Are They Human? Identifying The Bots

Author, Ryan Murray

As a marketer, there is nothing quite so demoralising as realising just how many of your campaign clicks are fake. The World Federation of Advertising has placed it as high as a third: that’s almost 21 trillion ads a year left unseen by the human eye. Considering the long hours and millions of dollars pumped into digital ads, this is a pretty depressing figure.

However, this is a monster of marketers’ own creation. For 10 years, the industry has poured money into chasing traffic volumes and conversations without considering who — or what’s — behind the clicks.

So while many of us are left sitting tight during COVID-19 lockdowns, perhaps now is the time to pause for reflection and consider the humanity behind traffic, the leads it generates and what is its intrinsic value to brands.

A mobile ambush

In Asia recently, selfie apps have become a major target for ad fraudsters. With the most popular of these apps garnering up to one billion downloads, it’s only natural that where people go, bots follow in pursuit. Our researchers recently unearthed 38 apps that duped users by offering selfie editing and beauty filters, only then to use their devices to call up fraudulent ads.

By the time these apps, all from the same fraud scheme, were removed from the Google Play Store, they had been downloaded on more than 20 million devices, throwing however much advertising money down the drain. Given that advertising expenditure is expected to plummet by 20 percent this year, this is not the time to let so much cash go to waste.

Admittedly it’s a tough game: fraud operations like the selfie scam have become increasingly sophisticated and its players now more innovative and highly adaptive to marketing trends. There is a need for constant vigilance when it comes to identifying and combating fraudulent activity, especially on mobile devices given Asian consumers’ propensity towards their handheld devices. But just how do time-constrained and budget-conscious marketers keep their eyes on this without entering just another game of whack-a-mole?

Millions down the drain

While the online marketing arena has come a long way in combating the prevalence of bot fraud, there is still little room to celebrate high traffic counts. 

A recent study we conducted with The Association of National Advertisers (ANA) estimated that the economic losses due to bot fraud were expected to total $5.8 billion globally last year. That number is hardly surprising when you realise the bots of today are not the familiar, stopped-by-a-Captcha-form critters the industry is used to. This new breed of sophisticated bots live on mobile devices and laptops and continue to work away in the background without the knowledge of the device owner. Worryingly, they also act and share behaviour traits that are alarmingly similar to real-life humans.

What’s more frustrating is when these bots fill out online forms on the back of a marketing campaign, which leads to sales staff contacting scores of leads that don’t even exist. During a three-month pilot programme conducted with a few big brands, we found that millions of dollars were wasted trying to capture or follow up with these increasingly sophisticated bots. To use a common metaphor for ad fraud, that’s a very expensive game of whack-a-mole in play.

Determining humanity through collective protection

One thing is for certain, scanning a list of leads from a web form and trying to deduce which ones are genuine is a colossal waste of sales teams’ time and more dollars down the drain.

What is much more effective is stopping bots from getting into the marketing technology stack in the first instance. This can be achieved by breaking down characteristics of a user session and assessing whether a visitor’s behaviour aligns with that of a human or a sophisticated bot. This is a cybersecurity problem, not a measurement issue so selecting the right partner is critical to ensure fraud is recognised, addressed and eliminated as quickly as possible as bad actors always follow the money. Applying this to specific campaigns will give marketers insight into which campaigns are driving fake traffic to websites. Any sites that are drawing bots should be immediately blocked.

Another key element in determining the humanity of digital interactions is through collective protection. By working together, we can protect against threats we find in one part of the internet, learn from the experience, and develop detection methodologies to benefit the entire ecosystem. It creates a mutual feedback loop of knowledge and insight into bad actors, which leads to new signals that protect everyone. As new organisations join the mission to disrupt the economics of cybercrime, the entire system becomes stronger to win the war against the bad actors behind it.

Getting to the heart of this will have a dramatic improvement in the efficiency of your marketing while making you reassess the way you viewed these metrics. After all, a marketer’s job is only done if campaigns are viewed by real people. Once you start to see your traffic for who — or what — it is you can finally focus on the ultimate goal: driving better business results for your brand.

By Ryan Murray of White Ops