Faster preparation. Less time spent organising information. More scope to explore the evidence. There are good reasons to bring AI into research, and agencies should be using it well.
But speed is only part of the offer. Clients also need a well-designed study, the right people taking part and findings that have been properly questioned. The role of an agency runs through all of that.
I think the clearest way to explain the relationship between AI and human intelligence is to follow the project.
Researchers shape the study
The work starts with understanding what the client needs to decide. What do they already know? Where are the gaps? Is the brief asking the right question? Which approach will give us useful evidence?
That takes research experience and a proper conversation with the client. Our researchers lead the design, develop the questions and work out how the study needs to run.
AI can support that process. It can help us explore existing information, consider another angle or test whether we have overlooked a line of questioning. It can also help with practical preparation, giving researchers more time to develop the study.
That is a useful contribution. The researcher’s role is more involved than checking a generated survey at the end. We need to shape the thinking from the outset and decide what belongs in the research.
Recruitment deserves a place in the argument
At Gusto, we believe direct human interaction has a central role in understanding people. Someone can challenge a question, mention an experience we hadn’t considered or explain why our initial assumption is wrong. That exchange gives us new evidence and new questions to pursue.
We don’t use synthetic respondents as a substitute for that interaction. We recruit people whose experiences are relevant to the study and give them the opportunity to tell us about them. Access to information about an audience is different from getting members of that audience into the research.
At Gusto, that means recruiting everyday consumers and customers, as well as specialist and harder-to-reach groups. We work with client lists and free find. Each needs an approach suited to the audience and the research.
A customer invitation needs to respect the relationship with the client. A professional audience may need a convincing reason to give us their time. Free find requires us to source and check people beyond a supplied database.
The aim is to get the right participants involved and understand whose experience may still be missing. Being good at recruitment is a research capability, and one agencies should be able to explain.
AI and researchers interrogate the findings together
In analysis, the relationship gets particularly interesting. AI can help us explore patterns, compare accounts and suggest explanations. Researchers bring their own analytical work, knowledge of the study and understanding of the client’s business.
We need an exchange between the two. Explore a pattern. Check it against the original responses. Look for evidence that challenges it. Investigate differences between audiences. Keep questioning until we can explain what the findings support.
Imagine customers repeatedly describe a service as difficult to use. An initial reading might suggest a simpler interface. But what does “difficult” mean? Are they struggling to find something, understand the language or feel confident about completing a transaction? Are we treating different experiences as the same problem?
AI can help explore those distinctions. Researchers need to examine the accounts behind them, consider the wider context and judge which explanations hold up. If the responses do not tell us enough, we may need to go back to customers.
The challenge should run both ways. We should use AI to question our own preferred explanation. Experience is valuable, but we can become attached to a finding or give a memorable interview too much weight.
Equally, speaking to real people does not make every answer reliable. People forget and rationalise. Questions can lead them. Samples can leave groups out. Good analysis needs to recognise those limitations, compare accounts and use other evidence where available.
Our researchers remain responsible for the conclusions and advice. We need to explain how we reached them, what they mean for the client and where questions remain.
The benefit should show up in the work
At Gusto, we see AI as support for project development and a partner in interrogating the evidence. Our belief in direct human interaction is backed by the recruitment expertise to bring the right people into the research.
The stages connect. Analysis can reveal a gap in the sample or show that we need another question. Researchers need to stay involved throughout, making those decisions and keeping the work focused on the client’s objective.
Clients should expect efficient delivery. They should also expect us to show how our design, recruitment and analysis have helped them understand the problem and decide what to do next. That is the offer we need to demonstrate.
Like this article? Follow us on LinkedIn.
