For the past few years, the ad tech industry seemed brimming with optimism. With new devices and media abundantly available to marketers today, advertising seems perfectly poised to make a positive difference for brands. We've seen bright-eyed headlines highlighting the "year of mobile," "rise of programmatic," the "era of wearables," and so forth. So, how, amidst all the creative possibilities, has the issue of ad fraud spiraled out of control in the past year?
The research is damning. Some verification companies are reporting ominous statistics that half of web traffic isn't human, and consequently, brands are losing $6 billion every year to "fake" impressions.
The numbers are certainly incredible. We've diagnosed the problem and, rightfully so, concerns have spread like the plague. Scammers have always existed in the ad space, but the sophistication of modern day fraudsters is starting to make the fraud tactics of yesteryear look like child's play. For this reason, all parties in ad tech -- both buyers and sellers -- are now scrambling for the solution.
When we magnify the problem, we have to take a closer look at several facets: how it seemed to proliferate so rapidly, why it has become ubiquitous in our industry, how numerous parties play roles that support fraud's existence, and finally, what can be done to eradicate the issue.
How fraud came to be
When we think of key performance indicators (KPIs) and success metrics, "clicks" still come to mind. Our metrics are still heavily performance-based, and advertisers still demand to see these numbers. These demands drive vendors to purchase cheap inventory, which may translate to more counted conversions, but little to no increase in sales, revenue, or ROI. And as more and more media are bought and sold in programmatic environments, the rate of ad fraud, as well as the sheer number of transactions occurring daily, is only going to increase due to a lack of human transparency. Open exchanges offer a cornfield of sorts for the fraud industry: The fake sites, or sites riddled with bots, are primed to inflate whatever KPI is needed to hit an advertiser's goal. And with their backs against the wall, buyers have had no choice but to shift budget to those "high performing" environments.
Ultimately, the buying and selling process within the industry has long supported the existence of ad fraud. Wherever media spending is significant, fraud has the potential (and demand) to grow. Only recently have fears reached critical mass due to technological advances on the programmatic front.
Common types of fraud
Though the term "fraud" seems to be a catch-all, it's oversimplifying how sophisticated the problem has come to be.
At a higher level, there exists impression fraud, click fraud, and affiliate fraud. And if we break it down further, we reveal several levels and options where fraud can occur within two main categories: site-level fraud and bot fraud.
This occurs when a botmaster commands its botnet to rapidly render page views on a phony site that has fronted itself as a legitimate publisher. The site owner, often the botmaster, then sells that [fake] inventory to brand advertisers through blind environments. This same method is used when legitimate sites start buying "cheap" traffic from affiliate programs to help boost impression numbers.
This occurs when fraudsters send zombie browsers to a legitimate advertiser's site so as to collect a cookie. After the cookie has been placed on the fake browser, the fraudster instructs the bot to go to their own site wherein the bot is retargeted a higher yielding ad from the advertiser it had just visited.
This occurs when there are multiple ads on top of one another in a single ad placement. These ads are often stacked above-the-fold to make it look "viewable."
This occurs when the page refreshes the ad after minimal view time, doubling or tripling the impression count.
This occurs when multiple cookies are "stuffed," or dropped, on a user -- even if the user has not seen, or clicked on, the brand's ad.
This occurs when a site is designed such that it sends its users between multiple domains in order to artificially inflate impression counts on each respective domain.
This occurs when an ad is "stuffed" into a 0x0, or a 1x1, pixel on a page, still counting the impression even though the ad is invisible to the user.
Incentive services (pay-to-click)
This occurs when real humans are paid to click ads, to drive traffic and boost click numbers, despite having no intention to actually convert or consume content.
Buying traffic (this bleeds into bot fraud)
This occurs when the seller buys a large amount of traffic for cheap in a bid to increase page views, and therefore, impressions.
Who benefits from fraud
While ad fraud hurts the brand, every other party benefits from its existence. This alone has buoyed ad fraud's overwhelming survival in the industry. Bot operators, of course, end up pocketing a significant chunk of the $140 billion of overall digital ad spend. But it's not just the botmasters or fraudulent site owners that benefit. Buyers in the space have long been winning incremental budgets from advertisers by buying artificially well-performing impressions. Open exchanges and supply side platforms (SSPs) are responding to a demand for inventory by buying cheap scale from unknown publishers with limited transparency into the quality of those sites.
With a clearer understanding of why fraud is increasing, the types of fraud, as well as who benefits from it, we can more closely look at the impact fraud is having across the advertising industry. If left unchecked, fraud will ultimately undermine the potential of, and confidence in, digital advertising.
There are six major detrimental effects ad fraud has on our industry:
Fraud is hitting all platforms.
