Decision Shapers, Not Insight Generators: Reframing the Value of the Insights Function

Hello, welcome everyone to our 3:00 presentation. We're very pleased to welcome Sarb here for the presentation decision shapers, not insight generators. I appreciate everyone who's hanging out this late in the day to continue to hear and learn some interesting stuff. I'll be helping moderate the session. What that means is I will keep an eye on the webinar chat, any questions posed, and be checking for any questions that are coming in through the live YouTube stream. And we'll bring those up toward the end of the conversation so we can facilitate a great conversation. Sarb, if you'd like to introduce yourself and take us away, that would be great. Thank you, Luke, and thank you for all who are still hanging on. It's pretty late for some of you in the evening. I'm Sarb Dana Brown. I'm a global insights lead and omni shopper insights lead for the Clorox company, and my career's mostly been in FMCG working for some of the the large companies that you all know, always in insights and in marketing. And today, I'm going to be signaling some thought starters on where our function is heading from my vantage point. And many of you may be head nodding already that our industry in the insights and analytics function particularly is being challenged quite heavily by AI. And I think about a year and a half ago in this same forum, I was presenting what it will take for insight professionals to be successful going forward where AI was only just beginning to impact our industry. And here we are a year and a half later, and it is a is a disrupter in a good way. There's some watch outs for us, and that's what I'm going to be signaling today. So, thank you for your engagement, and happy to obviously facilitate questions as we as we work through. But you know, the headlines is that there's never been a shortage of data or insights. They're not scarce. They are no longer scarce scarce. But decisions are rare. Outcomes are rare when it comes to our function influencing those outcomes. So, think about that for a minute because this is a really important not even a pivot, but a reframing of what we are in service of. And the problem is really hiding in plain sight. It's always actually been there. It's just been amplified and brought visible now given the advent of AI. There isn't a shortage, as I said, of of insights or data that our teams can bring forward to our marketing, our sales partners, our CX team members across any organization. But there have been some changes along the way that have been very rapid. No longer is weeks of analysis acceptable. It didn't really change influence any kind of decision or outcome. A lot of report generation was validating some of our assumptions, and I'm sure a little bit of that still takes place today. Multiple teams working on pretty similar asks from organizations within each of the functions. Shared insights create not clarity of the outcome we're driving, and it's not always easy to know who that decision maker is. But also, the democratization of data and insights inside companies has fostered a culture of that everybody owns insights, everybody owns data. So, driving clarity in well, who is the influencer of the decision becomes quite murky. And reporting insights, as I said earlier on just now about the the findings, is kind of too late. What our teams are certainly looking for is earlier signals. We don't have the time for those two-week turnaround reports. It's kind of making or influencing those decisions in the moment. So, I say this all all because that AI, although it is scary and exciting for many of us, it is also an enabler. And I'm going to talk a little bit about that in shortly. And and this is the great segue. While it is exploding at a rapid pace, AI isn't the disruption of our industry, it is the exposure of our industry. It's made the gap impossible to ignore, and I'm talking about the gap of moving away from creating decks and just report generators into what does the actual information mean in terms of driving a decision or influencing a decision? What are the choices that are in front of us in a business, and you as the insight lead, how can you shape those decisions going forward? And that constraint has moved because before back in the day, there was scarcity around data. Well, we don't have data for this particular retail, or we don't have data for this particular behavioral set. We have lots of data and tools at our disposal now. We have an abundance of signals. In fact, that's the overload, the overload and overwhelmingness of exactly what is available to us today. There's no shortage of companies coming towards us saying I can do research faster and better. I have this AI built in. Answers are now just table stakes in that respect from any of our teams. But what is the opportunity for our function? Is driving clarity. What does all of this mean? How can I frame it in a way that brings value and influences an outcome or an action that will be taken by our stakeholders? So, the human element of driving clarity and judgment still exists because the machine cannot do that for us because we know our business better than the machine. But what these algorithms, these