Why transparency is the real AI challenge in market research with Howard Fienberg

Listen

Watch on YouTube · Apple Podcasts · Spotify

Summary

Howard Fienberg, Senior Vice President of Advocacy at The Insights Association, on why transparency is the through-line that survives the AI transition in market research, how lawmakers are folding AI into existing privacy and consumer-protection frameworks rather than regulating it as a new category, why a fractured patchwork of 20-plus US state privacy laws makes a federal preemptive law urgent, what the 2025 AI updates to the IA code of ethics actually require (anonymity across the data lifecycle, disclosure of data provenance, and a human embedded in every AI research lifecycle), why respondents may open up faster to an AI but also turn angrier when it fails, and why the 2030 census and the participant bill of rights matter more to working researchers than they realize.

Guest

Howard Fienberg — Senior Vice President of Advocacy, The Insights Association · Marketing / startup

From this episode — top claims

• The third AI code addition holds that no AI system used in research should operate exclusively without human judgment embedded in its lifecycle — there has to be a human who knows what is going on inside the black box, both to treat research subjects with respect and to respect business partners; this principle is likely to end up in legislation requiring human involvement in consequential decision-making. — Ep. 57, Howard Fienberg
• The decennial census and the American Community Survey supply the frame that fuels essentially every piece of US quantitative research — and even shapes qualitative approaches — across both the private sector and government, which is why researchers should care about a complete and accurate 2030 census even if they never think about it day to day. — Ep. 57, Howard Fienberg

aytm's take on this conversation

Transparency runs on two fronts now: tell subjects they are talking to a machine, and keep vendors and clients talking about the data. The gap between your published privacy policy and how you actually operate is what sinks you, so audit it first.

aytm's perspective, voiced by Molly.

Full transcript

Auto-generated captions — speaker labels aren't always available and wording may be approximate.

0:00 Both policy makers and the insights industry are sort of in the same headspace. And I think all of us are kind of grappling with a certain amount of excitement and optimism about all the AI tools and the possibilities, but there's also skepticism. Like, what are we actually going to get? What are we losing in the process? I have a co-worker that always talks about AI as Skynet. So, there's apocalyptic things that people have in mind, which I don't expect, but there's kind of a broad range. But, we're all looking at a lot of the similar things and having some similar, yeah, both positive and negative feelings. And frankly, that's informing both how the industry is approaching it in many respects, but also how the legislators and regulators are looking at it.

0:43 Hello, fellow insight seekers. I'm your host, Molly, and welcome to the Curiosity Current. We're so glad to have you here. And I'm your host, Stephanie. We're here to dive into the fast-moving waters of market research, where curiosity isn't just encouraged, it's essential. Each episode will explore what's shaping the world of consumer behavior, from fresh trends and new tech to the stories behind the data. From bold innovations to the human quirks that move markets, we'll explore how curiosity fuels smarter research and sharper insights.

1:17 So, whether you're deep into the data or just here for the fun of discovery, grab your life vest and join us as we ride the Curiosity Current. Today on the Curiosity Current, we are joined by Howard Feinberg, Senior Vice President of Advocacy at The Insights Association. Howard spends his time at the intersection of public policy and the insights industry. He works directly with lawmakers, regulators, and research leaders on issues like consumer privacy, AI regulation, research ethics, and policies that shape how insights work actually happens in the real world.

1:53 Howard's background is, interestingly, rooted in both policy and data. His father was a statistics professor and advisor to the Census Bureau, and that connection sparked an ongoing interest in research, data, and the public policy that impacts them. So, today we are going to explore attention that a lot of researchers are feeling right now. The tools are getting more powerful and faster than ever, but the ethical and regulatory frameworks around them are continuing to evolve. What does this trust look like in that environment, and what does this mean for the people who are running research today?

2:25 This is going to be a good one. Howard, welcome to the show. Thanks so much for having me on. I appreciate it. Howard, I've I've known your work and I followed you throughout your career in the insights industry so far, and what really got me thinking about specifically having you on the show was something that you talked about when you spoke at the IA Ignite AI event recently in Los Angeles. There's this sort of tension that's happening right now where, like I just said, AI is making research more powerful, scalable, but at the same time it's raising a lot of questions in a couple different directions. Can we trust this data, and also can we trust the AI systems with the type of sensitive work that we're doing? And from where you sit, talking to policy makers and researchers every day, where do you think the industry's head is sort of at on that question right now?

3:17 So, interestingly, I think both policy makers and the insights industry are sort of in the same headspace. Um I think both I I all I mean all of us are kind of grappling with a certain amount of excitement and optimism about all the AI tools and the possibilities, but, you know, there's also skepticism and what are we actually going to get? What are we losing in the process? Um you know, I have a coworker that always talks about AI as Skynet. Uh so, you know, there's apocalyptic things that people have in mind, which I don't expect. Uh but, you know, like so there's there's a kind of a broad range, but we're all looking at a lot of the similar things and having some similar, yeah, both positive and negative feelings.

