What if every market research project could continue contributing to business decisions after its original question had been answered?
In this episode of Tech Talks Daily, I'm joined by Phil Ahad, Managing Director of Data at Cint, to discuss why he believes companies should move away from disposable research. For decades, the familiar model has been straightforward. A business asks a question, commissions a study, receives the answer and begins again when the next question appears. Phil argues that this process wastes useful information and repeatedly asks people for details that may already be available.
His alternative is an always-on human data engine that allows new studies to build on previous research. Existing responses can be combined with first-party information, third-party sources, transactional records and behavioral signals. Phil says this can help organizations answer new questions faster while reducing the burden placed on respondents.
That burden matters because survey fatigue is often misunderstood. Phil does not believe people have stopped wanting to share opinions. The problem is the experience. Customers are repeatedly asked long batteries of familiar questions, often after everyday transactions, because the structure of data collection has changed remarkably little since paper surveys. If researchers already know much of the background, they can ask fewer questions and focus on the reasons behind a person's decision.
We also examine synthetic data, a term Phil openly dislikes, and the growing use of AI personas or digital twins. At one end of the spectrum, a model might add 200 modeled responses to an 800-person study so researchers can work with a sample of 1,000. Phil says this extends an existing data set rather than creating genuinely new insight. At the other end, a company may create a digital representation of a person from survey responses, purchasing patterns, mobile activity and other signals, then ask that representation new questions.
The opportunity is faster research with less repeated questioning. The risk is believing the model knows a person better than the evidence allows. Phil says the industry must test how much information is required to predict an answer with an acceptable level of confidence. He expects progress to come from repeated comparison and validation rather than a single certification method or technical shortcut.
For business leaders, this makes transparency as important as speed. Before relying on AI-augmented research for a major decision, they need to understand where the original data came from, how modeled responses were created, how performance was tested and where human judgment remains involved. Phil also notes that strong decisions rarely rely on a single input. Companies bring together research, customer records, benchmarks and other sources before deciding what to do.
Cint's ambition, as Phil describes it, is to turn recurring tracking studies into a continuing source of insight. He says roughly one million people pass through the company's ecosystem each day, giving Cint an asset that can be combined with increasingly accessible technology. The larger challenge is making useful sense of growing data volumes at business speed.
Could continuous research help your organization ask people fewer, better questions, or would AI-generated responses introduce uncertainty that outweighs the speed gained? Listen to the episode and share your thoughts with me.
[00:00:00] The leading issue of agentic AI in businesses right now is ensuring agents act with compliance guidelines. And Denodo applies guardrails across your entire data estate. By aligning your company's data infrastructure under one system, these guardrails perform consistently across your platform. So start scaling your business and start with Denodo. Simply visit denodo.com to learn more.
[00:00:27] What if the research your company paid for last month could continue answering questions today rather than disappearing into a presentation deck? Well, my guest today is the Managing Director of Data at Cint. And he joins me to discuss why businesses should stop treating research as disposable data.
[00:00:53] And we'll talk about an alternative, which is continuously updated human data engine that combines fresh responses, existing studies, third-party information and tested AI models. And we will discuss survey fatigue, why people are tired of repetitive questionnaires rather than sharing opinions. And where AI personas or digital twins could extend real human research.
[00:01:22] And he'll also touch on why synthetic data is often asked to perform jobs that it was never actually built to do. So for leaders making expensive decisions, the question is no longer whether data is available. It's whether they can trust it and make sense of it quickly enough. And on that note, let me introduce you to my guest. So thank you for joining me on the show today, Phil.
[00:01:48] Can you tell everyone listening a little about who you are and what you do? Yeah, thanks for having me, Neil. My name is Phil Ahad. I'm in market research. I develop products and data solutions. Been in the industry for over 20 years. Recently joined Cint last year to head up our data and data solutions group. We're really focused on how we can maximize and empower people to make the best business decisions possible
[00:02:16] through access of global human intelligence and packaging up with data solutions. Awesome. Well, I'm looking forward to talking about all things MarTech with you and more today. And when we hear about market research, it's often commissioned to answer a specific question that's presented and then effectively discarded. But I've read that you believe a continuous human engine is more useful instead.
[00:02:43] So what does that model look like in practice and how does it change the way that a company can make decisions on this? I mean, look, I've been, like I said, in the space for a long time. And I was, you know, part of that push to give us a question. We'll answer it and then give us another question. We'll answer that one. And, you know, because of the ecosystem and the technology and the products around us, that's how we used to run and operate data. The world's, as you all know, from an AI standpoint, has changed a lot.
