While generative AI models are nothing new, Microsoft and OpenAI catapulted us into a new world where the practical applications for AI suddenly seem endless and available to anyone. Not only did they spark endless ideation on how to apply generative AI, but they also launched some very transparent models on how to best architect it as a capability within your organization.
The Copilot Analogy: AI’s Role in Market Research
One of these paradigms is inherent within Microsoft and OpenAI’s “Copilot” branding. Consider that analogy; the pilot is ultimately in control and responsible for all decisions made in the cockpit with the copilot playing a supporting role, owning many of the lower-risk actions while still being under the supervision of the pilot. While the copilot would be consulted, it’s the pilot who makes key decisions and is in control of taking action. While a copilot might be trusted at the controls through a little turbulence, it’s a safe bet that the pilot is in control when landing during a storm.
An AI-generated image from Open AI DALLE 2 of a pilot and co-pilot using controls in the cockpit
This forms the key tenant that generative AI is not about eliminating people from the work, it’s about eliminating low-value work and freeing people to create value in ways that AI is not yet able. As Fuel Cycle embraces these new capabilities, our goal is not to have AI be a complete replacement, but instead to let it enhance virtually everything we do.
By layering risk into our approach, we have the basic framework of how and when to apply AI. Lower-risk work and decisions are better candidates for high augmentation, while higher-risk work and decisions will see less augmentation. This paradigm will apply across any industry, not just Market Research – work that has clearly definable outputs and lower risk will be a great candidate for high augmentation, but when the risk goes up, the human factor will be much higher.
Applying AI to the Insights Life Cycle
When we apply that model against the insights life cycle, it’s easy to imagine augmenting some of the lower-risk areas such as providing summaries for qualitative research. However, it’s unlikely to replace the strategic market researcher who’s advising the business stakeholder on the call to action. Let’s think through this by looking at a few tangible examples.
Unlike quantitative data that seamlessly flows into statistical analysis, allowing the researcher to quickly draw out key insights, qualitative data is like a diamond field; there’s tons of value to be found, but sometimes you must dig through a mountain of data to find it. This is an area where generative AI is already shining by taking large amounts of data and producing summaries, themes, and insights quickly and accurately. By eliminating hours of upfront effort, the market researcher can focus on key areas and selectively dig into the data to create additional value. This forms one of the quintessential examples of high augmentation helping to free the researcher to create greater value.
Moderating Discussion Forums with AI
Moderating discussion forums is another tedious and sometimes time-consuming activity, and it’s also an essential part of collecting quality responses, driving engagement, and ensuring a positive experience for the audience. While interacting with real people introduces some increased risk, it’s another prime candidate for a well-trained AI Bot to provide engagement prompts, flag inappropriate comments and help keep the discussion productive. Keeping the AI bot boxed in helps reduce risk and makes it a great candidate for high augmentation, and there is still the need for a real person in the background checking in on things from time to time.
AI Support for Higher-Touch Work in Market Research
As key insights start to emerge, the focus turns to developing the story being told by the insights. This feels like one of the defining moments of the transition to higher touch work where the experience, intuition, and human factor of the market researcher takes over. It’s where the pilot takes back control and farms out tasks to the copilot while continuing to keep a tight grasp over the outcomes. It’s the researcher who builds the story, connecting the insights in a more natural and engaging way than we see from the current generation of AI. Along the way, AI will help create targeted slides, charts, and reports, but all under the direction of the researcher. AI acts as an assistant, helping with specific tasks rather than driving the whole activity.
AI’s Role in Strategic Decision Making
The call to action puts the researcher into the role of a consultant working collaboratively with business stakeholders to understand the insights and apply them to strategic decision-making. This is where augmentation from AI becomes vastly secondary to the people in the room. The work has been done and it is time to build consensus and act, and we are not yet ready to have AI make these decisions; while AI has helped give us the data, the pilot is going to land the plane. This goal is further enhanced by the fact that AI has been freeing up the market researcher, giving them back valuable time to focus on creating value instead of completing time-intensive tasks.
Embracing AI: Taking the First Step
While it feels like the rest of the tech world is suddenly in catch-up mode, the good news is that the barriers to entry for implementing these models are fairly low, and you’ve probably already seen them getting integrated into your favorite tools. As you start thinking about how to implement AI into your own workflow, it’s best to be on the lookout for time-consuming tasks that have a very definable outcome.
Set yourself a goal for today to identify that first opportunity to have generative AI becoming your copilot. Start small – get it to help summarize a document or help you generate ideas around a subject. It’s important to take that first step towards applying AI to your day-to-day processes!