3:59 Uh and frankly, that's informing both how the industry is approaching it in many respects, but also how the legislators and regulators are looking at it. And and both, I think, yeah, on both our side and on the policy makers, I think sticking to long-standing principles is going to be the biggest key. And this is why you'll hear me hit on transparency over and over again during our discussions, uh cuz it's one of the most important things out of our own ethical codes and professionalism in the insights industry, but also something that is being hammered pretty frequently at the it in legislatures and in Congress uh when it relates to AI and data, and AI especially. But, yeah, I think that that's a reasonable thing to focus on in most issues that we might come into contact with.

4:50 That makes a lot of sense. And uh related to that, I keep coming back to this idea that that trust in research is always depended, to your point, on transparency. Respondents, you know, they typically know someone, usually a brand, is asking the questions. They know why they're there to answer the questions, provide their information, their evaluation. And then they generally know how their information is being used. I wonder though, as we're layering on all of these technologies, that respondents may not see sometimes, may not understand sometimes. You know, AI moderators, automated analysis, synthetic data that's built off of their data to some extent. So, it makes me wonder, what does transparency look like in in an environment like this that's becoming increasingly complex? Is it I guess I wonder with with everything moving so rapidly, how do you think about transparency differently at all?

5:47 Um I I think you have to well and we're we're talking about it both in terms in in similar kind of terms, but there's transparency that you need within the industry itself with it and with clients. But there's also the transparency that's required when dealing with research subjects or potential research subjects. Um so the the principles are similar, but you know, it'll be exemplified in different ways. And you know, when it comes to dealing with a research subject, you know, legislatures are already charging ahead on certain things like if you deal with an AI chatbot for instance, California, New York, and Maine, they already have laws on the books that require you to notify someone who is using a you know, sometimes it's a companion chatbot, uh sometimes they they can you know, just broadly of any kind of conversational AI uh that you you that the AI is going to remind the human that they are not human.

6:45 Sometimes they need to provide that warning not just once, but you know, like regularly during the interaction. Um sometimes it requires inclusion of all sorts of other things like protocols against suicidal ideation, all sorts of uh fun and complicated things, but that transparency is the starting point there. And sometimes again, just a reminder that because sometimes the conversation can feel kind of natural. But you want people to understand at the end of the day that they are not talking to another human. And that's sort of a a base transparency that one would expect in any interaction. Um and as it relates to the business side within our industry, whether it's with your vend between vendor and client, between you know, brands and the research provider, anywhere within the chain, that transparency is is not just about disclosing lots of things because I mean

7:43 our contracts and our terms of use, they get really lengthy. Cuz there's a lot of stuff to cover. There's a lot of stuff to go over. Contracts, you know, they and even when they're looking relatively standard, there's a lot of stuff going on on the legal side. A good point to remind people, I'm not a lawyer. This isn't legal advice. I got to have the disclosure. Absolutely. You need to disclose that. It is important for people to communicate. And I think that's one of the most important fundamentals that I'm at this point reminding people when you're charging into the AI space or even just dipping in their toes, clients need to know what's going on with their information, you know, and what might be happening during the research process that may be different than they expect.

8:30 Um and you know, that's part of you know, we're going to get into it a little later what are the Insights Association codes of ethics and standards has to say about it. But we're we're hitting on base principles of some honesty and communication between all the partners in the research chain so that everybody has a reasonable understanding of what's happening, what's going to go on with client information, what should they expect on the other end in terms of results.

8:59 Uh these are all things that you need to set a good uh you need to set a good standard for in your own discussions with each other and expectations. So, it's not just about what you think everybody knows because what everybody knows is constantly changing and it's hard to keep up. So, things are on the move and so if you're not talking to each other about it, uh you stuff will get lost and that's what will lead to someone getting really mad at you or vice versa.

9:29 Uh and we want to avoid that. Um and you know, the you don't want to be surprised by that sort of thing. It's okay to be surprised by results uh cuz that can happen in research, obviously. But you don't want to be surprised by each other in a bad way. You said something uh a little bit ago where you had said that it's important to have that transparency and the disclosure to let participants know that they're speaking to an AI and they're not speaking to a human. And there's some interesting things that happened there in just our own work uh AYTM ran some research recently about weight loss and GLP-1 use, which can tend to be a more sensitive topic. And we found that interestingly enough, respondents tended to be a bit more candid with their responses and a bit more more forthcoming with what they had to say when they know they weren't being observed by a human being. And you mentioned something too at QRCA recently

10:28 um that people may open up to an AI faster but they also get much angrier when something in that system goes wrong again because they know they don't have to have that filter of talking to a human. So there's a lot of really interesting you know, psychological things that are going on here in just how people respond to these humans. What do you think is is sort of happening here? Can you break that down for us? Well, I would again recognizing I'm not an expert on that sort of thing necessarily, but I but I look at it sort of an extension of what we learn in the social media space where people will say and do things much more freely sometimes than they would in the real world because you're not It In that case, it's just you're not face-to-face with somebody. Or sometimes you're not you're not talking to people who you know in any real con space. Like they're just you know, digital representation at best. Or maybe you have no idea who

11:26 you're talking to at all and you have no expectation you'll ever encounter them outside of that space. And that doesn't necessarily bring out everybody's best. Uh and so in the case of dealing with uh an an AI system, you know, maybe you are more free with your id. Uh that won't necessarily I mean, I guess it depends to some extent what the client wants to understand.

