During a recent chat to a small business in Port Elizabeth wanting to target prospects in a specific area the topic of Click Fraud came up. Of the billions of adverts served every day the number of invalid clicks on adverts must be quite high. My clumsy fingers have made a mistake on occasion when browsing on my smartphone.
One needs to make a distinction between unwanted (invalid) clicks and fraudulent clicks. As fraud includes the deliberate intent to do harm by a human being it is difficult to actually determine which of the millions of invalid or mistaken clicks are actually fraudulent. In 2007 Google did release an estimate of fraudulent clicks received via their network.
But, to answer the burning question posed by my PE Client; “I have yet to see an advertising network that can accurately deliver completely targeted adverts to only certain narrowly defined geographical areas AND ensure that you don’t get unwanted clicks from people outside of that area.”
Frankly, the only way to ensure that your message reaches your geographically defined target market will be if you physically visit each and every resident. Or, you could troll through the white pages of the directory and phone every person in your area. At least then you won’t need to worry about your flyers being thrown in a bin, your Facebook or Google Ads being shown to and clicked on by people in Outer Mongolia and money wasted.
BUT, remember the wise words of John Wanamaker (1838-1922), a very successful United States merchant, religious leader and political figure who is considered by some to be a “pioneer in marketing” and is credited with coining the phrase “Half the money I spend on advertising is wasted; the trouble is I don’t know which half“.
The Extent of Click Fraud:
Determining Click Fraud is a whole science which is fraught with assumptions and can lead to many people reporting different numbers as the samples from the three reports identified here show.
The 2016-17 Bot Baseline Report, conducted by the ANA (Association of National Advertisers) and White Ops, concluded that fraud levels are down 10%, from $7.2 to $6.5 Billion, compared to the results of a similar study performed a year ago.
Full Report Here: http://www.whiteops.com/downloads/bot-baseline-2016-2017#downloadSection
In another report released in March 2017 on http://www.cnbc.com/2017/03/15/businesses-could-lose-164-billion-to-online-advert-fraud-in-2017.html it is estimated that ad fraud will cost brands $16.4 billion globally this year, and that nearly 20 percent of total digital ad spend was wasted in 2016.
So-called invalid traffic, where bots rather than humans view or click on adverts on websites, was estimated to cost advertisers $12.5 billion in 2016 by ad verification company Adloox.
Online advertising campaigns bought using automated, or programmatic, technology are at higher risk of fraud, according to The&Partnership’s research. It calculates that 29 percent of the $27 billion spent on this kind of advertising globally in 2016 was on invalid traffic, equivalent to $7.8 billion.
A third March 2017 report referenced on http://www.businessinsider.com/ad-fraud-estimates-doubled-2017-3 saying; “The amount of global advertising revenue wasted on fraudulent traffic, or clicks automatically generated by bots, could reach $16.4 billion in 2017, according to a new study commissioned by WPP and cited by Business Insider. That figure is more than double the $7.2 billion the Association of National Advertisers estimated would be lost due to ad fraud in 2016.”
Click Fraud is defined as: “The practice of repeatedly clicking on an advertisement hosted on a website with the intention of generating revenue for the host site or draining revenue from the advertiser.”
There are two primary incentives for committing click fraud:
- Advertisers may try to attack competitors by raising their costs or exhausting their budget early in the day.
- Publishers may click ads appearing on their own websites in order to inflate revenue.
The most common methods for carrying out click fraud attacks are:
- Manual clicking
- Click farms (hiring individuals to manually click ads)
- Pay-to-click sites (pyramid schemes created by publishers)
- Click bots (software to automate clicking)
- Botnets (hijacked computers utilized by click bots)
This video claims to show a Chinese ‘click farm’ where workers tap away on tens of thousands of iPhones all day long:
Russian man visited Chinese click farm.They make fake ratings for mobile apps and things like this.He said they have 10,000 more phones pic.twitter.com/qE96vgCCsi
— English Russia (@EnglishRussia1) May 11, 2017
Facebook and Google are two of the main sites that face accusations of Click Fraud.
Google’s own estimate of INVALID clicks is easily found using a ….. Google Search (official click fraud estimate from Google) and Google says:
“The vast majority of all invalid clicks on AdWords ads are caught by our online filters. These filters are constantly being updated and react to a wide variety of traffic patterns and indications of click fraud attacks. On average, invalid clicks account for less than 10% of all clicks on AdWords ads. At our current revenue run rate, the aggregate value of the clicks that we’ve identified as suspicious or invalid and excluded from what we’ve charged advertisers is in the hundreds of millions of dollars.”