tokens, these the all the AI applications that are out there that can help us is remove some of the drudgery. They can drive tasks in a faster way for us, for sure. So, our impact is not the output of reports. What what our impact is what value we drive on the outcome. That's the new currency. And the success of our function going forward is whether our work has changed a decision, whether it's driven an outcome, not how many reports we generate, not if we repeat another A&U or or large framework. It's what do we do with this in order to drive a decision that will ultimately grow our businesses, whether that is market share, whether that's entering a new category, a new space, or an adjacent business model. So, I mentioned AI has exposed the gap about the work. It forces a reckoning, you know, what what does really matter within our work and what decisions we can we can drive. Too much analysis causes the paralysis. We need the analysis to drive better outcomes and better decisions. And so, what I will say to you is and and the logos you see here, whether it's Google Gemini or Open AI with ChatGPT or Claude with Anthropic or Microsoft Co-pilot, whatever AI your organization is using, I encourage you highly highly to be making sure you use that, and you create your own agents because that will be the next pivot and game changing for game changer for us and how those agents can amplify the work we do, obviously within the confines of confidentiality and within the firewalls of your organization. But I will encourage you to experiment and use as much as possible because the trajectory of where AI is going, quite frankly, is scary and exciting, and we're in a historical moment, I find, for not just our insights and analytics industry, but for marketing, for any organization in that is operating with brands at scale. Uh with our sales partners, with the supply chain, finance, you name it. AI will and [clears throat] is impacting all of those areas at a rapid pace. So, experiment, pick a personal project, try and learn. It has to be embraced for our function to be successful going forward. So, AI creates the insights, and it can create them really really fast, right? You can template a research study. You can use DIY platform tools and farm that study out to panelists, and you can get results back in probably two two to three days. AI can make sense of that from analyzing it. But what we drive is the clarity of that information, the synthesis, the judgment. The action still remains human because we know our commercial goals, we know our marketing goals. And that is really through that relationship we have with our internal stakeholders. So, while the machines can quicken up the pace of some of our responsibilities, what it cannot do and should not do, quite frankly, is apply that judgment piece of what's right for the business going forward. And that is a very human task and and skill that is pertinent. So, our roles are shifting, right? And that is changing fundamentally the redefinition of what our function stands for from not just answering questions, which will still exist, but to framing the decision or the outcome and defining that choice before work begins. From delivering the report, from delivering the insights to then influencing those decisions and outcomes. I think, you know, in the industry to pause, um over the last 10 years or so, there's always been a conversation around what is the value of insights. We don't want to be order takers, but we want to be business partners. And that still remains true. I think the nuance is that business partner relationship is ever more close and intertwined with the internal stakeholders, because only then can those decisions be influenced. So, it's no longer a siloed operation of support function, but trying to understand, well, what are the business issues and how can then insights and analytics be in service and support of driving those decisions. We still have the same data. In fact, we have better data today. Um we have a more expansive toolbox of what we have available. We have some different questions, but quite frankly, a lot of the questions are still similar. But we are driving different outcomes. And what I mean by that is not the reporting, but what is the information illuminating in order to drive the outcome for the business. That frame does change the value of the function versus I'm going to conduct another brand tracking survey and update our brand equity scores, which does have a role to play, but what is that actually influencing? So, one of the things that I do encourage my teams to be thinking about as they receive demands from the business is to take a pause. And I call it the 30-second decision reframe. Before anything begins, think about the four questions that you should be asking yourself and asking of your business partner, whether the work really matters. And if it doesn't, how do we reframe that? Or we say no. And it's really starting with the end in mind as opposed to, well, what are your what are your objectives? That's still helpful, but really when you start to poke at what decision are you driving with the stakeholder, they tend to pause. They tend to think more intentionally about what they're asking, because