11:54 Because hitting the id unfiltered can be useful to the research, but isn't always. And it doesn't necessarily get you a reasoned response. It might get you an emotional one. Maybe that's what you're going for. Maybe that's you know, so that could be to your benefit. I mean, there are all sorts of ways in which this could play out that deserve a lot more study. And I think we're we're learning a lot of it in real time and you know, keep building from there. Yeah, absolutely. Well, I want to switch gears into the nitty-gritty that I'm super excited about, um which is the regulatory landscape, which we've heard from a lot of researchers that this can sometimes feel impossible to track. I mean, even for someone like yourself, it could be difficult to catch up with all of the different things that are required in all of the different jurisdictions cuz there's now 20 US states with comprehensive privacy laws that are currently in effect and each one is a little bit different. So, if someone was trying to run the same or

12:53 similar study across multiple states as utilizing an AI system or AI moderation, where would they start in understanding the space? And what's potentially something, like you mentioned, you don't want people mad at you. So, what's potentially something that people are potentially, you know, well-intentioned researchers could miss that may actually cause problems down the road? Well, first off, certainly the place to start is you should make sure you're an Insights Association member.

13:22 Um you know, from the perspective of being able to access all of our compliance information on say for example all the 20 different state privacy laws that are in effect um but also lots of other policy issues that are impacting the insights industry uh sometimes quite directly sometimes tangentially um and you know uh further to that uh you get to be a part of the advocacy for a federal privacy law that would hopefully preempt a lot of this fractured mess of 20 state laws which unfortunately is likely to be more than 20 state laws by the end of this year uh and that's just the way things go um but it and I could talk about lots of different steps that you can take but I think among the things you're most likely to miss uh when you're working on this is checking on all of your policies your notices and your contracts but how they relate to your actual procedures and your internal policies

14:20 and how you do stuff because that disconnect between what you're putting out into the world and what you're actually doing can get you into massive legal trouble in addition to you know certainly trouble with your clients and with research subjects when things don't match up um but that's really one of the most important things and I think it is missed by a lot of people and for years early on I mean I've been doing this 19 years now at uh at IA and its predecessor organizations um I spent years trying to explain to people no you can't just copy and paste a privacy policy from the another company because it might have no bearing on how you operate uh it I'm sure it looks good I mean I was reviewing I was reviewing a master services agreement from uh a platform not a research platform it's tech a tech platform yesterday and I'm going through like wow this is awesome

15:16 like I I wish more people did this but then I look at it like I don't think a lot of the other platforms that I deal with you know, smaller companies and such. They can't operate internally this way. It wouldn't make sense. Um so, along with all the steps that we recommend for people in dealing with all these different laws and regulations and expectations, you know, reviewing your your privacy policies and your notices you're supposed to provide and how you handle consent and what you're putting into your warranties and your contracts, your all these different clauses, how you respond to people making privacy rights-related requests, all of that is going to come back to how you actually operate.

15:59 So, it's not just about the forward-facing, it's about the back-end. And uh at with that, I would also point out that there's great benefits to ISO certification. Um it's something that I kind of sneered at for years cuz I looked at it and said, Well, I mean, seriously, I looked at ISO and said, "Wow, that's a whole lot of bureaucracy. That sounds like a huge pain." And, you know what? It's not an easy thing to certify to ISO and that's So, we have two that IA helps people with. One is ISO 27001, which is focused on data security, and another is 20 2252, which is for market research.

16:42 And, you know, something I can tell you for the data security one, that's it's something to make sure you're well positioned for dealing with data security laws in the United States. Um and there are a bunch of states that give liability protection against data breaches if you're ISO certified for data security. Um it's Connecticut, Iowa, Ohio, Utah, and Texas. Uh but, yeah, and there's the market research one in its own way also helps you getting a handle on data in the a broader sense because you have to understand your own internal processes and procedures, where what data you have, where is it, where did it come from, where is it going, uh what state is it in, not not US state, I mean uh how is it kept in what format and so forth. All those things help you comply with all all the data focused laws and regulations um and put you in a a much better position. It's not a guarantee of

17:41 anything, but it puts you in the right spot to really have a handle on your own practices and policies. I I just have a clarifying question on when you say that there's different states that will offer protection, is this the state that the company is operating in or the state in which they're conducting research in? Cuz especially for online where they're in all those different places, Yeah, it's about the state where the research subject is located. Yeah, that makes it hard.

18:10 Yeah, absolutely. And that's the same thing with the pri the state privacy laws. It's not about like it's not about where you are. You could be located in Timbuktu. But if your if your research subjects you're working with are all you've got, you know, hundreds of thousands of them in, you know, California, you got to worry about California. If you got a you know, more than a handful in Montana, um you need to pay attention to Montana. You know, like and that's how it gets very complicated.

