There’s a complete cure for panel fraud, and it’s available to any insights professional who isn’t wedded to online research. The cure, of course, is in-app mobile research, in which unique smartphone technology transports clients and their projects to a research habitat where real mobile consumers love to gather, where human fraudsters can’t hide, and where survey bots can’t enter.
A public window into the specific problems that are causing trouble for online panel providers recently emerged in what its participants call the Data Quality Download Podcast Series. The two conversations posted so far are intelligent, earnest, well-meaning exchanges between veteran online panel and online data professionals. Here are some of the key points from the first installment, which focuses on two leading issues: the need to identify and reject suspect survey responses, and the unrelenting problem of hard-to-detect survey-taking bots that mimic real respondents.
- Q: “Sometimes there’s a debate between the sample providers and research firms on how much sample should be rejected, and on what criteria…what’s reasonable?
- A: “There is no average, there is no norm you can expect [for] what you have to scrub out… it’s all over the map…because there are so many different sample sources out there…and so many blends of sample methodologies.”
- For studies involving multiple-source panels, “scrubbing can be as high as 30 or 40%.”
- “A constant vigilance has to be applied to make sure your data is going to be good, and you can have confidence in the insights that your customers are going to use.”
- “[Bot fraud] is a very dynamic, very evolving challenge for the industry. It’s getting increasingly sophisticated….You have to acknowledge that this happens in our space…so we can be really proactive and get in front of this issue. Putting our head in the sand does not help anyone, right?”
- An attempt at finding a more coordinated and systematic approach to developing fast, effective ways to detect bot attacks is getting started, “but…we’re in the early stages of that.”
- Fighting online panel fraud will require providers to “create predictive checks and algorithms that can move the industry forward….If we can rely on these predictive elements rather than human review on the back end, that’s a huge time savings and it really helps.”
The podcasts so far have not brought up a crucial opportunity that should be part of any conversation about solving panel fraud and safeguarding data quality: in-app mobile research, with its proven effectiveness in preventing fraud by bots or humans. Here are important additional points for you to factor into your own thinking about data quality, panel quality, and panel fraud:
- When they commission a properly sourced and fielded in-app mobile study, or conduct one themselves on an in-app mobile DIY platform, researchers see almost no need to reject any of the completes they receive.
- Bots can’t break into the in-app mobile survey space. In-app surveys load instantly into panelists’ phones. Respondents answer them offline, the entire experience taking place in-app, which means inside the phone.
- The bots are left lurking online, where they can and do find other prey.
- “Mobile optimized” research simply substitutes a smartphone screen for a desktop or laptop. “Mobile optimized” respondents have to click on a link and enter the treacherous online space to take what’s actually only a mobile-mimicking survey.
- Each smartphone is individually identifiable, and each identified smartphone is in the hands of a validated human consumer. Fraud-minded respondents are detected quickly and permanently frozen out of the survey app.
- More than 1.3 million active panelists who use the Surveys on the Go research app will take your surveys in a mobile safe space that’s a realm apart from the online universe where bots run rampant.
The podcast series about online research’s struggles against panel fraud is fascinating and well worth listening in on, including this second installment. But if you’re serious about panel fraud and its consequences for your data, it’s even more important to have a clear, informative and productive conversation about in-app mobile. Just contact us at email@example.com.