The Value of Time in Research: How AI Shifts the Equation

When most people think about market research, they think about methods, deliverables, or sample sizes. What they don’t think about — but what they’re always paying for — is time. Every survey, every interview, every focus group is really an exchange of minutes for money. Respondents give us a slice of their day, and we compensate them for it. Yet in all the years I’ve been in research, I’ve noticed that few people consciously measure value this way. We talk about “completes” and “fielding costs,” but what we’re really buying is human time.

In B2B, that time is especially expensive. A short survey might cost $100 per person, and a qualitative interview with an executive could easily run $1,000 or more once you account for recruitment and moderation. If we wanted to combine both full quant and qual, the cost of a study quickly rises well above the $80,000 mark.

For most of my career, that meant I was constantly choosing between scale and depth. Surveys gave me breadth but not much nuance. Interviews gave me richness, but with small sample sizes. Rarely did I have the budget to do both at the reach I wanted. Research wasn’t measured by its effectiveness — it was measured by the hours I could afford.

I lived in this tension for years inside large corporate research teams. My job was to equip product managers, marketers, and leadership with insights, but the reality was that we never had the budget to get the level of understanding we wanted. I was always making trade-offs, but I was looking for value.

That changed when I shifted into the agency world. Suddenly, instead of being boxed in by a single company’s budget cycle, I had the chance to step back and really think about the bigger issue. How could we stop treating research as just an expense line item and start redesigning it for efficiency and impact?

The Breakthrough: Insights Gained per Minute

Before I could rethink how to measure research, I had to step back and define what an insight really is. To me, an insight is a piece of actionable information tied to a specific research question — something that actually changes what a business can do. Slides, transcripts, and datasets are outputs, not insights.

Once I made that distinction, the next realization came quickly: the true measure of research isn’t just about how many interviews or surveys — those still matter. The real question is what you can extract with the limited time you have with each respondent.

That’s why I now think in terms of insights gained per minute. Time is a scarce and costly resource. If I can have my research questions answered using a more efficient and less costly method, I will choose that almost every time. Why ask five questions when you can get the same insight from one question? Once again, I want the full value of every minute a respondent gives me.

AI is Creating More Value

AI doesn’t just make research faster or cheaper. I firmly believe it’s changing how insights are created. At its core, AI introduces efficiency: making every minute of respondent engagement work harder. Instead of treating surveys and interviews as separate, disconnected phases, AI allows us to rethink how insights are generated, connected, and scaled.

Where I have seen explosive growth is AI-moderated qualitative research, with many new solutions emerging and organizations starting to fold them into their insights portfolios. Traditionally, a moderated interview with an expert professional could cost hundreds of dollars and take weeks to schedule and run. Now, AI can automate much of the probing and exploration, allowing respondents to go deeper in real time at a fraction of the cost. Plus, I can increase scale by conducting more interviews with a given budget.

Where I am starting to see tremendous value is when quant and qual are combined in one seamless experience. A respondent can finish a 10-minute survey and immediately flow into a 20-minute AI-moderated deep dive tailored to their answers. What used to require two recruitments, two incentive structures, and weeks of delay is now completed in a single session. The result is both breadth and depth, with far greater efficiency. This is where insights per minute becomes powerful — every minute compounds in value instead of duplicating effort.

I've also been focusing heavily on synthetic data over the past year. I believe it represents the future of research, offering a new way for AI to create value by reducing our reliance on purely real respondents. With synthetic augmentation — or “boosting” — we can use fewer total respondents to generate the same insights, which increases our insights gained per minute.

For example, if 500 people each spend 10 minutes, that’s 5,000 minutes of data. With synthetic augmentation, we don’t need all 500 to be real participants. Recruiting 400 real respondents and supplementing with 100 synthetic cases still preserves the 5,000-minute total.

The measure isn’t just completes anymore — it’s how many insights are generated across those minutes. Synthetic data allows us to boost underrepresented groups, fill in gaps, and reduce cost, while keeping the overall quality and balance of the study intact.

Looking ahead, the next evolution is pure synthetic research. Instead of paying for people’s time to answer every question, we’ll pay them to refresh the data and keep personas accurate. Respondents contribute periodically to update their profiles, while synthetic personas carry the ongoing weight of answering, probing, and scenario testing. This model dramatically reduces cost per complete and further increases the value of every real-world minute we do spend with respondents.

A New Way Forward

There needs to be a mindset shift. I don’t explicitly frame projects in terms of completes anymore. I frame them in terms of return on minutes. When I deliver results, I don’t think, “We ran 20 interviews.” I ask myself: “How many insights per minute did we generate, and how did those insights shape the stakeholder's decisions?”

That’s what separates activity from impact. That’s what tells me whether the research was truly worth it. At the end of the day, research shouldn’t be measured by the hours paid for. It should be measured by the insights gained per minute. That’s the metric that really matters.

That said, there are times when the traditional, non-AI approach still applies. For one-way door decisions — the kind that can’t easily be reversed, like a major product launch or repositioning — it can be worth investing in both a full survey and a dedicated round of qualitative research. Those projects demand extra certainty.

But for most decisions, the smarter path is clear: maximize respondent time, connect methods where possible, leverage AI solutions, augment, and focus on the only unit that matters — insights gained per minute.