Human insight in the AI era Five ways leading brands are redesigning agile research
Agile insights programs have become a much bigger part of the research toolkit over the last several years.
Brands have been looking for faster ways to answer business questions for a long time, but the pandemic accelerated that need in ways few organizations anticipated. More recently, sustained economic volatility, hypercompetitive markets and rapidly shifting consumer priorities have created situations where teams need answers while events are still unfolding.
Research teams are also operating in an environment where stakeholders have become accustomed to getting answers instantly. AI has changed expectations around how quickly questions can be answered, even when those questions still require fresh feedback from real people.
Many organizations already have agile research programs in place. What we’re hearing more often today is a different conversation. Researchers are evaluating whether those programs are delivering the flexibility, engagement, qualitative depth, data quality and AI readiness they need today.
This guide explores five observations that continue to surface in those conversations. Each reflects challenges and opportunities researchers are navigating as they support more stakeholders, answer more questions and adapt to changing business conditions.
Many agile programs aren’t actually agile
Agile insights programs were designed to help organizations answer business questions quickly. In most cases, they give researchers access to an established framework for running tactical surveys across different audiences without starting from scratch every time a new request appears. When those programs are working well, feedback starts arriving almost immediately. Reach3, a Rival Group company, reports that 40% of participants complete conversational chats within the first hour, allowing teams to begin reviewing customer input while projects are still in motion.
Yet many researchers tell us their agile programs still feel surprisingly similar to traditional project-based research, sometimes amounting to little more than packages of ad hoc studies sold together.
Launching a study can involve more setup than expected, approvals can stretch timelines and expanding the program beyond its original design can increase both complexity and cost. What was intended to be a faster, more flexible way of working can begin to feel a lot like the process it was meant to improve.
What businesses expect from agile research today
The definition of agile varies from one organization to the next. Some organizations need faster answers to support business decisions. Others want a single program that combines qualitative, quantitative and video-based research. For many, the broader goal is to embed research into the way the business operates so customer feedback can inform product development, marketing, customer experience and other decisions as work moves forward, rather than becoming a separate project that happens afterward.
Researchers who once used agile programs primarily for quick-turn surveys are now looking for ways to combine qualitative, quantitative and video-based research within the same program, alongside AI-enabled tools and ongoing audience engagement. Before evaluating any platform or program, it’s worth defining how agile research is expected to support the business, from product development and marketing to customer experience and innovation.
One financial services client came to Reach3, a Rival Group company, with an agile program that had been running for years across multiple business units. As the program evolved, the team wanted access to capabilities that hadn’t been part of the original design, including video, integrated qualitative and quantitative research in one stream and AI-enabled tools. The organization wanted a more engaging participant experience while preserving the depth and sophistication of its existing research.
Today, the program supports research among audiences from high-net-worth individuals to Gen Z consumers and incorporates tools such as AI-based smart probing, which automatically asks relevant follow-up questions, helping researchers collect richer, more detailed feedback without adding time to the research process. In fact, our research-on-research has shown that this approach delivers 5x longer responses, and, when video is used, responses are nearly 8x longer.
Questions to consider
0 of 5 considered
Researchers need more than surveys
Many agile programs were built around surveys, and surveys continue to play an important role. The challenge is that many of the questions researchers are being asked today require more context than a survey alone can provide.
Organizations want to understand not only what people think, but what they were experiencing in the moment, what influenced their decisions and how those decisions fit into their daily lives. As we sometimes put it, we want to see beyond the “what” and get down to the “why.”
Those questions often can’t wait until the experience is over. Conversational research allows organizations to collect feedback while experiences are still unfolding and quickly return to participants when new questions arise. That helps research keep pace with product development, marketing campaigns and other business decisions that continue to evolve after a study begins.
As a result, many organizations are changing the way they think about agile research. Rather than treating qualitative and quantitative research as separate activities, they’re looking for ways to bring them together within the same program.
A participant might answer survey questions, upload a photo, record a video response or complete a diary activity. Researchers can then recontact those same participants as new questions emerge, building a richer understanding of the experience over time rather than relying on a single point-in-time survey.
Hershey was looking for a more consistent way to understand shopper behavior across key seasonal moments such as Halloween, Christmas and Valentine’s Day. While the company had strong seasonal data, the team wanted a better way to compare results across occasions and over time while understanding what was influencing consumer decisions from one season to the next.
Working with our sister company, Reach3 Insights, Hershey combined ongoing quantitative tracking with in-the-moment video and conversational feedback from shoppers. The result was a more complete view of awareness, consideration and purchasing behavior across seasonal windows, supported by the context needed to understand what was driving those behaviors.
Questions to consider
0 of 3 consideredBetter participant experiences lead to better feedback
Researchers spend a lot of time thinking about data quality, but participant experience doesn’t always get the same attention. Most people have completed a survey that felt more like a test than a conversation. Long grids, repetitive questions and survey experiences that require considerable effort can make it difficult to keep participants engaged. When people disengage, the quality of the feedback often suffers.
Participant experience is only one part of the equation. Organizations also need confidence that they’re hearing from real people. As concerns about bots, fraud and low-quality sample continue to grow, many research teams are paying closer attention to how participants are recruited and validated alongside the experience itself.