what they asked for may not be really what they're looking for. And so, these four questions help us decode what really matters. What is the decision? Call it out in terms of what you are trying to inform. Who is the decision maker? You know, we obviously work with assistant brand managers all the way through to VPs. So, getting clarity on who the decision maker is helpful to help the influence once we start looking through the insights and the analytics themselves. Really define what the alternatives are. What options do we have? What do we know already versus reacting? Um and if there's a gap, there is a gap, and then we then we frame that. And what changes? So, let's try and understand what will really shift the choice or the decision, and that helps clarify the actual inquiry of insight or analytics versus potentially maybe repeating something that's done in the past. So, asking these four questions of yourselves or the partner that you're working with, not only allows for a clearer brief, often it stops work, because we find ourselves, actually we already have that answer. Here it is. But what you're really asking for is over here, and that is a gap, and let me help you frame that and drive that choice. And so, inherently you become a business partner because of the way in which you're showing up and the way in which we're facilitating the conversation. So, it's not necessarily about driving more reporting or stopping delivering insights, it's about starting influencing the choices through the insight work. And that's the conversation narrative. That's where I feel our storytelling and our conversations need to elevate, and some organizations are really great at this, and some are in the middle, and some are just starting their journey. So, in order for to become a decision-orientated kind of thinker, you know, some organizations are structured in this way, and some are not. But drive that clarity on who is actually making the decision. And if you're aware of that, I would very close relationships with with that decision maker. Um but drive that accountability, especially in matrix organizations where there are multiple teams working on very similar goals and outcomes, and driving that decision-making um owner is very, very important, because that will also drive integrated inputs. The last thing anyone on this room wants to do is work on the same request asked by different people. And so, we're in that unique position as insight folks, where we do have a purview to many functions, and we see all the data. We have a real leadership role there to provide synthesis of, well, here's what we know, we've seen this over here, this person's asked for it here, to provide that integrated view and drive more value with our work. And that will drive clarity, too, in terms of uh not just the outputs, but what do we do with it now? End with a clear recommendation timeline. This is all kind of standard stuff, but the pivot here is influencing the decision, influencing that outcome versus here is what the findings actually tell us. So, insights in themsel- in in of them itself, I would say, they're not differentiating anymore, right? Everybody can generate insights. AI has allowed that. But the decisions are driving clarity on the outcomes and what those outcomes actually mean, and there'll be more than one, um is where our value, I feel, is beginning to shift. And that can cause a blending of roles even between, you know, a classic consumer insight professional and an analytics person. Um so, the future to me belongs to teams who are able to think at that business level, from the strategic or tactical outcome level, being able to synthesize and blend their knowledge across various parts of the insight function, and leveraging AI where it can really help them by moving quicker and faster, but where the value is from the human side is applying your judgment, knowing what you know on the business, on the category, or for that retailer in order to drive the better outcomes, because your business partners will not be able to do that. They're running at 100 miles an hour, and we have access to all of the information and all the data. So, I wanted to kind of pause here and just take questions, really. Um this is meant to be kind of a provocative piece in terms of, you know, are your organizations driving you to be thinking about more outcome-based insight work? Or do you agree or disagree with with the ethos of decisions, knowing that in many companies insights are seen to be a support function, but in many companies also a strategic driver. So, I'm going to pause here, Luke, and hand it back over to you. Okay, thank you. Absolutely. Uh so, questions from the YouTube stream. Um you had mentioned earlier in your talk at agents, uh encouragement to design agents, use agents. Uh for those of us who aren't uh quite as up to speed on AI, uh maybe just starting to put a toe in the water there, could you give a quick definition of agents and maybe some examples or case studies of of uh agents that were designed, what they did, and how they facilitated that migration you're talking about from from producing data to enabling decisions? Yeah. So, you know, let's just back up a little bit. You know, when you use