[00:03:12] And I think that type of data practice and process, we're just wasting data. And I consider it disposable data. When actually every single one of these data elements and these pieces of information that we collect can be part of your human data engine. It's constantly on. It's always on. It's repeatable. It feeds and builds on something that you want to grow with going forward. And this is where I believe the industry needs to go to.
[00:03:38] These one-off projects, ad hoc projects, every single data element you consume doesn't have to be discarded. It can be part of your foundational data set. And I'm not saying it's going to replace everything you're doing going forward, but it can amplify, it can augment, and you can do more with existing data sources. And, of course, there's another cost to repeatedly starting research from scratch. And survey fatigue and declining response quality is something we're seeing.
[00:04:08] I think just about everything you buy, everywhere you go, you get a survey, how did we do afterwards? So you start to ignore them after a while. So how serious has that problem become? And could always on measurement actually give us better human insight while actually asking less respondents? I think it's a, obviously I think it's a real problem. If anyone has followed anything I've said over the last, even two years, you know, as an industry, you know, in some cases we don't provide the best experience to give us information, to give us the answers that we're looking for.
[00:04:38] We take you through a battery of questions because that's just how we've conducted market research and data collection for the past 50 years. If you go back to paper surveys, to online surveys, to what we're producing today, the construct of the data collection, the survey hasn't changed much. Today, we don't need to do all that, right? And this is kind of the shift. It's like, you know, I've collected a lot of, in some cases since collected, thousands of pieces of information on a single person, right?
[00:05:08] Do we really need to go back and constantly ask the same 25 battery questions? Most times, no, right? And what we can do is we can start to create a really, really great engaging experience for our audience by just asking them what we really need to know. The why behind the action in a lot of cases. And I think, you know, this is kind of where we need to find a balance between what we need to ask and what we really, really should be asking. We're not quite there yet.
[00:05:33] I think by leveraging some of these AI tools and capabilities and by leveraging an always-on data engine, you can get there. You can get there faster. And, you know, the notion of, like, people are getting tired of giving us information. They're not. They're just getting tired of that specific experience. People love to give their opinions. I get it. Unrequested opinions about things every single day from people for free, right? This is where we need to find that balance. We need to change the experience. We need to make a better experience.
[00:06:02] And since working with partners in terms of doing that, in a way, and, you know, I'm not saying we're running out of people. We have one of the largest accesses to people across the world. But we want to keep these people engaged. We want to keep them coming back. And by doing that, it's providing a better experience for it. And something else we're seeing a lot of recently is synthetic data. Getting enormous attention as AI enters the market. But I've got to ask, is synthetic data creating new insights? Can it do that?
[00:06:32] And what do business leaders need to understand about exactly what synthetic data can and cannot reliably tell them? I hate the term synthetic. I think it's one of the worst definitions of what we're trying to do or what the industry is trying to do here within data. I mean, obviously, this is opportunity, in my opinion.
[00:06:53] I think there's a multitude of use cases where this works really, really well and where you take an application expecting you to replace thousands of human insights with it and you miss, right? This is where, as what we're trying to do here at Synth, but also the industries, we're trying to find that balance and where it makes sense to leverage these solutions and these capabilities to potentially create human insights, augment, append, and expand data collection, right?
[00:07:24] We're testing the full spectrum of it from simple boosting. So I've conducted a survey to 1,000 people, maybe only got 800. I really need 1,000 to cut and slice that data for statistical significance. I'll boost the other 200 based on the data I've collected. This doesn't necessarily create new insights. This expands in existing data sets, allows you to do more of your research and data cuts, right?
[00:07:50] That is, to me, the, I'm going to say call it simplest, but it's a really bad term for it, but it's the starting point where I'd say with synthetic works, and it works as kind of like a production, a data production environment. And then you go all the way to the other end of that spectrum. We call them AI personas. The industry calls them digital twins, where it's like, Neil, you know, I've asked you a bunch of stuff about yourself. Maybe I'm tracking your mobile activity or social activity. Now I can create a digital representation of Neil.
[00:08:18] You know, I know what Neil likes to eat, what Neil likes to do, where Neil likes to shop, how much Neil makes or how much his family spends, right? Every single month. And now I can start asking your digital twin questions, actually conducting the research through your digital twin. This is where the industry wants to go and where we actually have the capability of going to.
[00:08:39] What we're trying to close the gap was like, how much do I need to know of Neil to be able to accurately predict in a case, those specific insights on those data? And this is kind of that spectrum of where we're exploring. Synthetic capabilities, AI capabilities, digital twin technologies allow us to then reduce some of the burden on the people giving us that information. But what we're trying to figure out is how much do I need to know of that person to be able to accurately potentially replace that person as part of the process?
