I sat down recently with Mark Menig, CEO of TrueSample for Episode 3 of the Innovate Podcast Series. At the top of the list for discussion was the recent spike in fraudulent activity affecting online publishing and advertising, and the impact it is having on market researchers.
The question we wanted to touch on was whether this spike was anomalous or part of the regular peaks and valleys we see in fraudulent activity in research panels. Largely international, but in some cases affecting US users as well, the coordinated activity we saw in the industry was substantial and it had a major impact on the industry. What does it mean and how should we respond?
Listen to the full episode below:
Determining Context in Online Fraud
One of the most important factors in understanding and responding to sample fraud is the context of that fraud. As Mark noted, this Spring’s surge in fraud represented a peak across the board for market research, online publishing, and advertising. TrueSample’s data trends show the rise across almost all indicators, but what does this mean?
To start, we should look at the global economic conditions in places like Russia, Brazil, Uzbekistan, and especially in places like Venezuela. There are large pockets of fraud in these markets as poor economic conditions and currency collapse affect almost the entire population. It’s easy to look at a $0.30-$1.00 incentive and see it as nominal, but to someone in a local economy that has been hit particularly hard by recent events, that $1 could be extremely valuable. Just how hard is it to earn $1 in their world?
We operate and live in a global society more than ever before. This changes what fraud means because there are a variety of motivations that can drive people to react. Sure, there are fraudsters out there looking to do bad. The recent spate of malware and ransomware shows the inclination of many people to break the system for the sake of breaking the system. But there are also those who act out of desperation, and while it is still fraud, the nature of what they do and the mechanisms employed are different.
Understanding the motivation of fraud can help in better understanding the approach of the individuals acting, and help in preventing it in the future.
Desktop vs. Online Fraud
As a natural offshoot of the discussion around recent fraud rates, Mark and I also looked at both sides of a conversation that has cropped up in several publishing circles of late – the comparison of mobile and desktop panels.
Specifically, there have been claims that mobile and in-app samples are better because they are more naturally impervious to data quality issues like the ones faced by desktop sample. I’ve lived and breathed mobile for years, having worked on one of the first mobile panels, and find this conversation fascinating because of the way it has evolved over the years. Is there any truth to the core claim that mobile is more secure, and what does it mean for the space?
As Mark states in the podcast, the industry has a habit of getting myopically focused on the buzzword of “mobile.” There is no silver bullet for data quality issues. There are two things to consider here:
- Not all respondents want to use a dedicated mobile app
Everyone has their own preferences. Whether a smart TV with browser, a gaming console with internet connectivity, a desktop, or a mobile device, the line between connected devices has blurred more than ever before, and while half of traffic is mobile, that’s still only half of total traffic.
Our jobs as researchers, for the longevity and prosperity of the industry, is to put the respondents first and ensure a positive experience regardless of device.
- Mobile is not impervious to fraud
One of the core assumptions here is that mobile is a silver bullet – that it cannot be hit with fraud in the same way as a desktop or laptop. Mark shared his experiences with joining mobile panels through apps. He illustrated how he could successfully join surveys that used Facebook Authentication with 100% non-real data, including a fake email that was able to bypass Facebook’s own authentication.
These things can still happen, and as mobile devices have become more customizable and the technology and resources around mobile development have increased, so too have the abilities of fraudsters. Another factor here is the potential incentive for mobile fraud. Historically, mobile has been a blend of qualitative and quantitative, with a more engaged audience, and as such a higher incentive. This of course means a higher incentive for fraudsters.
A recent story on Motherboard.com perfectly illustrates what fraudsters gain from this kind of activity. Three men in Thailand were recently arrested after running a click farm with 500 unique cell phones and 350,000 SIM cards.
Is mobile better than online? The risk of fraud is there for both and the best thing we can do is continue to provide the best possible experience for respondents while keeping this dialogue open. Through awareness, continued discussion, and a focus on staying ahead of fraud in every way possible, we can make the industry stronger.