18:40 Well, to stick on the topic of regulatory frameworks because they're riveting, um I do find it fascinating right now and I would love for for you to kind of impact this for us. It seems to me from where I sit that AI is is largely being regulated right now. It's really treated as its own category and instead of lawmakers seem to be kind of weaving it into existing privacy and consumer protection frameworks. And I'm curious, like is that approach, like is that the right approach? Does it leave gaps anywhere um that that are, you know, researchers, suppliers, brand side should maybe be paying attention to you if we're not going to be building these very specific AI-related frameworks and just build these into existing frameworks.

19:27 I think it's a useful approach certainly right now when getting policymakers to agree how to define AI is not the easiest thing. So, how are they going to charge in and regulate it as its own separate thing? When you know, they can't even really get to grips with some of the basics of it. So, like you know, there are states, California being the leader in it trying to restrict the use of what they call automated decision-making systems.

19:56 Which you know, we we fought for years with them about this cuz they started they originally started with a definition that was less about AI and more about really any kind of automation. So, Excel spread Excel spreadsheets, they were going to basically regulate those out of existence. And so, it was a it was a hard back and forth. Um it So, we're still it and we're going to keep struggling to get our heads around what AI is both from the industry side and from for the lawmakers. So, trying to apply long-standing principle like we were talking about earlier makes the most sense.

20:33 Yeah, and there's been discussion even in Congress about, you know, why do we need a new AI regulator that specializes in AI. Like, well, that's but we have existing government authorities that have a certain expertise and they don't all have expertise in everything that relates to it, but the idea that you would task say the National Institutes for Standards and Technology NIST under the Commerce Department with looking at practices and you know, very technical aspects of AI and AI and security and and other things. It's it's a wonky area.

21:13 You want your very you know tech heavy uh scientific focused people focused on that. And if you're worried about consumer protection and you know the as things about you know notice and transparency, how the data is going to be handled as it relates to individuals, you're talking about the Federal Trade Commission. Which has decades of experience relating to those issues. And yes, are they all going to be experts on the latest things with AI? No, not necessarily. But they're in a good position to be able to understand it and grow their understanding over time because they're focused on those kinds of issues and that kind of scope.

21:56 So I think it's the same sort of thing for researchers perhaps to to look at this as yes, it's a new tool, there are new capabilities, but you still need to come back to first principles and what you're strong at and what are you weak at and you know try to adjust that way. That's so interesting. So it's almost like at least in this nascent period of AI when people are not even in agreement about exactly what it is and it's rapidly changing that our policy should be focused and our regulations around the areas of impact because there's already infrastructure there where people have a deep understanding of their area matter and can layer in these sort of AI regulations over top.

22:38 Correct. And like and you can see it even uh with California of course being the the leader with regulating, they have mostly turned to their privacy regulator, the California Privacy Protection Agency, to take the lead on a lot of these things because that's where they see the big impact. And if people are going to fight about the energy use of data centers, they're going to talk to their energy and resource departments. as a you know as a good example. And so I think it makes sense that we would be approaching things in a similar way within our own businesses.

23:11 I I'm curious on that. Do you think that's going to stay like that forever? Or as AI becomes more embedded into everyone's everyday lives, perhaps that could change? I mean, of course it could change. But it's hard to predict what it's going to look like. Um because certainly the idea that I'm sitting here and you know making use of chat GPT professionally at least you know once a day. Uh I would not have thought that I would be doing that even you know certainly a year ago. I would have laughed at the idea. It was this this stuff is stupid.

23:46 Well, no, but there's some usefulness to it. Um and that's and it's it's creeping in in all sorts of tools and all sorts of platforms and for all sorts of purposes. So who knows? I had a a guest speaking lecture that I just did for a a grad class on market research at Cal Poly Pomona and um I I mentioned that a year ago I did the same talk and I have my little cards here and it would say AI is going to be part of your life. It's like forget it now. You it it is you are going into the market research space. You need to know AI not just as a shiny thing. It needs to be an intrinsic part of the way you operate.

24:26 Cuz even for us um you know at AYTM I work in marketing and I my last gig I wrote SEO articles 60% of my job and that's just like now I can ask Claude to do that. I can ask I can ask any of those systems to do it for me. And then I can actually move on to other things. So it's it it'll be it'll be an interesting thing to watch of course. Well, you touched on this at the beginning. So let's dive into the AI code of standards and ethics. You recently helped to lead a lot of the updates and the 2025 version now includes specific provisions as it relates to AI. So, walk us through the practical side of that and and what new things came out in that document that perhaps was different and recommendations that you're offering now that perhaps you didn't in the past. And what do researchers from from those guides need to understand today?

25:23 So, all quite out. I think the the new parts of the code that are AI specific are they're not new in concept, they're just new in their specifics. Cuz it AI is again, constantly changing, becoming more pervasive. So, you got to you roll with it and you have to figure out what you need to focus on next. Um so, one of the points that was added is something that to me seems straightforward, which is you know, I'd say the respondent anonymity is essential across the data life cycle.

