including CTV, mobile applications and web. “Spectrum Reach ultimately selected Oracle Moat Analytics for their superior technology, service and forward-thinking roadmap. Moat allows us to measure and verify all of our inventory and proactively flag any IVT before it propagates across our streaming ecosystem. As an MRC accredited partner, they are investing in sophisticated detection capabilities that allow us to stay one step ahead of some of the bad actors that are trying to infiltrate our supply chain,” he stated.

In addition to Moat, Spectrum is also the first local media sales organization to become a TAG Platinum member. “The Trustworthy Accountability Group (TAG) is the leading global initiative fighting criminal activity and increasing trust in the digital advertising industry. TAG advances its mission by connecting industry leaders, analyzing threats, and sharing best practices worldwide,” Sheahan noted. This group was created by the 4As, the ANA and the IAB to work collaboratively with companies throughout the digital ad supply chain. “TAG member companies that receive all of TAG's available seals (now just 3 – Brand Safety, Ad Fraud, and Malware) achieve the coveted Platinum Member status. These certifications require in-depth application process as well as third-party audits to ensure they meet ongoing requirements for each particular seal,” he added.

Spectrum’s Brand Safety guide recommends the following actions to prevent ad fraud:

1.       Pick trustworthy partners with good reputations in the industry and use best practices to insure that ad campaigns are free of all types of fraud.

2.       Verify ads using an accredited third party verification partner to be sure your ad inventory, delivery to the advertiser and ad dollars are kept to the highest quality.

3.       Demand transparency from your marketing efforts. Ask to understand how the technology

works, what makes their inventory premium, where your ads are running, and who is seeing them.

And yet, with all of this effort, the industry’s efforts to combat ad fraud are in early stages. “We are still at a point where we are educating our advertisers about ad fraud, that it exists, and what we are doing to protect them. I believe it has increased trust between our company and our advertisers because we are educating them on the ecosystem while also providing them with a solution,” Norris explained.

Fortunately, Spectrum’s ad fraud initiative is well timed. “In the next 2 years, we expect ad fraud to become more sophisticated as more industry players put an emphasis on fighting anti-fraud and weeding out the low-hanging fruit with tools and processes that will drive bad actors further into the shadows,” Sheahan concluded.

 This article first appeared in www.MediaVillage.com

Artwork by Charlene Weisler

Building Programmatic Trust with TRUSTX. An Interview with David Kohl

David Kohl, President and CEO, TRUSTX, has been working on issues like fraud and lack of transparency in programmatic since 2013. As a consultant, “a major media industry trade organization engaged my firm to evaluate and quantify the impact of these issues on their members,” he explained. “I gained a first-hand perspective on these critical challenges and saw a lack of urgency among many in the ad tech side of the business to address what was clearly a growing industry problem. It was then that I realized it was time for me to set out and take on this problem in a more active way.”

Kohl noted that “The complex programmatic supply-chain – originally created to make digital advertising more efficient – instead has made it harder for digital advertising to be effective.”

Charlene Weisler: How did TRUSTX get started and what is it?

David Kohl: I proposed the concept of TRUSTX to industry trade leaders, recognizing the importance of cooperation if we were going to make a positive impact. In September 2016, TRUSTX launched as a collaboration between the publisher trade organization, Digital Content Next, 30 of its members – all premium, professional content publishers – and some of the smartest ad tech software engineers in the industry.

TRUSTX is a purpose-driven programmatic marketplace singularly focused on restoring trust, transparency, and safety to digital media. We are a public benefit corporation (B-Corp) wholly-owned by the premium publisher trade association, Digital Content Next. This corporate structure enables us to go to market as a non-profit, with a mission to restore the promise of value to advertisers and their agencies, and to our exclusive membership of professional news, sports and entertainment publishing companies. We are the only purpose-driven technology company in programmatic media.

Weisler: What do you mean by purpose-driven.

Kohl: Good branding is about purpose and understanding your place in the world:  Why do you exist?  How do you change the world for the better?  What is the true impact you have on customers?  Leading companies are making purpose-driven business decisions around the things like the sustainability of natural resources and advocacy for equality, because it is core to their values and beliefs. 

In digital advertising, it seems that we’ve lost our values, or our sense of purpose. Low price, rather than quality, has been driving many decisions. But there is a high cost to that low cost of media: fraud, hidden fees, murky business practices.  Particularly in the last few years, we have let technology lead and we’ve lost not just control and oversight, but also a commitment to a baseline set of values in how we trade media and hold partners accountable.

