It has become critical to understand what digital ad fraud is. Globally, advertisers are losing at least 20% of their Ad Spend due to digital ad fraud. Ad fraud is constantly increasing, especially on the mobile platform, and as digital mobile penetration and the spends on mobile devices is going high, we are all now in a situation where we can’t afford to not to advertise on mobiles, since more than 70% of the impressions served are on mobile devices itself. Even though Brands have now started investing time and effort in identifying the ad fraud and taking measures to counter it, it is still not completely controlled.
As per the whitepaper released by The 614 Group, 78% of publishers report experiencing nefarious traffic on their sites.
Source: The 614 Group, whitepaper “2017 State of Digital Publishers’ Fight Against NHT: To Block NHT or Not to Block?”
While the ad fraud conversation is nothing new for digital marketers, let’s try to get the public to comprehend how ad fraud works, the types and methods used for ad fraud and possible actions that can be taken to keep a brand safe from Digital Ad fraud.
When running any Digital Campaign, basically three questions needs to be answered
- Are You Reaching the Real (not just Right) Audience?
- Are You Reaching your Audience at the Right Place (for Real People)?
- Are You Paying the Right Cost?
If your instant answer is No to any of the above 3 Questions, then it is vital to understand Digital Ad Fraud, including means to prevent it.
1. Reaching the Real Audience happens: This is the most effected ad fraud in digital industry which predominantly has 2 elements
A) Click Fraud – Click Frauds are perpetrated by humans or machines (bots or algorithms) that imitates and acts like a real user. It has been seen that almost 30% of reported clicks on some campaigns have been invalid clicks perpetrated through click fraud.
Although Google does not charge for any invalid clicks, it is still a valid concern and is one of the most recurring examples of click fraud. Click farms, Botnets, Ghost Sites and Ad Stacking are some of the examples for ad fraud – this includes clicks, likes, app installs, engagements, leads and even reviews. Below is an example of a recent ad which advertises the whole process
What can be done to control it?
- The campaign performance needs to be monitored closely, and campaign insights need to be scrutinized, despite the performance.
- Either a custom set or a purchased set of spam/blacklisted sites need to be excluded from the campaign.
- Start working with the trusted sites/network who are improving performance from time to time and take learnings from them.
- Deploy the tools that are accredited by Media Rating Council, (Adloox (mw), ComScore (mw, ma), DoubleVerify (mw, ma), Integral Ad Science (mw), Meetrics (mw), Microsoft Bing, Pixalate (mw), White Ops Fraud Sensor (mw) etc.,)
B) Targeting Fraud: This is another important aspect of ad fraud in digital marketing, which can be perpetrated either by the user, the digital marketing manager, the agency, the Publisher or the Ad network.
These are the four Parameters where targeting can go wrong.
i) Demographic Targeting: This can be controlled by getting console level targeting filters and asking for the proof of targeting from any third-party vendors or ad networks, which becomes critical. Comparing the console or publisher data with the Google Analytics Demographic Data also will also help to identify and take corrective measures. If money is not an issue, tools like Neilsen can be deployed which will track the accuracy of the demographic data for campaign optimization.
ii) Geographic Targeting: This is a type of targeting which is easy to setup but difficult to control since people often open emails from different location, there is a change in IP address (dynamic IP), usage of mobile devices, usage of VPNs etc. Despite concerted efforts, there will still be some decent margin of traffic from non-targeted cities present (this applies even from Google Adwords clicks to Google Analytics sessions.)
iii) Psychographic Targeting: Even though most of the targeting settings are optimized at Geographic/Demographic levels, Psychographic targeting is also a very important option for the campaign manager to make sure that the right people are watching the ad for the brand that is running. These days, most kids watch videos/read content on their parents’ devices, hence it is important to block content watched by kids (since the chances for the ad to be served to kids increases) if it is not being targeted to them. Imagine a scenario where a Youtube campaign is being run for a premium apparel brand ad to reach out to mothers in LA and her kid ends up watching that video in her mobile.
What can be done to control it?
- Check if the targeting settings are correct, (Is the target location of the targeted audience correct?)
- Check with the device/browser to verify for an error as the most common cause of the issue are wrong geographic reports.
- Check if the settings are correct, as there is reliance on Google Analytics data to verify whether the ads are directing the audience to the website
- Platforms that use mobile operator data and can be combined with publisher campaigns like Facebook/Google can be used to serve ads only if the Campaign and Operator Location Matches – (Vserv is an example of the same)
2. Reaching Audience at Right and Appropriate Place:
One of the most critical things that most digital marketers ignore is about reaching the audience at the right place. It is very important for any brand to appear on website which are safe to advertise. In this world of UGC, the possibility of serving ads on low quality, suspicious or fake news sites is always high.Though there are many default and automatic exclusions in some channels, it is still critical to monitor the campaign regularly and act against any errant channels that might be present.
Upon observing the image above, a brand advertising cruise vacations on a site where the content relates to cruise accidents – this happens when negative terms are not excluded. It is recommended to have a separate list of negative terms which are not safe for the brand and use them before setting up any campaign.
3) Paying the Right Value for Reaching the Audience:
Along with reaching the right audience at the right place, the next thing that should definitely be considered is the value that is being incurred for the campaign, which needs to be checked for.
A) Viewability: In the programmatic advertising world, there needs to be more focus on checking whether the ads are viewable or not and to check non viewable impressions are also getting paid.
B) View Quality: When direct media buying is happening with publishers, it is important to work with those who don’t do ad stacking, auto refresh (atleast capped with 5+ sec), pixel stuffing etc.,
Best way to avoid ad fraud is regular monitoring of the campaigns for engagement rates, eliminating small ad sizes, reducing the weight of the ads and also using trusted partners to serve ads. Additionally, third-party ad servers like Sizmek can be used to track the viewability (with internal filters where it doesn’t serve ads if the spot has more than 6 ads, if the spot has 4 or more impressions in 2 milliseconds, serving ads in pages with high refresh rate etc.,)
Since there only exists a regularization board but no monitoring for any digital campaign and the big players are at war to get maximum share on digital spends, there is no (or very less) scope for regulation body to audit digital spends in the horizon. Hence the onus of controlling the brand safety is fully with the brand – it might be the internal brand team or agency of the brand or publisher and tech firms who work with the brand.
Accidental Views, bot clicks, wrong targeting, bad placement, misguided influencers, etc are the problems that persist and it is time to act if the brand must become better.
#Digital #Digitaladfraud #adfraud