Raising the Bar on ‘Acceptable Loss’ Levels
Blog | December 1, 2020
Over the past decade, countless articles, keynote speeches, and podcasts have highlighted online market research risks. Whether you are reading this as a research agency professional, a panel company (supplier), or an end client, you are likely aware of terms like speeders, cheaters, bots, and survey farms. Rather than reinforcing what we already know, let’s explore the piece typically swept under the rug – ‘Acceptable Loss.’
‘Acceptable Loss’ is the embodiment of the lowering of standards in the online survey world. It is expected that a buyer of sample will undeniably need to build in time and budget to clean data, which in most cases is a very labor-intensive initiative, and replace a large proportion of their survey completes due to low-quality responses. Whether the data cleaning process happens on an ongoing basis throughout fieldwork or as a single process post-close, the result is never pleasant. Acceptable loss leads to missed deadlines, overspending, or both. This concept also reinforces the notion that we all have a tolerance level for fraud and low quality in our data.
The average data toss rates across the industry range from 15-20% in consumer research to 30-40% in the B2B space. Answer me this: were you aware of how high toss rates have stretched in the last few years? Does this meet your tolerance level for “acceptable loss”? Or has the cynic in you accepted that this is just what’s expected? “They’re all bad.” Or “You can’t trust anything anymore.”
What if you did not need to accept this? It is time to raise the bar, increase your expectations, and redefine what’s acceptable. After all, end clients cannot make proper business decisions with inferior data. Luckily, high-quality data is attainable when mechanisms are put in place to validate the device, the person, and the project associated with your research study.
In the spirit of adding insight to the discussion, OpinionRoute will share updates on recent trends surrounding fraud prevention and data quality over the next couple of weeks. We have been busy gathering data from our ID Product Suite, (CleanID, ValidID, VerifID, RecruitID), and macro-level findings around what drives low quality. Increasing your protection against survey fraud should be simple and results quantifiable.
Follow us on LinkedIn to never miss a beat.