any of these kind of um AI chats, right? What they're doing is they're learning all the time from what you're feeding it. So, you know, that's how that's how they're learning, and they're getting smarter and smarter and smarter. The agents, uh let's give you an an example from the travel industry, right? There are travel bot agents out there. Once they learn your patterns, they can organize your holiday for you, because they know your preferences, they know your taste. Um you don't have to spend the time in drudgery trying to look for where to go on vacation. If you've trained your chat and created kind of within the chat a bot, think of that as a virtual person, >> [laughter] >> who knows all your preferences, knows what you've been feeding it, it can organize that all for you. It can tell you even where to stay. That's what we mean by a bot, right? Um that that's a really simple simple example. You know, a bot that I've created, you know, for me personally offline is or one on travel. And from a work standpoint, you can create bots that are very analytical if you're feeding them, you know, kind of the same data set. Uh and it's looking at the same data set, it's very quickly going to learn the patterns you're asking it. So, if you're always looking at a particular age group or you're always looking at a particular uh profile of of of a shopper or a consumer. The next time you feed it this spreadsheet, it's already going to be prompting you, "Oh, do you want me to look at this in the same way or or a different way?" So, you're training the bot as an enabler for you to go quicker and faster. And if you don't know how to create a bot, ask your open AI, whichever one you're using. You know, I showed a few on the page, whether it's Microsoft Copilot or if it's ChatGPT or if it's Claude or if it's, you know, Google Gemini. It will tell you how to create it. And it will teach you how to use it. Okay, thank you. Uh you had shared some examples of AI tools at the beginning of your talk. Do you have any preferences, suggestions, or recommendations for insight-specific AI tools that might be worth taking a look at for folks in insights-related roles? Yeah, absolutely. I mean, uh again, this has to be within the firewalls of the company that you work with. Um you know, for us it's Microsoft Copilot and it's pretty good from an insight level in the in the sense of analyzing information. Uh ChatGPT is very, very good, too. I mean, they're very similar, but you know, there's certain guardrails around that. Um I would say one of the unlocks is Gamma. If you're spending time creating PowerPoint decks, uh I would stop doing that. I would use Gamma. It will create your presentation if you feed it an outline um within 30 seconds. I mean, [laughter] it's it's powerful. Yeah. Thank you. Uh our next question comes from uh the webinar chat, and this is Andrew. What if the client isn't asking for decision recommendations? Sometimes I struggle with whether to provide only insights and then offer a suggestion if they're just looking for data because I don't want to intrude on their business too much. How do you find a balance? You know, I you know, this is the struggle on the agency side. I would put yourself in the position of who's asking you for the information. You know, I have always gravitated towards agency partners who are going to make my life easier. If they had come to me, yeah, absolutely. It's like, you know, what I'm paying for is not or, you know, partnering for isn't necessarily just the information. It's help me understand what do I do with this and what can it inform. You know, I want you to accelerate my thinking. So, to me, I would I would not be worried too much about the balance. I would try it out. I would say from all this information that we see, and if you're close to the business or or your partner that you're working with, is to say, "These are the outcomes the work is driving." Because that's the value then of not just the financial outlay, but the value of working with you as a partner. So, um I don't think any insight again, depends on the culture, whether they would feel, "Oh gosh, you've stepped on my toes." I mean, it's the way you frame it up, you know, thought starters, right? Here are some thought starters for you to be thinking about if you're if you're aware of what's keeping them up at night. So, partly yes, here's what the data tells us, but here's what it really means. Here is what we feel you can do with it going forward. You know, and this goes back to the what, so what, now what kind of framework as well. Okay, thank you. Uh if anyone has a question they'd like to pose either in the YouTube live stream or in the webinar chat, uh feel free to drop it in there. While we're waiting for those to queue up, uh here's a question from me, and uh you know, it's it's kind of a perennial. I think it comes up quite a bit. There's a lot of discussion about how insights is evolving rapidly. Uh how do we future-proof our careers both on the corporate side and on perhaps the market research vendor side? What skills should we be getting? How should we be positioning ourselves to be successful going forward when things are changing so fast? Yes, I love that question. You know um I would say