AI as Thought Partner: Why Trained Researchers Will Always Ask Better Questions

By Discuss

• article

• Insight Communities
• Online Communities
• Online Qualitative
• Remote Qualitative Research
• Qualitative Research
• Mobile Qualitative
• Qual-Quant Hybrid
• Online Focus Groups and Forums
• Group Discussions/Focus Groups
• Depth Interviews
• AI
• Artificial Intelligence

There’s a trend I find genuinely concerning. The idea that ‘we’re going to get rid of the market researcher’ keeps surfacing in conversations about AI. And I want to be very clear: it’s not going to happen.

Learn more by watching or listening to Daniel on the Founders and Leaders Series podcast here:

The quality of questions that a market researcher asks is so much better than someone who’s not trained in research. If you were to use Claude to code and you had a decade of experience building amazing technology, you’d build higher-quality code. The same is true for a true market researcher.

You using AI as a thought partner in building research are going to get so much more out of it through the quality of questions you’re going to ask. Versus a lay person who is going to get what sounds like a good answer, but may miss the entire point or opportunity that the decision could make.

The Shallow AI Problem

When we did our listening tour, speaking with almost 50 companies’ executives, researchers and analytics firms, we heard a consistent frustration. They’re facing a lot of really early-stage DIY AI startups that are really shallow in what they look at. They’re facing very antiquated methodologies that are incredibly expensive. And they’re trying to understand how they can utilise an AI-forward tool that has the best methodologies and doesn’t skip over things using automation.

This is the tension. Companies need the benefits of innovation, but many of the tools they’re encountering don’t have the pedigree or rigour they require. The shallow tools give fast answers, but those answers miss depth, nuance and often the entire insight.

The Thought Partner Opportunity

The one thing I think people should pay way more attention to is AI as a thought partner in research. If you have this amazing ability to bring insights from many studies and data sources together to be able to talk to that data, the concepts that you’ll be able to study will be better.

This is where the real power lies. Not in replacing the researcher, but in giving them a tool that can instantly surface connections across dozens of previous studies, identify patterns they might have missed, challenge their assumptions and help them frame better questions.

I think AI is more of a tool in the hands of that market researcher to speed things up and ask better-quality questions with that thought partner, rather than something that overcomes needing them.

Why Methodology Still Matters

We’ve always had this belief that’s no longer true: that rigour and speed are opposites. If we accept that slow processes are a proxy for quality, we’re going to end up like the dinosaurs.

With the right methodology that is enabled by AI, you can have both. You don’t have to fall behind, but you want it to enable the best-quality methodologies in qual and quant, and not skip over steps for the sake of speed.

This is where trained researchers make the difference. They know which steps matter and which are administrative friction. They understand when you can accelerate and when you need to slow down. An untrained person with an AI tool doesn’t have that judgement.

The Quality of Questions

Let me be very direct about this. A trained researcher asks fundamentally different questions than someone who isn’t trained in research. They understand cognitive bias, know how to avoid leading questions, and recognise when a research brief is actually asking the wrong question entirely.

When you give a researcher an AI thought partner, they use it to pressure-test their thinking, to explore alternative framings, to identify gaps in their research design before they’ve fielded anything. The AI makes their expertise more powerful, not redundant.

And when you give an untrained person an AI tool, they get answers that sound plausible. But those answers may be based on faulty assumptions, biased questions or complete misunderstanding of what the data can actually tell you.

Cross-study Synthesis

One of the most powerful applications of AI as a thought partner is in bringing together insights from multiple studies. In most organisations, research accumulates in silos. Even when it’s well-documented, the connections between studies are often missed because no single person has time to read through hundreds of previous reports.

AI can surface those connections instantly. But it takes a trained researcher to know which connections matter, which are spurious correlations and which represent genuine insights that should shape the current research design.

This ability to have a conversation with an AI that has unbelievable depth and breadth of understanding of all the research your company has ever done represents a genuine step change in research quality. But only if the person having that conversation knows how to evaluate what the AI surfaces.

The Validation Question

When we combined Voxco’s quantitative capabilities with Ascribe’s AI-powered analytics and Discuss‘s qualitative platform, we saw something interesting. Researchers wanted to validate quantitative patterns with qualitative depth. They wanted to take qualitative insights and understand their breadth quantitatively.

AI makes those connections faster. But it doesn’t tell you which validation is needed, or how to design that follow-up research, or how to interpret apparent contradictions between what the quant and qual data suggest. That’s where researcher expertise becomes critical.

Building Better Concepts

When you can talk to an AI about all the previous research your organisation has done on a topic, you can frame better hypotheses. You can design studies that build on what’s already known rather than rediscovering it. You can identify the genuine gaps in understanding rather than conducting research that feels necessary but won’t actually inform decisions.

This is AI as thought partner at its best. Not giving you the answer, but helping you frame the question in a way that will yield genuinely useful answers.

The Three-year View

Over the next two to three years, AI-assisted research is going to be the default. Cycles will be faster and insights will be more continuous. But that won’t change the need to have trust, rigour and human voices at the centre of those decisions.

The researchers who thrive will be those who embrace AI as a tool to enhance their expertise, not those who expect AI to replace it. The organisations that get the best insights will be those that invest in both great researchers and great AI tools, not those who assume the technology alone is sufficient.

What Researchers Should do Next

AI is not going to replace market researchers. What it’s going to do is make the gap between trained researchers and untrained people trying to conduct research even more visible.

The trained researcher with AI as a thought partner is going to ask better questions, design better studies, spot patterns that matter and deliver insights that actually inform decisions. The untrained person with AI is going to get fast answers that sound good but miss the point.

If you’re a researcher worried about AI, stop worrying about being replaced. Start thinking about how to use AI to ask questions you couldn’t have asked before, to make connections across studies you couldn’t have made manually, and to deliver insights faster without sacrificing the rigour that makes them trustworthy.

That’s the future, and it’s one where researchers matter more, not less.

Learn more by watching or listening to Daniel on the Founders and Leaders Series podcast here:

Author

Learn more about