25:54 So, if personal data is being used for an AI training data set without informed consent, yeah, researchers need to make sure that any personally identifiable information can't be reverse engineered by AI inference. Uh this is just going back to basic principles of privacy and security with research subjects data. Uh it's it's not a new thing. I mean, it's we're we're talking about new and you know, more easily doing interesting things with data.

26:25 But the base concept hasn't changed. And let's see, we have second one was if AI tools are selected, then their use, the purpose, the technique, uh the accuracy of the model, where the data came from, whether it's primary, if it's a secondary data source, if it's synthetic data, uh you need to disclose that. So, AI generated data, whether it's predictive, whether it's generative, you need to clearly distinguish where the data came from, whether it came from human research subjects or somewhere else.

26:58 Um again, these are basic principles of transparency with business partners, um yeah, with clients, and between vendors and providers, making sure that everybody in the chain knows what is happening and is not yeah, there are no negative surprises, right? Going back base principles. Um and the third one that's specific was that no AI system using research should operate exclusively without human judgment embedded in its life cycle.

27:29 Um and this is something uh likely to end up in legislation yeah, in one form or another, requiring human involvement when it comes to decision-making uh relating to anybody. And yeah, at at its most basic, that kind of concern is focused on decision-making for the big picture things, the yeah, the the major stuff like employment, healthcare insurance, getting a job, getting yeah, your getting a loan, but you know, we're talking about something that's at a very different level, but the basic principle still applies if you are going to be treating a research subject with respect.

28:11 Uh it also involves making sure that they're not going to be abused in the process, um that, you know, there's a human that knows what's going on in some capacity. And that's also about the respect for your business partners that someone knows what's going on inside the black box. And you can't know everything that's going on within the black box, but there's got to be some kind of human involvement.

28:38 I'm so glad that you're talking about that cuz I really that is the part of the code that I look at and I certainly am a fan of um and agree with, especially where AI is today. I just I, you know, I do enough with AI that I know where the the, you know, the softnesses are, right? And it's performance, let's say. And so, I love that conceptually, and I think a lot of researchers would really resonate with this idea that like the human in the loop, essentially. But it does raise the question of like, what does that mean in practice? And I think you kind of hit on two parts of it. One is in, certainly in the design, right? And that's that goes back to, well, two things. One, that you're doing right by your client, but also that you're creating a survey instrument that is, you know, going to be respectful of your, um, respondents. And then also in the analysis, but specifically the sort of dis- distillation into these key insights, and ensuring that like what we

29:36 learn is vetted by a human to say, I can see all the evidences for this, right? Cuz the AI has laid the map for me, and I can look at this and be confident that this is what the data says. And that's my sort of human oversight of that, looking over that stuff and saying, I see how you got there, and so I can trust this and put my stamp of like, you know, human expert approval on it. Is that kind of how you think about it?

30:05 Yeah, I think that it is it'll depend on the study and the tools where human judgment is needed. But yeah, it it is it's needed somewhere in there, and it's going to it'll vary depending on the situation. Um, but humans still have a role to play, and we shouldn't chuck them out the window. Yes, let's not. Yeah, that I mean, that that whole thing of just human in the loop is a whole other concept that we've gotten into a a couple of times on the show, talking about ensuring that humans are at base level still con- collecting data from human beings to inform things that will impact human beings.

30:47 And yeah, and I should have even mentioned this if from a legal context, you know, California has new regulations as of late last year, I believe, early this year on going back to automated decision-making, but because they they drafted it really broadly, it also includes research subjects when they receive incentives because they become independent contractors. Um and at the moment, it's mostly focused on notice and as it relates to the big picture interaction with the research subject in that context. And that's their, you know, the hiring and firing and in our case, that's, you know, I automation as it's involved with bringing them on to a research study, choosing them for a sample, you know, on board it bring them into a panel or kick them off a panel.

31:36 Um it's which seems real basic, but it's something that I think a lot of people are not looking at yet. Um and so it is a useful reminder not just to go up and look up our information about that regulation, which we have on the IA website, but also just to think about it in terms of just your own respect for the research subjects, even remember not happy with them when we're kicking them off a panel or rejected them from a study, you know, whether or not they're actually, you know, a real person or not, you know, those are other fights for another day, uh but maintaining a reasonable expectation that they're human and that we're human uh will help both sides.

32:22 Did I I was watching your face there, Stephanie. Did you Have you heard about a lot of these things? I did not know. I I'm so like fascinated by this idea of, if I'm hearing you right, Howard, you're saying that, you know, by law in California that research subjects are in that relationship or designated as independent contractors essentially? If you're receiving an incentive, I'd for a participation research study, you are unless unless you're an employee of whoever's doing the research, you are an independent contractor of that company or organization.