Video and mobile advertising continue to grow and show no signs of stopping. According to eMarketer, web video ad spending will hit 8 billion by 2016. Mobile will hit $18 billion this year. However, as more brands direct dollars toward these areas, they have become larger targets for fraudsters. Video advertising is especially vulnerable as its ad rates are often ten times that of display. Therefore, video is a huge target for fraudsters, with advertisers wasting a reported $6.8 million per month. And with mobile, 40 percent of all clicks are predicted to be fraudulent.
While fraud has always been pervasive in display, the larger issue is that fraud is agnostic, and the monetary incentive is enough to spur proliferation across other well-established platforms like video and mobile.
If fraud is growing alongside these budding media, what's to stop a brand advertiser from avoiding digital marketing, and instead sticking to TV? Nothing.
Fraud is forcing us to rethink inventory sources.
Though they contribute to ad fraud, bots make both media sellers and buyers look great. They help to artificially increase click-through rates and conversions. Unfortunately, since those "clicks" and "conversions" didn't come from a human user, there is absolutely no chance that any additional revenue was seen for the brand. Agencies and brands must evaluate their inventory sources carefully, not letting attractive numbers fool them. The "highest performers" can sometimes attract the most robotic traffic.
Fraud is driving checks and balances.
Growing fears have spurred demands for checks and balances -- finally. The fight against fraud has raised the standard for vendor transparency in the name of brand safety. Ad networks that have maintained direct, transparent relationships with publishers are being rewarded. Premium and mid-tier publishers that have maintained quality inventory -- and have steered clear of buying "cheap" traffic -- will thrive. Advertisers should be willing to pay more for better quality. Everyone else will get weeded out.
Fraud is impacting all publishers.
Even good publishers have bots that they cannot prevent from coming to their site. Therefore, publishers need to know how people are discovering their content. They must analyze and monitor the quality of traffic, especially if they are working with any third parties to help their SEM or marketing efforts. Ultimately, good publishers will likely start to avoid outsourcing.
Advertisers will be expected to pay more.
To this point, there has been very little incentive to eradicate bot fraud because brand dollars, and the agencies that control these dollars, keep pumping money into the companies that are buying the cheap traffic. Agencies and brands need to realize that they can't have it all. They have to ask themselves if they're willing to pay premium rates for clean, safe, human inventory.
Also, if 30 to 40 percent of all inventory is fraudulent, then all advertisers have less access to inventory than we originally thought. As with anything, as the supply shrinks, the demand, along with the price, increases. Publishers can ask for increased CPMs, which may hurt short-term as expectations are readjusted.
Fraud is potentially ringing the death knell for open exchanges.
Will they go away completely? Maybe. But ad fraud has forced open exchanges to put a mirror up to their ways; there is no doubt that media selling within these systems has to change. Expect to see brand marketers and agencies start pulling out in waves, especially now that agency conglomerate GroupM announced it is dropping out of open exchanges entirely.
Perhaps private exchanges could be the dawn of a new era in media buying. It almost feels like a pseudo-back-to-basics strategy, but with the intelligence now to support the infrastructure, the realm of programmatic buying is about to get a hell of a lot smarter.
Ad fraud's biggest impact
Brands have been shifting their ad spend from television to digital for years, especially with the growth of new devices and media. Ominously, the relatively low barriers to entry for a botnet could make brands think twice about how quickly they shift those budgets. Wasted dollars may result in ad spend shifting elsewhere, which is good for no one in our industry, including (and perhaps especially) the fraudsters.
How to eliminate ad fraud
For advertisers interested in helping to eliminate ad fraud over the long-term, there are a number of effective strategies you can take.
- First, take a closer look at your inventory, and only work with partners and publishers you trust. It is key to work with partners who have monitoring and reporting tools in place to help safeguard against fraud.
- Secondly, because fraud has become much more sophisticated, the industry -- and especially the IAB -- has developed a keen eye as to who is good to work with and who is not. Your partners should have processes and regular quality checks in place to illustrate their compliance and commitment to a more transparent industry. Publishers that participate in the IAB's Quality Assurance Guidelines (QAG) program are a good place to start. These companies have made the QAG commitment to audit their sites regularly and are transparent about sourcing their inventory.
- Third, don't hesitate to work with partners who also work with third-party verification services to safeguard against fraudsters.
In the end, ad fraud is, indeed, one of the biggest challenges in the industry today. It's causing enough noise to make buyers and sellers listen to the problem. However, if we work hard to put out the fire, standards for quality, transparency, and safety will only increase.
The ultimate goal here is to ensure all parties in ad tech are responsible for eradicating fraud across all digital media, so advertising may thrive both effectively and safely.
On Twitter? Follow iMedia Connection at @iMediaTweet.
"Hacker typing on a laptop with binary code background" image via Shutterstock.