Weisler: Who are you are your direct competitors?

Kohl: While we compete with other SSPs when it comes to supply choice, we do not have a direct competitor in how we enable 100% viewability across our premium marketplace. Unlike the major SSPs, we have a real-time viewability capability that pre-certifies an impression before it’s billed. Our methodology enables us to guarantee viewability on every individual impression. No averages. No samples. No questions of which impressions were and or were not seen by humans. Our approach ensures the cleanest, most effective path to quality inventory, and it has two clear benefits:

·         For brands, we not only eliminate any media costs for non-viewable inventory, our approach reduces buy-side technology and data costs, which results in higher media buying power on every dollar spent. 

·         Since we measure viewability on every impression, buyers benefit from real-time pacing to human and viewable inventory. This eliminates the time and resources normally spent reconciling impressions and managing makegoods.

Weisler: What are the areas of Programmatic that are untrustworthy? How does your company improve this and increase trust?

Kohl: Opaque trading practices, high fraud rates, low viewability and a general lack of accountability are big flaws that have given programmatic advertising a bad name. With TRUSTX, an impression must be certified as human and viewable in accordance with MRC standards before an advertiser is charged. The approach ensures that there is no waste. Buyers also experience 100% transparency down the publisher URL, along with 100% financial transparency from bid to delivery, reporting and billing.

Weisler: Please explain the sales process from planning to stewardship.

Kohl: At a tactical level, we try to educate our clients and prospects on what 100% human and viewable inventory can do for their campaigns. With average viewability at around 50% on most exchanges, that means that only half of the impressions in a campaign are seen; the other half are wasted.  

Spending time on comparative CPMs is important. The ability for TRUSTX to deliver 100% certified human and viewable campaigns may look more expensive at first, but once buyers understand the apples-to-apples comparison and see the results on branding or sales, they realize how cost effective we actually are.

After a campaign is live with TRUSTX, we monitor the campaign and offer bid reporting, insights on placement optimization and guidance on publisher supply.  Additionally, we offer our clients Media Buying Power scorecards and summary reviews that help them understand the full value of their media investments. Several clients have taken advantage of our transparency promise to execute custom studies, knowing we’ll provide both buyers and sellers with granular, transparent reporting. 

Weisler: What does the future hold?

Kohl : In the next five years, we anticipate the media supply chain to transform dramatically, with more control in the hands of buyers and sellers. TRUSTX intends to be how buyers engage with premium publishers in the evolving programmatic ecosystem.

This article first appeared in Mediapost

The Fraud of Fake Influencer Marketing

Fraud remains one of the top concerns for marketers and digital companies. A new study by cybersecurity company CHEQ found that 15% of all influencer ad dollars are spent on fake followers costing brand advertisers a projected $1.3B globally in 2019.

A study titled The Economic Cost of Bad Actors on the Internet; Fake Influencer Marketing in 2019 by Professor Roberto Cavazos, Merrick School of Business at the University of Baltimore, reports that the total cost of fraud and the resulting loss of trust, “extends beyond the affected business. Entire sectors and economies are adversely impacted.” Influencer marketing spend is big and growing; Mediakix reports that 2019 influencer marketing spend is $8.5 billion globally and is expected to reach up to $10 Billion in 2020, up twentyfold since 2015 when spend on social influencers globally was $500 million. 

When it comes to Fake Followers, there are four main ways to game the system:

Automation can be used to build followers to appear as if they are among the top influencers. Nik Speller, head of campaigns at influencer marketing agency Influencer says: ‘‘They use  automated services, to act on your behalf to follow people, unfollow them again, like content, comment on content, just to do all that underlying stuff that an account has to do sometimes to grow and do it in a  turbo-charged way. In this way all of a sudden you follow 500 people, 200 follow you back, you unfollow them, you’ve got 200 followers and it looks like you are important because your follower number is bigger than the number of people you follow.’’ Speller believes this practice is fraudulent and against social media network’s terms of service as it consists of spamming. 

Use of “Pods” which allow influencers to trade engagement back and forth on each other’s posts, as part of a community. This involves one influencer commenting on or liking a certain number of posts, and is reimbursed in kind with comments on their own activity and posts. Cavazos noted that, “This is clearly not what brands have in mind when creating and paying for campaigns.” 

Sponsored posts. Some bad players publish what is purported to be sponsored posts on behalf of brands they are not actually working with. “In some cases, they engage in this dubious behavior with fake sponsored posts to dupe brands into believing they have a proven track record – and in order to get hired for a future engagement,” according to Cavazos’ study. 