32:59 I think the reason it fascinates me so much is because it really butts up against our data quality initiatives as well, right? Where it's like there's so much pressure to make sure that our panels are clean and full of people who are ready to be attentive and and serious about answering questions, but at the same time when you start to think about these as as independent contractors, to what do we owe our respondents in this sort of conflicting, you know, this relationship where everybody has a stake in it, it becomes a lot thornier. It's fascinating.

33:28 A longer conversation there. But certainly one I think I but and but certainly one that I think is worth more people thinking about because there are other there are all sorts of basic legal issues that come up um and it's one of the things I lobby on all the time is trying to make sure that we can continue to treat them as independent contractors and not have to turn them into employees. Sure. Yeah. Yeah, that's a whole other can of worms.

33:58 All All of our research participants now need health insurance. Uh so, since you bring it up, I mean, there are all sorts of uh there are bills in I'm I think I've counted out like 25 states this year that are trying to look at legislation that would allow for uh benefits to be provided to an independent contractor without them automatically being considered an employee. And again, like a lot of stuff related to independent contractors, it is about gig workers. They're not thinking about research subjects, obviously.

34:33 But it would impact us and does that put pressure suddenly on the The insights industry is saying, do we need to start, you know, allowing and providing you know, contributing to benefit structure for our panelists? I would hope not. Uh but it is something that makes me nervous. Brings new meaning to the phrase professional survey taker, doesn't it? And that's a huge piece of it is the presumption under under these laws that someone needs to hold themselves out as a professional research subject to be considered an independent contractor.

35:08 I think this gets us into the next part of our conversation that we had, which is there's a legal side, but then there's also a human side and an ethics side. You know, we're all I mean, and this is not a new conversation, but we've been all competing for the same finite pool of respondents who are willing to take surveys, to join panels, and to donate their time, and participate in these studies. So, I want to talk on something that that is I don't think new and it's relatively simple, but it's important, which is what does genuine respect for a research participant look like, especially in the age of AI where we can perhaps more quickly reach audiences or or we're we're tightening timelines, all of those things. How do we maintain that genuine respect for their time given all these new advancements in the industry?

36:00 Um oh, we could go down all sorts of rabbit holes. We could talk about surveys Surveys that are too long uh and yeah, research studies that collects way too much information, including like insanely detailed demographic information that's not actually relevant to the uh results of the study. Uh like all things that add time and burden on the respondents um and create greater risk of anything that happens with their data.

36:31 Um, you know, if people the more cognizant you are of that, the greater respect you're going to have for the research subject that's taking their time and energy and putting them towards your project. Uh, so again, it was things like that. I mean, respect and transparency, they're key aspects of the professional code of ethics uh, for IA and really for in in most cases or most industries. So, why shouldn't you apply it here? Uh, but that's only it frankly, that's only one side of it. We could go down lots of different areas. Um, did you guys want to talk about the research subject bill of rights?

37:11 Uh, participant bill of rights. Yeah, so yeah, I as part of the global data quality initiative, uh, folks developed uh, a participant bill of rights. Um, and yeah, there's I think there are like 16 different pieces to it. Uh, and some of them are very simple, you know, like I have the right to communi- communication that is easy to read and understand. Well, you know, you look at that and say, well, I hope so. It's like uh, if you're failing that one, uh, yeah, that's a real problem and not just with the research subject. Um, but there are ones that mean a lot to me uh, just for my own experience here at IA, uh, like where it says you you I have the right to know how to contact the company that invited me to the research study.

37:56 For a for years the dating back to when this was like Seymour, uh, Seymour had a program in which we would provide yeah, background and information for companies to share with research subjects about what research is and yeah, sort of a a generic introduction to what is research, how does it work, etc. And there was a phone number that went along with that for a while that people could call to learn more of the basics of research. And some companies, having seen that, thought that, "Hey, that's great. Instead of letting somebody know how to contact me if there's ever a question about the study that they're involved in, that they've been contacted about sometimes accidentally in the middle of the night or in some poor fashion, we're going to plaster on that phone number so they can contact Seymour and we don't have to deal with it."

38:49 Uh so I took angry phone calls for years. Just and emails of all the time from people that were confused about what was going on in a research study and didn't have a way to contact the the research company that was running it. To me, you know, one of the most important things in the Bill of Rights is this very anodyne num- piece in there. But, you know, some of them, again, straightforward, I have the right to be free from harassment or intimidation when it comes to joining or continuing a research study.

39:21 You know, I have the right to know how I can leave a study at any time. What are some other good ones? Oh, I have the right to know if I will receive an incentive for my time, in what form, its value, how and when I will receive it. Being clear with that sort of thing seems a good idea to me because it avoids all sorts of pain and suffering later when people, again, don't have the right expectations. Uh and they're expecting that the incentive will arrive in the, you know, their mailbox or be in their hands uh before they've even finished the study, for example.

39:54 Uh you know, you want to have everybody on the same page. Uh what else? Oh, I have the right to not be sold anything or asked for money as part of a research study. Good first principles of not mixing your marketing and your market research. Marketing with your research, yeah. Uh well, again, whole 'nother discussion. We have lot lots of fun things there, but obviously in an ad tech world the bright lines get blurry.