BUT, Google makes a distinction between INVALID Clicks and Click FRAUD. In 2007 the amount of click fraud on the Google Adsense network was estimated at 0.02% of all clicks. See: https://adwords.googleblog.com/2007/02/invalid-clicks-googles-overall-numbers.html
Facebooks official click fraud estimate is unknown to the writer as Facebook’s approach seems to be one of ‘blame the other guy and things not under our control’. In a post on Facebook Business Facebook says: “If you use a third-party reporting tool to gather insights about your ads, your reporting may not match your Facebook Ads reporting.”
Facebook then cites some common reasons why which include:
- Cross-device reporting – Third-party reporting platforms aren’t able to measure cross-device conversions well due to cookie-based measurement. Because people are their real selves on Facebook, conversions can be captured more accurately in Facebook reports. Cross-device conversions from ads are common. Mobile conversions tend to be underreported by third-party platforms that measure with cookies.
- Clicks vs impressions – If a person sees a Facebook ad for your product and doesn’t click, but later browses your website and decides to purchase, Facebook is able to attribute this conversion to the Facebook ad the person saw. Third-party platforms are unable to capture these.
- Time of impression vs time of conversion – Facebook shows conversions in your reports based on the time of impression and not the time of conversion. For example, if someone sees your ad and then buys a product on your website a few days later, we show this purchase based on the time someone saw your ad and not when they purchased your product.
- Attribution windows – Check that your conversion attribution windows match. By default, the Facebook conversion reports are set to a 1-day after view or 28-day click window.
- UTM Parameters (Referrer Links) – Many third-party tracking providers use referrer URLs to credit conversions back to ads. Some providers may under-report Facebook conversions by as much as 40%. This is often due to the fact that roughly 40% of people browse Facebook using HTTPS instead of HTTP and when someone clicks on an ad on Facebook and converts on a site, the referrer cannot be recorded since they left an HTTPS environment and entered an HTTP environment. In addition, if someone opens a new browser tab and purchases there (ex: they click the ad, run to a meeting, then go back to the website after work to purchase the product), the referrer URLs will no longer be available. The analytics tool will consider the purchaser a different person and not attribute the sale to Facebook.
- Ad Blocker Software – Your Facebook pixel may not fire if someone is using an ad blocker while browsing your site. This can cause Facebook to underreport conversions as compared to your internal data.
- Pixel implementation – You may also be seeing discrepancies because your pixel isn’t firing correctly.
Facebook has had to answer many questions about their reporting vs third party reporting. What many people have done is add a Google Analytics tracking code to their ads on Facebook and then compare the number of clicks as reported by Facebook to the number of clicks as reported by Google Analytics. Some clients say that the number of clicks reported in Google Analytics is as low as 6% of the total as reported on Facebook. See: https://moz.com/community/q/why-does-my-google-analytics-show-a-massive-discrepancy-from-facebook-s-reported-website-clicks
Under the heading – Click Fraud? – Thomas Fitzsimmons wonders why his FB Stats show 1 917 ‘Shop Now’ clicks in 7 days but he has only sold 13 books. He goes on to say; “The last time I advertised with FB i caught them at “click fraud” and they returned my money.” See: https://www.facebook.com/business/help/community/question/?id=10207372047313778
In an article titled – Here’s how Facebook click fraud works – Adarsh Thampy says; “Facebook says 192 clicks and charges me for it while Google Analytics tells me 19 clicks. That’s some seriously flawed reporting. It could be an error with Google as the rep alleged when I reached out to them.” See: https://medium.com/@ConversionChamp/what-happens-when-facebook-decides-to-do-click-fraud-43b187b81bc7
Key findings from the Bot Baseline Report include:
- Traffic sourcing is still the major risk factor for fraud. Traffic sourcing, or the process of purchasing traffic from inorganic sources, was again a large source of fraudulent activity. The report said 3.6 times as much ad fraud came from sourced than non-sourced traffic.
- Nine percent of desktop display and 22 percent of video spending was fraudulent. This was a decline from the previous year when display advertising fraud was reported at 11 percent and the fraud rate for desktop video was 23 percent.
- Mobile fraud was found to be considerably lower than expected. Overall, participants saw less than two percent of fraudulent activity in app environments and mobile web display buys. However, this does not include fraud in mobile web video and pay-per-click fraud which remain high and problematic.
- Fraud in programmatic media buys is no longer riskier than general market buys as media agencies have improved filtration processes and controls.
The report provides a number of action steps for the industry to fight fraud going forward, including:
- Demand transparency for sourced traffic.
- Refuse payment on non-human traffic in media contracts.
- Avoid excessive restrictions.
- Encourage Media Rating Council (MRC) accredited third-party fraud detection on walled gardens.
- Support the Trustworthy Accountability Group (TAG).
Fraudsters continue to focus their efforts on the most lucrative areas of the digital advertising ecosystem with billions of losses predicted in 2017 due to sophisticated fraud techniques. White Ops conducted the study from November 2016 to January 2017, analyzing more than 30 billion online advertising impressions from 49 ANA-member companies.