[GRIT COMMENTARY] Revisiting Automation: Efficiency, Transformation, and the Path to Smarter Research

Every GRIT report for the past several years has chronicled both the promise of and the uncertainty about automation. This report continues the trend, showing tentative usage across the various areas categorized as “automation.” The results demonstrate that, while it’s reasonable to conclude that although we've come a long way, we've still got a long way to go.

Part of the challenge is the broad range of things that fit under the automation umbrella.  Currently, the discussion spans from visualizations to machine learning, to name just two key components, extending from the obvious to the truly experimental. The question we are all left with is how to determine where to invest time and resources in automation?

One way to frame automation is to view it from the benefits perspective. Said simply, what problems do the various approaches to automation solve? What impact will they have? One way to parse out trends in automation that can help us realize immediate opportunities and future potential is to look at automation as the genesis of both efficiency and transformation.

The GRIT report illustrates that the primary interest – and traction – around automation focuses on the efficiencies that can be realized. From concept tests to trackers, whether you standardize the entire research study or just a portion of it, you can increase your productivity, reduce the likelihood of errors, and streamline execution and collection of data. The flexibility of having standard templates that can be customized, as well as having modular versions that address the most common kinds of testing – such as customer satisfaction, brand, and share of wallet – supports speed and low cost for iterative data-gathering.

Efficiencies of automation can also be realized in other parts of the research process, from sample selection to real-time analysis within dashboards, that enable the researcher to interact with and visualize study results. The ability to conduct real-time “what if” analysis and to show and interpret survey results among different sample segments on the fly means that researchers are investing their time in decision-oriented, rather than task-oriented, activities. Increasing speed to insights and making more time available for strategic-level problem-solving – while reducing spend when compared to traditional approaches – adds up to significant competitive advantage.

Things get interesting when we think about the transformational side of automation. New approaches to research, often built atop efficient technology-driven platforms, are enabling us to dig deeper into participant behaviors and motivations. They are allowing us to ask fewer questions – or ask them differently – while simultaneously increasing the depth of the data collected. As researchers become more comfortable with automation, they will seek out more ways to use it. The result is a process that balances simplicity with improved participant engagement and responsiveness.

An exciting corner of automation, although one in which we are just beginning to crack the code, is artificial intelligence. Natural language processing, for example, is a technology that has great potential and is beginning to be used to fundamentally change the survey process. It can be applied in multiple ways, from powering conversational surveys to auto-generating reports across both structured and unstructured data.

While there may be some uncertainty and even hesitation around automation, the benefits are ours to realize. Do we want to spend our time collecting data or interpreting it? Framing the question or answering it? Executing studies or delivering results? Automation can provide a means to that end, making us more efficient and smarter while ultimately driving better business decisions.