40:22 Uh and not because of us, but because of what brands want to do with information and the goals that they have. But you know, it's important for us when we are conducting studies, designing them, uh working on and interacting with platforms and vendors, making sure that people understand uh that that's something that needs to be relayed to research subjects all the time.

40:49 That this is for research purposes. We're not going to be using this to sell to you. And frankly, uh principle-wise, if it's going to be you know, a mar- used for marketing in some capacity, you're making sure you're letting them know, getting some form of consent for that. Because it's a change of purpose in what's happening in the interaction and what's going to happen with their data. Well, and I think in the context of you know, and these are not new, but all of the DIY platforms, a lot of this kind of thinking and decisioning sat with people who were very experienced at thinking about these kinds of things. You know, supplier-side researchers who were pretty steeped in it.

41:29 Um on a DIY platform, you've got a lot of people in a lot of different roles coming in and running a survey. They might be a marketing intern, they might be a product manager, right? And they're not going to have that background. And so it it sets up a very interesting tension of like, how much can we build into our guardrails for them with the tech, you know? And that's part of our That's That's a huge part of our responsibility to the respondent experience and to the brand. But some of it you can't do for them. And you know, I just we've certainly had times where I've caught like a a survey in field where someone's like, "Hey." And it's it's been a few years, but you know, kind of doing a soft sell in the survey and just catching it and being like, "Whoa, whoa, whoa, we can't that's not what this is. We can't do that." You know, and it's new rules for people who are outside of market research and it can be really challenging. So, the more I think that tech companies especially can do to help those DIYers, the better

42:27 it will be for everyone. Definitely, it's the the scope of people involved is always growing. Yeah, it is. Um well, you know, to switch gears on us again, let's get into the census if if everyone's cool with that. Howard, you are among other things co-director of the census project and that is, from my understanding, a coalition advocating for strong and accurate 2030 census. I think most researchers probably don't think a lot about the census on a day-to-day basis, but obviously a lot of our sampling frames that we rely on for US-based surveys and population research are absolutely rooted in census data.

43:11 So, my question for you really is, why what why should researchers uh be paying attention to the 2030 census at this particular point in time? And I have a little bit of a pointed follow-up on that from my own perspective, which is that I know that there's a lot of political debate about like whether to count non-citizens in the next and census, uh which we have historically done, right? And so, that's a big change and I imagine it matters or doesn't based on what you use the census for, but in market research, I mean, non-citizens have buying power.

43:47 So, it feels like a fundamental change that could have an impact on market research that's potentially negative. How do you think about it? So, yeah, there's well, there's a lot of issues that go into this. Yeah, our our prime concern is support for the decennial census and also the American Community Survey, the ACS, which is which used to be the census long form, but now goes on all decade long. That data fuels every piece of quantitative research in the country, but also how people put together even their approach in qualitative research.

44:22 And that's not just in the private sector, but also every other government study. It all comes back to a frame that comes from and the upcoming 2030 census and results out of the ACS. So, we're focused on trying to make sure it is as complete and accurate as possible. So, that means day-to-day we're and year-to-year we're battling for funding, but also focus. And that means trying to keep the Census Bureau focused on these core constitutional responsibilities instead of doing what they've been up to since around about 2019-2020 when they started very specifically trying to build their own probability-based online panel to compete directly with our industry, which is something they did really badly and continue to do so badly that they don't have anything to show for it.

45:15 But they've spent money keeps spending money on it and they want to put more into it and put make it the source for a lot of things instead of just buying this research off the shelf or the services off the shelf even from our members and from our industry. They know they're looking to compete with us, which is ridiculous including from the fact that we spend all this energy trying to support them in both advocacy and otherwise. So, that's one piece of it and the other is just generally trying to make sure that the process that leads to the decennial is coherent and well-researched, well-founded, well-tested. And frankly, the debate over the citizenship question being on the decennial census in the last decade for the insights industry that was mostly concerned with are you going to test it are you going to put it into place

46:13 early enough in the process that you'll have time to test it figure out what the impact real impact is going to be and then you know adjust for any you know fallout from that. So in the last decade we were joined with lots of other groups who had all sorts of different motivations and all coming together to try to kill the addition of that question because it was being done in a haphazard way against regular procedure and testing at late in the decade and you know we helped you know win on that at the Supreme Court and everybody's happy but yeah it doesn't matter it doesn't all shake out the same way in the next decade because right away there was already discussion earlier in this decade about adding a citizenship question and again this time around at the beginning of a decade it's not as big of a deal.

47:06 Um again as long as you're committed to a a sensible process of figuring out what you're going to do and testing it determining what the pitfalls are what are the upsides not just assuming it one way or another and following procedure including in this case as it was framed in the decision out of the Supreme Court. So we're not so focused on it and I think a lot of the the focus at you know the federal level um among people that want to not count non-citizens it's mostly focused on the apportionment count. So it's not that they would not be counted in the the census count as a whole but that it would not matter for purposes of apportioning congressional districts or yeah redistricting of uh state legislative districts for example.

