Artificial Intelligence Making Research More Efficient and More Precarious at Once
By Ken Donaven and Chelsea May
Artificial intelligence continues to permeate nearly every industry out there; and market research is no different. Like so many professionals grappling with both the allure and alarm that AI tools offer, we as insights professionals and market research consultants are keeping a close eye on its evolution, analyzing what innovations should be adopted and which temptations should be avoided.
In an age where data reigns supreme, our team is continually seeking innovative tools to extract actionable insights from the vast troves of information at our disposal. The integration of artificial intelligence into market research processes has been embraced by some as a game-changer, promising to streamline data collection, enhance analysis, and drive informed decision-making. However, as with any technological advancement, the advent of AI brings both opportunities and challenges, prompting researchers to navigate the terrain with caution and curiosity alike.
The Rise of the Machines
At the forefront of market research lies the integrity of survey responses, the cornerstone upon which data-driven insights are built. Yet, the rise of AI poses a significant threat to this integrity, undermining the reliability of survey data.
One of the primary concerns in our industry is the insidious nature of AI bots, capable of mimicking human responses with alarming precision. These bots, often motivated by human incentives, such as gift cards or monetary rewards for survey participation, can infiltrate surveys en masse, skewing results and compromising data quality. They are only there to complete the survey and capture the rewards, with no regard for the survey itself. Left unchecked and undetected, the counterfeit responses generated by AI machines can render results anywhere from meaningless to malignant.
To combat this, human intelligence must be applied to data collection to undermine and counteract the nefarious actions of AI-powered bots. Thankfully, the machines often have a “tell” — which is often undetectable until scrutinized by trained professionals. And we are getting better and better at spotting irregularities. By generating responses that appear too polished or uniform, AI bots trigger red flags for researchers, which allow us to weed out respondents that prove not to be genuine. We market researchers find ourselves engaged in a constant battle against AI bots, employing sophisticated tools and techniques to safeguard the authenticity of survey responses and uphold the sanctity of the data we collect.
An emerging secret weapon at our disposal? Artificial intelligence itself!
By leveraging machine learning algorithms, researchers can now quickly identify and eliminate fraudulent data points, minimizing manual effort and maximizing efficiency of the data collection effort. Perhaps ironically, AI, while creating problems, is also providing the solution: the ability to expedite the data cleaning process through automated flagging of suspicious-looking responses. These tools can take what were once human tasks — detecting “straightliners” and identifying respondents who spend far too little time on the survey, etc. — and perform them at scale more quickly than ever before.
Score one for AI. And for humans as well.
It’s Not All Bad
Despite the challenges posed by AI bots, the integration of AI into market research processes offers a beacon of hope in the quest for efficiency and accuracy. There are, truth be told, several ways in which AI augments traditional market research practices, streamlining data collection, analysis, and interpretation.
For one, AI empowers researchers to extract deeper insights from complex datasets, uncovering hidden patterns and trends that might elude human analysis. Natural language processing (NLP) algorithms, for example, enable researchers to analyze open-ended responses, extract sentiment, and categorize themes with unprecedented speed and accuracy. By automating time-consuming tasks and empowering researchers with advanced analytical capabilities, AI unlocks new possibilities for innovation and discovery in market research.
Additionally, machine learning tools can be applied to raw data sets to allow researchers to quickly find the most common phrases and most common sentiment expressed by large pools or respondents. For years, we’ve been applying more primitive forms of AI for web scraping of consumer reviews from public review sites to quickly scan and learn the topics, the trends, and the common and universal pain points, and to observe large-scale trends over historical timelines — without (or before) ever conducting a survey. These tools only mature and get more sophisticated and reliable as they evolve. We advocate for adoption in this area, but also to avoid relying on them without human supervision.
Lastly, AI can be trained. It can learn. It can even teach. Just as we would with human apprentices, we can and should be training these machines so they can get smarter, be more useful, and prove more reliable. If supervised and trained by human professionals, AI learns as it goes, and can begin to “understand” what a good survey response looks like. Such tools allow professional market research consultants to spend less time on data gathering and sorting and more time on analysis, understanding, forming insights, and delivering intelligence to clients.
Artificial Intelligence + Human Intelligence = Augmented Intelligence
We are huge proponents of the power of AI — augmented intelligence, that is. That’s when human intelligence applies the promise and precision that artificial intelligence offers the market research world to deliver carefully monitored and meticulously vetted insights, leaving nothing unsupervised by the “adults in the room.”
While AI-driven efficiencies promise to revolutionize data collection, the proliferation of AI bots threatens the integrity of survey responses and data quality. Navigating this terrain requires a delicate balance between leveraging AI’s capabilities and safeguarding against its perils.
By embracing AI as a tool for innovation and insight, market researchers can unlock new possibilities for understanding customer behaviors, identifying trends, and driving informed decision-making. However, this journey must be undertaken with caution, mindful of the ethical considerations that accompany AI-driven research.
In our opinion, AI will not replace humans in the art and science of converting data into information, and information into intelligence. It is the humans who are appropriately applying AI tools who will outperform (perhaps replace) those who don’t. Just as WebMD didn’t replace your heart surgeon, AI alone cannot and should not be thought of as a replacement for the nuanced art and skill of market researchers.
As researchers continue to explore the frontiers of AI in market research, we must remain vigilant in upholding the highest standards of integrity, transparency, and ethics. Not only is it not going away or something that can be ignored, innovations and AI pose an actual threat if not counteracted by humans trained to overcome them.
By navigating the complexities of AI with foresight and integrity, market researchers can harness its transformative potential to shape a future where data-driven insights are not compromised for the sake of expediency or the illusion of cost savings. The true cost comes in getting it wrong, when all the “intelligence” in the world confidently says otherwise…artificial as it may have been.
The Martec AI Task Force
The moral of the story: AI is evolving quickly and the market research community simply must evolve with it. It is a fool’s errand to ignore or dismiss the new tools emerging in our industry. The rate of change is accelerating exponentially; what was once thought to be impossible just six months ago is commonplace today. What the future holds is both exhilarating and daunting.
What we can ensure is that whatever the future holds, Martec pursues solutions in the furtherance of our clients’ goals. We have intentionally created an Artificial Intelligence Task Force at our firm, comprised of both veteran industry stalwarts and enthusiastic, growing market research professionals. Rest assured, if there are better, smarter ways to achieve our objectives, we will adopt them. But when artificial actors conspire to get in the way of pristine data integrity, we will continue to doggedly work to counteract them.
Ken Donaven serves as Senior Director with Martec, and Chelsea May serves as Project Manager.
