对话式研究调查：为什么传统调查正在失效——以及2006人认为什么方法更有效

Why Traditional Surveys are Failing—and FOREWORD I’ve spent my entire career in research, and if there’s one thing I’ve learned, it’sthis: how we ask questions matters just as much as what we ask. For decades,we’ve relied on long, rigid online surveys that often feel more like homework than Response rates are plummeting, data quality is under pressure, and in somecases, bots are the ones “participating” in our studies. As an industry, we needto be honest with ourselves: traditional surveys aren’t cutting it anymore. Seven years ago, we set out on a different path when we introduced aconversational approach to research that felt more natural, more human, andultimately more effective. Since then, we’ve seen firsthand how this method Over time, we’ve refined our techniques, scaled whatworks, and most recently, expanded our capabilitieswith the help of AI. This moment felt like the right This new research-on-research study is our lateststep in that journey. We wanted to validate howfar conversational methods have come, and whatthat evolution means for researchers today. Inside JENNIFER REIDCo-CEO and ChiefMethodologist,Rival Group Let’s face it: traditional online surveys feel They’re rigid and outdated, misaligned with how people communicate andbehave today. At a time when attention is fractured, screens are crowded, and It’s not just low response rates that are hurting data quality. More bots are takingsurveys, gaming incentives and skewing results. Meanwhile, real people are This is where conversational research makesa difference. Instead of forcing people throughrigid, impersonal forms, it meets them where 5%average survey responserate today—a major SURVEY RESPONSERATES IN FREEFALL According to Pew ResearchCenter, survey responseratesdropped from just five responses to key labormarket questions. In theU.S., the Current Population CONVERSATIONAL RESEARCHIS DEFINED BY THESE Here’s what you’ll How conversational surveys unlockricher, more thoughtful responses The ChannelDistributed via mobile-first1 Why the experience matters—and The ToneA more human,friendlier approach The InterfaceA familiar, intuitiveexperience that mimics How the data holds up in The design principles that make These elements make theprocess feel easy, engaging, andeven enjoyable — three qualitiesthat matter deeply to participantsand to the quality of the data What better input means for AI- If we want better answers, we needto ask better questions—and do itin a way that makes people want to METHODOLOGY How was the research conducted? We know researcherslove a good methodology section, so let’s get right to it. This research-on-researchstudy was designed byReach3Insightsto answer a key question: Here’s who took part: • 2,006 total respondents•1,001 took the traditionalonline survey•1,005 took the mobile-first To find out, we ran a paralleldesign with the same topicand questions, but two very Both groups were asked abouttheir participation in onlinecommunities like Reddit, Discord,and others. But just to be clear:this report isn’t about how people GROUP 1: One group received a traditionalemail-based online survey The other completeda chat-like, mobile-first Fieldwork ran from April 16 to May 5, 2025. All statistical testing Richer, deeper responsesthat drive better insights Ask a good question in the right way, and people will tell youmore than you expect. That’s the power of conversationalresearch. When questions feel like a natural back-and-forth In our parallel study,the same questions wereasked to two separategroups—one using a Thoughtful Wins It’s not just about length, although volume doesprovide more detail and context — giving you apeek below the surface.Primarily, though, it’s about quality. To measure this, we used our proprietaryThoughtfulness Score,a metric that evaluates The difference wasstriking: conversational These results highlight the power of video feedbackcaptured via mobile. When people talk instead oftype, they share more. Speaking reduces cognitiveload and feels more natural than writing, especiallyon mobile devices. One study found that writing 2.5x longerresponsesusing a chat-like 5x longerwith SmartAI Probe, a tool thatgenerates contextual Better responses lead to stronger analysis and moreconfident decision-making. In an AI-driven research Nearly 8xlonger whenparticipants responded A better experiencefor every respondent When people have a better experience, they give betterdata. It’s that simple. One of the strongest findings WHEN COMPARED TO A TRADITIONAL SURVEY, CHAT-LIKEAPPROACHES CLEARLY WON OUT AMONG ALL AGE GROUPS: These three factors are essential to driving better participation and richerfeedback, especially on mobile. When we looked a bit closer at the databy age, we found that these numbers rang true across all demographics Importantly, this conversational experience had an impactbeyond just one interaction. Participants who took the chat- Experie