Goosing Inactivity. Even real followers can be problematic when audience inactivity on many networks is considered. In a situation where 30% of some social media accounts have been claimed to be inactive, digital analyst Brian Solis says, ‘‘Many influencers have no access to 90% of their audience simply because it no longer uses the social network where they were followed. This doesn’t stop them from touting millions of followers, who will, of course, never see your content.’’ 

Complexity is a well-known cause of fraud in many domains. Cavazos explains that, “Ad fraud losses to the economy are expected to be more than 20-times the costs of Influencer marketing losses in 2019. However, the growing popularity of influencer marketing brings with it greater opportunities for fraud as the number of players increase and more players are involved around this economy across an expected 4.4 million influencer promoted posts in 2019. The opportunities for accountability and monitoring diminish as the sector increases. In addition, the industry is moving towards greater depth and complexity through other means, with so-called micro influencers and nano-influencers joining the payroll of brands.”

While there are efforts to combat fake influencer fraud, marketers should be aware that we are not there yet.

This article first appeared in Cynopsis.

Research and Fraud: An Interview with the ARF’s CRO, Paul Donato

Research and fraud: As Data Science is increasingly adopted in media companies, research departments, especially in those areas dedicated to quality research such as focus groups, are feeling the pinch. Not only can pure data focus attention and resources on real time reporting, the margin of error in its output can be minimized with protocols such as blockchain. The two seemingly different but actually related topics of quality research and ad fraud were discussed by Paul Donato, CRO, the ARF at SXSW:

Charlene Weisler: Is there room for quality research in this age of data driven decisions? 

Paul Donato: Quality and quantitative research lie on two different dimensions. Qualitative and quantitative are two ends of a single dimension, and both are changing dramatically.
The quality of the research is essential for both qualitative and quantitative. Social listening is electronic, real time, continuous and benefits from very large samples. Sentiment algorithms have improved dramatically. Open source APIs for natural language and sentiment and the abundance of data on the social networks have changed qualitative in a way believed to have impacted the last presidential election.

Quantitative data is now drawn from online behaviors, loyalty cards and other sources of consumer and media data.  However, for many applications, panels and surveys are still needed for modeling holes in the data and demographics of the digital signals.  Because, digital data often lacks the depth of what we can collect in surveys, the quality of the models that fill in that data is essential. There is so much data available, with so little measure of accuracy. Therefore, methods of assessing quality are more important than ever.

Weisler: Will fraud ever go away and if not, can we tamp it down?  

Donato: Possibly. There are many who are testing blockchain as an immutable ledger of the digital supply chain that can eliminate most of the fraud. Maybe, the major digital companies are working on a device and person’s identity with the possibility of a universal ID after there are no cookies. However, this is looking out 3 to 5 years.

Today, many things are happening to tamp down fraud.  Digital is here to stay, so advertisers are taking control of their digital campaigns. Consortiums such as TAG required registrations that support safe and valid traffic. Standards such as those set by the MRC have supported independent firms that specialize in identifying invalid traffic. Viewability standards will help tamp down fraud. Advertising will find a way, so it is a matter of time before fraud is less important than most other challenges in marketing.

This article first appeared in www.Cynopsis.com

It’s An Advertising Jungle Out There. Protecting Brands from Ad Fraud

In the world of cybersecurity, experts are finding that both the demand and supply side of business requires greater protection from unsavory advertising environments and malicious advertising threats, eliminating digital risk while increasing ROI on ad-spend. 

To that end, CHEQ and RiskIQ formed a partnership that advances the industry, creates visibility and protects assets across the entire digital advertising supply chain, affording advertisers, platforms, and publishers a greater sense of security and control over their online presence.

Daniel Avital, Chief Strategy Officer, CHEQ noted, “The rise of ad fraud has become a major threat, with up to 30% of all online ads served to bots, rather than humans, and ad fraud costing $51million a day in 2018.” In addition, he asserts that “brands are increasingly facing the threat of their ads repeatedly being served against violent, graphics and offensive content, such as rape or terrorism significantly damaging their brands.”

Launching the first of its kind cybersecurity-based study for the ad industry (as opposed to traditional ad-verification companies from the ad-tech space), CHEQ, examined four billion ad requests from October 2018 to February 2019, and found the following:

• While just 18 percent of that online ad traffic was deemed fraudulent, 77% of it was considered “sophisticated invalid traffic” (SIVT).
• SIVT vastly outnumbers more basic forms of ad fraud “general invalid traffic” (GIVT). Therefore, CHEQ concludes that ad-fraud threat is much greater than many imagine.