47:58 An issue that we don't have a personal stake in as an industry. As long as we still have access to the information, um, an extra question is not necessarily a bad thing. We don't want to jam tons of questions into it cuz then you're going to get less people responding. But, adding a question is not neces- inherently a bad thing. So, it's not that the census just goes to citizens. It's It's that there's a citizens that question and it becomes like a parsing or filtering tool. But, we wouldn't have to do that. We would still have access to the total population level estimations.

48:33 Correct. And And there will still be fallout concerns like if you add this, does that mean you have less people responding because they're concerned about it being there and what might happen with that information compared to the rest of it. Um, I think some of that concern is overblown, but some of it is real. And, again, as with most things, more research needed. Absolutely. Well, Howard, we have a recurring, uh, segment that we do on the Curiosity Currents called Current 101, where we ask our guests the same question, same set of questions.

49:07 Uh, the first is, in your experience, what's a trend or practice in market research that you would like to see stop, just end as quickly as it can. And then, what's one thing that you would like to see more of in market research? So, the one thing I'd like to see end, uh, immediately, is the way that This is the government's fault. Um, is it's how the the US Office of Management and Bud- Budget approaches, uh, research policy for the US. And they demand a ridiculous response rate.

49:41 And I'm talking There's a It's not a specific thing that they have in their own rules. It's how they've allowed it to be interpreted across all the federal government agencies. That they should see you like 70 80% response rate on any survey for it to be considered good or acceptable. Which is absolutely insane and this is across all kinds of methodologies. It's almost unachievable in almost any context, but what results from that is that they are ruining the research subject experience for everybody else.

50:14 Because if you are working on a research study for the federal government, you end up being told that you're going to have to do dozens of for example if it's a phone survey, you're doing dozens of callbacks in the course of a short period of time and that is harassment. The government is requiring harassment of potential research subjects and that's it is offensive beyond all measure and I wish they would just cut it out. I would like that to stop, too. Yes. And certainly you know, things I would like to see more of I go back to transparency. I want to see more communication between you know, business partners and I want to see more communication uh with the public.

50:56 People should know what's It shouldn't be black boxes that they're dealing with on either side. I mean, you don't have to open up the doors to everything, but let's let's open it up a little bit more at least. So, Howard, thank you so much for taking the time in chatting with us today. I feel like this is something that undercuts every single piece of our activities that we do every single day, but not something that I feel gets the spotlight as often as it should. So, let's say there's somebody listening to the show and they really want to triple check, cross the T's, dot the I's and make sure that they are on top of everything that we've discussed today, the regulation, the ethics, the census, being part of that. What what's the first thing that they should do? What what's the best starting point and where should they go from there?

51:42 So, the checklist at top of the checklist is joining the Insights Association. Yeah, it represents the whole industry and you should be a part of it, obviously. But along with that, there is an insane amount of information on our website, insightsassociation.org. Uh lots of useful information there on compliance, on ethics. We have the code there, which anybody can uh review and should be looking at regularly anyways. Um most of the information is really behind the members only wall when it comes to legal and compliance issues. Um which again, why you should be a member.

52:16 Uh we've got a daily electronic newsletter. I like all sorts of things happening across the industry, but also what's going on in the advocacy and see and compliance sides. Uh if you follow us on LinkedIn, we do regular IA on the Hill posts about what's going on on Capitol Hill and on the advocacy side. And even a monthly fighting for you newsletter that I do uh on just rounding up everything that's gone on in a given month cuz there's a lot going on every month as it relates to advocacy.

52:45 I get that every month. I love to look at it. Yep. Yeah, and then yeah, we even have specific things like the General Counsel and Privacy Officer forums for our company and department level members. Um very candid discussion on all sorts of issues uh several times a year. Uh like there are a lot of opportunities that in which you can learn a lot uh sometimes very quickly and get up to speed, but also you can be a part of something very important in advocating for your own industry. We're going to be launching a grassroots tool in very shortly that will allow anybody to contact their policy makers on behalf of the industry in support of industry interests as it relates to everything from a data privacy bill in you know, a given state to support for the census, uh the latest regulation on AI, everything's going to be on the table and you want to be a part of it.

53:42 Amazing. I think that's a I have I have that checklist same thing as Stephanie. I get all your newsletters all the time and that they're super helpful context to always be aware of. Well, thank you again for joining us today, Howard. We really appreciate your time and for for spending the time to share your knowledge with us and and with the industry. Absolutely. I really appreciate you guys taking the time to talk to me cuz there's always lots going on. The Curiosity Current is brought to you by AYTM. To find out how AYTM helps brands connect with consumers and bring insights to life, visit aytm.com. And to make sure you never miss an episode, subscribe to The Curiosity Current on Apple, Spotify, YouTube, or wherever you get your podcasts.

54:26 Thanks for joining us and we'll see you next time.

Produced by aytm

Curiosity Current is made by aytm, the consumer-insights company. Guests speak for themselves; the synthesis is ours. About aytm