you have to have AI literacy. Prove that you are using AI in a way that enables you and drives those outcomes. That is mandatory table stakes going forward. Um if you do not use AI or have a good literacy, solid literacy understanding around it on how it enables you, you will not be chosen going forward. It's as simple as that, you know, um I've heard of recruitment agencies now not just recruiting for the human, they're asking, "What bots are you bringing with you as part of your portfolio?" Just think about that for a second. >> [laughter] >> Not your skill now isn't academic only. Your skill is what you have been able to create technically um that arms you and gives you a point of um advantage versus the the next person. So, I'm not saying become coders all of a sudden, right? But you need to be knowing enough to be dangerous and to be able to talk to it and know how it fuels the work that you do versus not participate in it. If you do not participate in it, trust me, you you will be left behind. Thank you. Uh the part of your discussion about timelines condensing, I think most of us in the industry have felt that. How can we, if we're on the corporate side, set up a good relationship with our market research vendors? I mean, there are typical project lengths that a big deck is produced. How do we enable our external partners to help get that timeline as tight as possible? Yeah, I would always have the conversation with what AI are they using on on their end. Right? I think gone are the days of 100-page PowerPoint decks. And if those are still happening, there should be really good reason for it. >> [laughter] >> Yeah, you know, because no one's reading it. You know, we're in an attention deficit economy. Nobody has the time or inclination to be reviewing those types of decks anymore going forward. Certainly not in CPG, you know, um and again, depends on the a lot of culture in the company. But I would I would say to you that when working with the agency partners, the value of working with the agency partners is whether that technique doesn't exist in-house, whether there isn't a, you know, fully resourced insight function where you're having to rely on your ecosystem, but moving quicker is absolutely mandatory. Um you know, I had a marketer say to me, "God, 3 weeks? Going to take 3 weeks? That's too long." >> [laughter] >> You know, um so, I would guide our partners to be thinking about how are you moving quicker through the traditional methods by leveraging AI. What are you using in order to move quicker and faster? If it's still a lot of human work, then I would challenge them on that. You know, in in most instances, we are not even asking for PowerPoint decks. We're asking for a two-page written summary and say, "Okay. All right, so this is what it means." And and and decisions are being made within 24 to 48 hours. That's the speed of challenge in the function. So, agency partners, I would get, you know, hot in the press with tools that quicken up data processing, quicken up analysis, and more importantly, automate visualization. Cuz that's where the the slow down tends to happen because our time, agency time, should be spent on the interpretation. Tell me what this means. What is the narrative here? Um that is going to drive an action going forward. Okay, thank you. Uh next question comes from our YouTube live live stream. Could you give some advice on overcoming AI objections within leadership for my org? You know, uh if there are objections still, it depends what part of AI I would say, right? Um I would encourage you to look at some white papers. Here's one of Here's one I came across recently, and it's from Kantar. You will pretty much have access to Kantar. The Palmer Group. If you and eMarketer, if you just look at the thought leadership in this space, it is, like I said, both scary and exciting. There isn't a company that I I've not come across yet that is not using AI in some shape or form because they're forced to. So, the advice I would say is the positioning of it. It's not there to not do the work that you've been hired to do. It's there for you to go faster, enabling faster analysis if that may be a challenge, Um not necessarily a cheat code for doing your work. So if that if that is the hurdle, then I would position it in terms of how does it enable us to go faster? The second is I would always encourage thought leadership to be elevated to those leaders. The recent CMO conference here, you know, the CPG conference, the main topic was AI. That is what is keeping organizations up at night. So I think keeping that external lens of what is happening in the industry to help influence leadership is always always helpful. Okay. Thank you. Uh those are all the questions we have at the moment through the webinar chat and the live stream. Uh while we wait for any more to come in, any uh questions that you've been asked recently that uh you'd like to take a chance to share and answer here that might be of interest to to folks. Questions just just on the topic of AI or decisions or the function any any of the above? Any of the above. Uh what were you hoping to be asked? Yeah. Yeah. You know, I I I think the the reframing from, you know, generating to driving outcomes is an