[00:09:08] And that sounds a million times better than synthetic data. So we will banish that word. But I mean, it does, in some circles, people might be listening and say it might raise a big question around trust. So if an executive is about to, I don't know, make a multimillion dollar decision using AI generated or an AI augmented research, what should they ask about the underlying data, the validation, mythology and transparency before trusting that answer from, I don't know, let's say digital Neil, for example?
[00:09:38] Yeah, the good thing is like the industry went through this, right? They went through this with offline to movement to online in the early 2000s, right? And if you go back to those use cases, that was the same discussions. How am I sure, right? How do I know? How is this validated that I can actually rely on this data? And we probably spend years doing it. And all of a sudden, it just transitioned.
[00:10:02] You'll see like phone surveys, offline surveys were this much and online surveys were this much. And then the whole flip happened, right? I predict the same thing will happen, right? At the end of the day, there's not going to be one true way to test and validate, right? The methodologies, the data collection process, and then how we predict, which is actually what we're doing when we ask a thousand people questions, right? We're collecting all this data and then smart people on the research side or the business side takes that data, takes that insights.
[00:10:32] And what are they going to do? Based on that information, they're going to predict how those people would potentially behave when that product hits the shelf or that marketing campaign goes live. It's exactly what we're trying to do here within a hybrid approach of human-based research, synthetic and AI and digital twin-based research. It's the information that I get, how confident am I going to be to make that business decision on it? And the only way we're going to get through that is to continuously test and start running data and start running research projects.
[00:10:58] My opinion is there's not going to be one silver bullet here where we said, oh, this is validated. This is proven. We're going to have to go through it through the work. And where some of these synthetic solutions have failed is that we've taken, like what I mentioned, a boosting model and have said, hey, this is going to be an insights creation tool. And the results haven't made sense. That's where the failure point is.
[00:11:21] This is why we're testing right now and we're exploring how much information do I need on an audience or a person to be able to run them through that model and predict it. We know this works. We know it works because the big tech companies, I mean, Facebook, Meta, I can tell you right now, they can predict what I'm going to do tomorrow better than I can if you ask me. Because they know so much about my behaviors, what I like and what I don't like. So from an infrastructure standpoint, we know it works.
[00:11:49] From a data collection research standpoint, what we've got to figure out is how much of that gap do I need to close to be 85, 95% accurate in my prediction. And we might have a few business decision makers listening, thinking research is simply too slow or too expensive. So for what you're seeing out there, I'm curious, are businesses increasingly choosing good enough data or social listening or gut instinct instead?
[00:12:18] And what does anybody listening in the research industry, what do they need to change to win them back? What should they be doing? I mean, disruption and then innovation comes from those specific needs. It's like it's too slow and it's too expensive. Now, on the expense side, I don't think it's too expensive. I do think it's too slow, right? And I think this is where we're pushing for these type of solutions to be as part of the game.
[00:12:45] And I mean, everyone at P&G, brand decision makers, insights decision makers, they're all kind of dealing with the same thing. They're not relying on one single source of data to make that decision, right? They're taking multiple different inputs of data across first-party data collection, third-party data collection, their CRM information, their benchmarks, and then they're making that decision. And that's really how good decisions come through. And that's really what we're trying to do from a persona-based work, right?
[00:13:15] Is, yeah, we're a first-party data provider. But every single person that comes through the SynthExchange, we also have layers and layers of third-party data integrations from transactional data to SKU-level data to voting habits to social habits. And we're expanding. If you ask the person 10 questions, we have the potential to add another 100,000 data points on top of that person so that you can make the best possible decision off of the first-party data that you've collected.
[00:13:42] To me, that is the future of insights and decision-making. It's like you're not making your decision based on one input. You're making it on multiple facets of inputs. And now with all the technology at our fingertips, we're able to digest this data in a very scalable, quick way to make those decisions. And one of the things that stands out in what you're doing here is this hybrid approach, combining real human input with AI prediction, digital twins, et cetera.
[00:14:08] But for people listening, how should organizations decide which method to use for which problem and where should humans remain firmly in the loop? How do you use this? Honestly, it's on a need-by-need basis, right? There's going to be certain audiences where these tools work really, really well because they're so enriched and they're so nuanced on that information.
[00:14:29] And there's going to be some work that's really, really narrow, specific healthcare work or specific B2B work where maybe a model, even a SYNT, doesn't have enough information. Right? Now, in the live stream of data collection, we could start to collect and the model could say, you know, actually, you've collected 70% of a human, right? The next 30%, we can create it through our AI persona technology. Right? So, you know, you got to take this project by project based at the moment.
[00:14:59] Again, some verticals, it's perfect fit, right? You're going to get really, really good data out of it. And some you may need to collect in real time and then some you may need to collect at all. And what excites you about where we're heading at the moment and the work you're doing at SYNT? You're probably working on a lot of things there. What excites you about where all this is heading and what you're working on right now? There's no shortage of data right now. So, obviously, that gets me excited.