To help analyze and combat this ad fraud, CHEQ realized the need to develop more than 700 unique parameters to block ad fraud in absolute real-time including the use of Dynamic Code Patching, OS and device fingerprinting, and honeypots (bot traps) performed on every single impression. Adding to this effort is the use of AI-based image recognition to help reduce brand safety concerns where ads are placed near videos or sites served in unsavory contexts (for instance graphic and offensive content).

“The use of AI and more robust cybersecurity techniques is vital in moving away from simplistic measures relied upon by the industry such as using blacklisted keywords, or in the case of ad fraud, buying static IP lists purchased from data centers covering only 3-5% of all fake traffic,” stated Avital, adding, “Using these latest advances we can prevent almost all fraud and brand safety issues in real time.”

This article first appeared in Cynopsis.

Military-Grade Ad Verification. Interview with Daniel Avital, CSO CHEQ

With all of the industry talk about ad verification, it may take a former Israeli intelligence officer to develop a military-grade protocol to detect and prevent ad fraud. 

CHEQ, founded by Guy Tytunovich, is positioning itself as “a cyber-security company looking to replace traditional ad-verification platforms with a fully autonomous and pre-emptive solution.” It is certainly an area of importance for agencies and advertisers who are concerned about brand safety, ad viewability and fraud prevention.

I sat down with CHEQ’s Chief Strategy Officer, Daniel Avital, to find out more:

Charlene Weisler: How does CHEQ prevent ad fraud?

Daniel Avital: Every time a user is about to be served an ad, our system rapidly analyzes almost 700 parameters and sets multiple honeypots (bot traps) to determine the authenticity of that user and ensure that only human traffic gets served. The name of the game is speed – after analyzing all of that data and making a decision, we need to prevent that impression from being acquired by our client. All of that needs to happen in record speed.

Charlene Weisler: Does it protect privacy?

Daniel Avital: Ultimately, our mission is to protect the digital ecosystem and privacy is a big part of that effort. To that end, we are fully compliant with the new GDPR regulations and have brought on board designated personnel to ensure that our privacy practices continue to be protective of user privacy, which we believe to be a basic right.

Charlene Weisler: How does CHEQ work in real time?

Daniel Avital: To enable real-time prevention, our engineering teams have had to continuously optimize our algorithms to ensure all our fraud prevention and brand-safety NLP (natural processing language) modules run simultaneously, cost-efficiently and at great speed. This allows CHEQ to make an accurate decision before an ad is ever served.

Charlene Weisler: Could it be said that it is in the programmatic space or works with programmatic in some form?

Daniel Avital: We cover all digital advertising environments, including programmatic, whether it be display or video - where we are introducing unique and cutting-edge capabilities to the industry.

Charlene Weisler: Can CHEQ be used for other media applications?

Daniel Avital: Absolutely. At this point in time, CHEQ’s focus is on transforming the ad-verification space (ad fraud, brands safety and viewability), but our ultimate goal is to help advertisers take back control of their digital media, a goal that transcends ad verification. Our R&D teams are developing exciting applications for our tech which will give advertisers new tools like Autonomous PR (crisis management, reacting to competitors, breaking stories).

Charlene Weisler: How does the technology work?

Daniel Avital: For fraud prevention, CHEQ's platform applies various proprietary modules to detect suspicious non-human activity, data discrepancies and behavioral anomalies. On the brand safety side, we apply advanced NLP (Natural Language Processing) modules to intelligently analyze content at page level and ensure advertisers only serve ads alongside brand-safe content.

Charlene Weisler: What advice would you give a digital advertiser?

Daniel Avital: Stress-test your ad-verification vendors and don’t blindly trust the accuracy of their measurement and reporting. Do not put up with a vendor who offers a “black-box” solution. Always ask them to provide you with raw data (flagged IP’s, flagged “unsafe” URL’s, traffic sources etc.) so that you can assess their work with an external party.

Charlene Weisler: Where do you see your part of the business going in the next 3-5 years?

Daniel Avital: We want to put the control back in the hands of advertisers. Today the hottest issues are brand safety, ad fraud and viewability but in the next few years we plan on transcending the verification space and launching innovative products which will help advertisers get an even stronger grip on their digital presence.

This article first appeared in www.Mediapost.com