interesting you know, pivot for our function to drive value. And so it's not a question that was asked to me necessarily, but I've been seeing it coming, you know, this this old age question of, you know, from insight functions I just don't want the report. Tell me what to do. Right? Tell me what to do. What is your recommendation? And then the frustration of our group historically being, well, you know, I did this great piece of work and then it's out in filing cabinet. Nothing happened with it, right? Like And then the conversation in the industry was about taking insights to action, right? That's still true. I think that the the nuance on the narrative is that action piece is driving a decision you're aware of through conversations with your stakeholders. It it's not like I do this piece of the work and then I kind of go away. You need to work end to end with them. So my my, you know, not questions necessarily, it's an observation of the work and the way in which companies are operating is that end to end through line. And [clears throat] so once you get a better understanding of those decision points and what will influence them insights then becomes not just here's the report and here's the study I conducted. It's here's the work we can do to drive that outcome. And so we have a better success chance of taking insights into action, right? Versus here's a recommendation of things you can do. Okay. Thank you. >> so that's my observation. Um It's it can be quite challenging, right? Especially where insight functions are seen to be support functions and not strategic driving functions. But that shouldn't stop you because then you're demonstrating the value you bring to the table. Okay. I don't see any more questions at this time. Um what So the industry is one for me then. The industry is changing pretty rapidly. I what's coming next? What's going to be happening in the next few years? Do you have uh do you have a take on that? Yeah. You know, I I think, you know, I'll give you an example from my my current position, which is really on the commercial side and looking at how shoppers decide, choose, and buy, right? And so for me, I think AI where, you know, where we're all learning it and it's getting faster and deeper. Well, what's the next pivot? We're seeing agentic AI influencing our decisions, right? We're now seeing if you're on um you know, watching streaming a program, you're going to see a pop-up from Walmart come up, right? Um on on Tide or or a particular product with a with the QR code. You're being influenced right there. You may not even thought of laundry at that time, but you could just put your camera up to your TV and buy right there in that. Um I think the future is um smaller basket sizes. They're being compressed because of the way in which people choose, buy, and decide products now and brands, right? Because the agentic AI is also beginning to recommend if you're on Amazon, Rufus can recommend to you straight away, well, here are the top five, you know, whatever mattress cover you're looking for. Um you're you're not discovering in the same way anymore. So again, these algorithms as they're learning more about you and become more and more mainstream in the way in which we just live, um the future for me becomes more digital. The future for me becomes more challenging on brand loyalty, right? If the machine is actually recommending something to you. So I see on the retail side, our retail partners from the Walmart's all the way through to, you know, Kroger's, etc. Um playing quite heavily in that tech space when it comes to agentic AI within their own platforms. Yeah. So to me that you know, and that's going to be a rapid increase quite quickly. It's very in its infancy stage right now, but it's already becoming highly highly used. Yeah. So again, you get we're all going to have to learn that kind of well, well, the influence of how people shop or behave is is rapidly changing. Okay. Thank you. Uh I don't see any additional questions that have come up. Thank you so much for your time. Any any closing thoughts or anecdotes or comments you'd like to share? >> Yeah. Well, first of all, thank you for having me. Um it's always always a pleasure. I would say experiment. Pick a personal project. Experiment with Claude. Experiment with ChatGPT. You know, obviously nothing confidential. Start creating a chatbot. Um you're going to become quite enamored by it, but also then distrusting of it. You're going to go through this cycle of oh, I'm not sure. Oh, I really love it. Oh, I'm not sure. Oh, I don't trust it. And then fascinated by it and then probably back to distrust. There's there's going to be some sort of cycle that you're going to go through. It will be useful in certain parts and not helpful in other parts, but you will only learn that by experimenting with it. You know, we've always said in our industry experiment. This is no different. It's just a new way of working. Um it's an evolution and a revolution of the way we work today and we'll work in the future. So we're going to have to get on this bus. Okay. Excellent. Well, thank you again so much. This This has been a great presentation. We really appreciate your time. Thank you for having me. Have a great day. Thank you.