[00:15:25] I get excited when we integrate a third-party data provider and we match at 80% of our respondents. And then all of a sudden, you're like, whoa, I've just added 3,000 transactions to this person. And then you're like, oh, shit. How do we make sense of this? Right? And you got to build tools to be able to digest or ingest and then make sense of it. It's, you know, we are getting to a point where we are, there's no shortage of data.
[00:15:53] There is shortage of how do you make scalable sense of it. That's what really gets me excited right now. We've never been in a world where you're so empowered and so enriched on making that decision, but you're challenged with how quickly you can make that decision in an effective way. And that's what I'm excited about. It is an incredibly exciting space at the moment because researchers can collect human and advanced data at machine speed.
[00:16:24] And if you look at yourselves at Sint here, you've got a global footprint and programmatic access to hundreds of millions of consumers across 130 countries. And you also connect researchers, advertisers, and brands to reliable data from the world's largest network of high-quality sources. So is there anything you can, I don't know, give me a use case or something that you've seen out there at the moment to bring to life how you're helping companies overcome so much of the problems and the noise
[00:16:53] that we're hearing online? Is there anything you can share? You don't have to mention any names here, but just to bring to life the problems of solving. I joined Sint late last year. And again, I've been in the industry for a while. And I'll just tell you the reason why I joined Sint. Obviously, I have a long-standing history with Patrick Comer, our CEO here. And a lot of the questions when I was kind of exploring, what do I want to do next? Like, why don't you start your own thing?
[00:17:21] Well, what got me really excited is that there's not many assets out there that actually run through, have a million people coming through their ecosystem every day. And the technology has never been so easily accessible to anyone, either in the market research or inside space. Like, in my career in this industry, we've always been chasing that technology because we've had to create it, right? And now it's never been so easy to adapt and integrate into it.
[00:17:49] And actually, the gap is the data. And when you have an ecosystem where you have access to a million people across the globe, you can close that data gap very, very quickly. So where we're helping our current clients right now and where we're seeing the best traction is, look, you're launching studies with us consistently, right? Just what you've collected in that moment, that's just one piece of the puzzle. As part of our ecosystem and our data exchange ecosystem,
[00:18:16] we can amplify every single data point that you've collected and actually take it a step further. If you're doing a tracking study with us, there's no reason that every single one of those waves become a static thing. They can be an always-on data engine where we can ingest all your waves of tracking. You don't have to wait till the next wave to add insights to that because you're so enriched and it's so deep. You can use the persona model to start creating new insights off of the back of that.
[00:18:41] And then the tracker can constantly give you that real-time nuance in the moment data to continuously feed that model. These are things we're exploring and we're building on today since it's not just a basically data and survey collection tool. We are an always-on data engine and that's what we're moving towards. I think that is a thought-provoking moment to end on. But for anyone listening wanting to dig a little bit deeper on anything we talked about today, start a conversation with you or your team or just keep up to speed with some of the big announcements
[00:19:10] that could be coming out later this year, where would you like me to point them? Go to our site, Synth.com. Link in with me, Philahod on LinkedIn. I'm happy to talk to anybody. Awesome. Well, I will have links to everything you mentioned there. I advise anyone listening, if you're interested in anything you've heard today, please check out the links and get in touch. But more than anything else, thank you for bringing all this to life today and talking about it in a language that everyone can understand. I really appreciate you, Tom. Thank you, Neil.
[00:19:40] Appreciate it, buddy. I think Phil's argument gives businesses a different way to value research. A survey does not have to become a static report, especially once the original question has been answered. Because with the right permissions, supporting data and testing, each study can then go on to contribute to a growing source of insights that improves the next decision. And the caution is equally important.
[00:20:08] Yes, AI personas can fill gaps, extend samples and test possible responses. But confidence must be earned through repeated comparison with real behaviour. After all, a convincing prediction is not automatically a dependable one, especially when the commercial decision attached to it carries a very large price tag. So I will include links to Phil's LinkedIn profile.
[00:20:35] And remember, you can also find out more about Sint at Sint.com. But a big thank you to Phil for joining me today. But over to you. Would your organisation gain greater value by treating research as a continuing asset rather than a succession of isolated projects? Let me know. You can find me at techtalksnetwork.com. You can find more there about how you can work with me,
[00:21:01] contact me, leave me an audio message, or meet me at a tech event. We've got a lot of big events coming up, so it'd be great to meet you there. Other than that, that is it. We're out of time today. I'll be back again real soon with another guest. But thanks for listening today. Bye for now.