This is one reason many organizations are rethinking how research is designed and delivered. Rather than treating participant experience as separate from data quality, they are looking at how the two influence one another.
Mobile-first, conversational approaches are built around the way people already communicate. Participants can respond in their own words, share photos or videos when relevant and provide feedback closer to the experience being studied. The result is often more detailed responses and a greater willingness to engage with future research.
A recent research-on-research study conducted by Rival Technologies and Reach3 Insights compared conversational, mobile-first surveys with traditional online surveys among more than 2,000 respondents across the U.S. and Canada.
The study found that conversational surveys generated open-ended responses that were up to eight times longer than those collected through traditional surveys. Participants also rated conversational surveys higher on engagement, enjoyment and ease of participation, while quantitative results remained consistent across both approaches. Conversational studies also achieved average completion rates of 87% and response rates exceeding 60%, suggesting that participants were more willing to begin and complete the research experience.
The findings suggest that when research feels more natural and less like a task to complete, participants are often willing to share more thoughtful and detailed feedback.
Questions to consider
0 of 5 considered
AI is changing what stakeholders expect from research
AI adoption extends well beyond research teams. Marketing teams are using it to generate campaign ideas, product teams are using it to summarize feedback and explore concepts, and leaders across the organization are becoming accustomed to asking questions and getting answers immediately.
According to McKinsey, 78% of organizations now report using AI in at least one business function. As AI becomes part of everyday workflows, expectations around speed, access to information and decision-making are changing as well.
Research teams are feeling that shift. Questions that might once have waited for a quarterly study or a custom project are now surfacing every day. Stakeholders still need reliable answers, but they expect those answers to arrive much faster than they did in the past.
AI can summarize existing information and identify patterns within it. When organizations want to understand how consumers are responding to a new campaign, product launch, pricing change or market event, they still need current feedback from real people. Without that input, teams risk relying on historical information, internal assumptions or AI-generated outputs that haven’t been validated against current customer behavior. Many organizations are also beginning to think beyond individual research projects. Fresh human insight becomes part of a broader human context layer that supports decision-making, strengthens internal knowledge bases and gives AI systems access to current customer understanding rather than historical information alone.
This is one reason agile research programs are taking on a larger role within many organizations. They provide a continuous source of fresh human insight that can support business decisions as they are being made while also contributing to the organization’s broader knowledge and AI capabilities.
During Agile Insights in the AI Era, a masterclass webinar from our sister company Reach3, the speakers discussed how AI is reshaping expectations across organizations. Stakeholders are becoming accustomed to getting answers immediately, creating new pressure on research teams to respond more quickly to emerging questions and business needs. Organizations still need a way to validate assumptions, test ideas and understand how consumers are responding to changes in the market, which is why many teams are using agile research programs as an ongoing source of customer feedback.
Questions to consider
0 of 4 consideredSuccessful agile programs are designed, not deployed
Many organizations focus on technology when evaluating agile research programs. Technology matters, but successful agile programs are shaped just as much by the operating model, the people who use them and the business decisions they’re designed to support.
Agile programs often touch many parts of an organization and support a wide range of stakeholders. That makes alignment an important part of the setup process. Teams need to agree on who the program is intended to support, how research will be conducted and how insights will be delivered. One stakeholder may need a dashboard, another a video highlight reel, another an executive summary and another a direct feed into internal AI systems. Those requirements are much easier to address before a program launches than after it has been operating for months.
Every organization defines agility differently. One team may focus on speed and turnaround times, while another is trying to combine qualitative and quantitative approaches or support a larger group of stakeholders. Expectations around turnaround times, methodology and rigor can vary considerably across teams, which is why those conversations are worth having upfront. Teams need to define what isn’t agile from the beginning. Organizations that establish those boundaries early often find it easier to manage expectations, allocate resources and scale adoption over time.
CareFirst BlueCross BlueShield wanted an agile research program that could support multiple audiences and different types of business questions within a single framework. The program includes healthcare decision-makers at small and medium-sized businesses alongside consumers making health insurance decisions for themselves and their families. Rather than relying on a single research method, Reach3 combined conversational surveys from Rival with qualitative interviews to match the needs of each project while providing quick-turn insights over an ongoing five-year program.
The result is a flexible research program that supports both tactical business questions and longer-term learning. By designing the program around CareFirst’s stakeholders and research needs from the outset, the organization can quickly gather feedback from different audiences without creating a new research process each time a question arises.
Questions to consider
0 of 5 considered
So what actually does happen after the survey?
Successful agile programs are not defined solely by technology, although researchers have more options than ever before. Video, conversational methodologies, AI-enabled tools, mobile-first experiences and new ways of combining qualitative and quantitative feedback have expanded what agile research can do. So how do those capabilities fit within the research function and what role is agile research expected to play?
Organizations approach that question differently. One team may be trying to answer tactical business questions more quickly, while another is focused on bringing qualitative and quantitative learning into the same workflow or supporting a larger group of stakeholders. AI has added another consideration as teams look for ways to gather and validate customer feedback quickly enough to keep pace with internal demand.
Across the examples in this paper, organizations spent time defining what the program was expected to do, who it was designed to serve and how it would fit alongside other research activities.