How much of the bolded AI-ready tasks have you implemented or are planning to implement in the coming year?

Priya said she is currently in the use-case identification stage, surfacing these opportunities internally at Google Cloud and seeking engineering support to build capabilities. In parallel, she is actively engaging research agencies to find out which of these tasks they can already deliver. She estimated a realistic timeline of 9 to 12 months to see some of these in action.

For early-career researchers who spend most of their time in templated, repetitive work, what is your advice on career path?

Priya recommended over-indexing on prompt engineering and AI QA and design, both through formal training and practical experience. She suggested seeking roles at AI companies or at agencies building AI functionality, and volunteering for proof-of-concept work. She also emphasized developing people skills, particularly change management, as a long-term differentiator.

Aren't you working yourself out of a job by volunteering for AI proof-of-concept experiments?

Priya said she does not feel concerned, because the majority of her current work sits in the blue-box, high-judgment categories that AI cannot replicate in the near term. She added that she has not seen evidence from any AI tool she has tested that it can do what experienced researchers do in those areas. The risk, she argued, is greater for those whose work is predominantly templated and repetitive.

Exploring the Line Between Human and AI Value in Market Research

Priya Mathur, Senior Customer Intelligence Manager at Google Cloud, walks through where generative AI can realistically take on market research tasks today, and where human judgment, context, and collaboration remain irreplaceable. Drawing on 20 years in the profession and a front-row seat to the AI category's rapid growth, she challenges researchers to map their own daily work against this line and to proactively shape how AI enters their practice rather than waiting for it to be imposed on them.

Key Takeaways

Map your daily tasks against an AI/human line before tools are chosen for you. Researchers who do this proactively will have more say in how AI enters their workflow.
AI is already capable enough to handle templated, repetitive, and first-draft tasks, covering an estimated 30 to 40 percent of a typical research process today. Those gains in time and attention should be reinvested in higher-judgment work.
Human value concentrates in collaboration, organizational navigation, adapting to change, and connecting research to broader business context. These are the tasks AI cannot yet replicate.
Prompt engineering and AI quality assurance and design are the two skills most worth building now, especially for early-career researchers. The quality of AI output depends directly on the quality of the prompt.
Not all methodologies will be affected equally. Researchers should evaluate where AI plays a larger or smaller role across the specific methods they use regularly.
Being frontfooted matters. Volunteer for AI proof-of-concept work, surface use cases to leadership and agencies, and treat experimentation as progress even when individual tests fail.

Questions & Answers

How much of the bolded AI-ready tasks have you implemented or are planning to implement in the coming year?: Priya said she is currently in the use-case identification stage, surfacing these opportunities internally at Google Cloud and seeking engineering support to build capabilities. In parallel, she is actively engaging research agencies to find out which of these tasks they can already deliver. She estimated a realistic timeline of 9 to 12 months to see some of these in action.
For early-career researchers who spend most of their time in templated, repetitive work, what is your advice on career path?: Priya recommended over-indexing on prompt engineering and AI QA and design, both through formal training and practical experience. She suggested seeking roles at AI companies or at agencies building AI functionality, and volunteering for proof-of-concept work. She also emphasized developing people skills, particularly change management, as a long-term differentiator.
Aren't you working yourself out of a job by volunteering for AI proof-of-concept experiments?: Priya said she does not feel concerned, because the majority of her current work sits in the blue-box, high-judgment categories that AI cannot replicate in the near term. She added that she has not seen evidence from any AI tool she has tested that it can do what experienced researchers do in those areas. The risk, she argued, is greater for those whose work is predominantly templated and repetitive.

Session Notes

Context: Why This Topic, Why Now

Priya Mathur opened by framing the moment: generative AI exploded as a category in roughly the nine months before this talk, and Google Cloud, where she works, sits at the center of that conversation alongside Microsoft. Every profession is being asked what AI means for it. Market researchers are no different.

Her lens for this session was deliberately narrow and practical: not AI as a broad technology story, but AI as a day-to-day question for the working researcher on either the agency side or the client side.

Mapping the Research Process Against AI Readiness

Priya walked through a representative set of tasks across the standard research process, from scoping and design through fieldwork, data processing, and reporting. She identified which tasks AI is already positioned to take on, at least in part, and which still require human involvement.

Tasks Where AI Can Add Meaningful Value Today

Generating a draft RFP from a set of parameters and a template
Crafting a question bank for a given research type (segmentation, persona, market opportunity) once business objectives are supplied
Running sample quality checks and flagging unnatural variances during soft launch
Identifying standard cross-tab views and charts given the survey instrument
Generating a report wireframe or populating a standard deck template
Producing a first-draft analysis with recommendations from data, written in conversational language
Monitoring data variance across tracker waves, a task currently done with conditional formatting and prone to human oversight

Her estimate: AI can get a researcher roughly 30 to 40 percent of the way there on many of these tasks today. The gap is largely context, not capability.

Tasks Where Human Judgment Remains Necessary

Collaborating with stakeholders, managing political dynamics, and navigating how recommendations will land with different audiences
Adapting scope, design, and timelines in response to org changes, shifting priorities, or evolving team structures
Connecting a given piece of research to adjacent work across the organization
Extending insights to new use cases beyond the original audience (for example, surfacing sales research findings to marketing or operations)
Ensuring insights reach the right people at the right time to influence real decisions
Understanding industry sensitivities, audience nuance, and organizational context that shapes how findings should be framed

AI is not going to understand that hey there was an org change and now so and so teams roll up to this other team. We need to change how we've crafted our recommendations.

Three Categories of AI Advantage for Researchers

Standardized and templated tasks. Rinse-and-repeat work like standard question banks, smoke-jump sample monitoring templates, and tracker wave variance checks. AI can surface red flags in real time without the researcher having to stare at a complex spreadsheet.
Repetitive tasks prone to human error. Sample quality checks, data variance reviews, and quick calculations that suffer when attention is divided. AI can deliver more consistent quality here.
First drafts of creative work. Survey instruments, discussion guides, report frameworks, and initial recommendations. AI gives the researcher something to react to and refine rather than a blank screen to fill.

Not All Methodologies Are Affected Equally

Priya flagged that in any given day a researcher might run an IDI, send a survey, review social benchmarking data, and discuss a competitive intelligence project. AI's role differs meaningfully across each of those. Methodologies that lean on creative judgment, context building, and direct human engagement will be less disrupted than those built around data processing and templated outputs. She encouraged attendees to evaluate their own methodology mix with this lens.

Skills That Will Make Researchers AI-Proof

Audience polling during the session ranked change management first, followed by influencing for impact and storytelling in a tie for second. Priya endorsed those results and added her own priorities.

Change management. Navigating how AI enters an organization, who adopts it, and how its outputs are integrated into decision-making.
Influencing for impact. Making research land with the right people and drive action, a skill that depends on reading relationships and organizational dynamics.
Storytelling. AI can write bullets. It cannot read the room, track body language, or shift emphasis as a conversation evolves.
Organizational behavior. Understanding how decisions are made, who owns what, and where accountability sits.
Prompt engineering. Priya's personal top pick. The quality of AI output is directly tied to the quality of the input. This is a learnable skill with immediate payoff.
AI QA and design. As research tasks become automated, someone must evaluate the automation, write the requirements for AI functionality, and catch hallucinations and errors. This is an emerging role that early-career researchers should actively pursue.

What Researchers Want AI to Do More and Less Of

A second audience poll asked what researchers most want AI to enable. Results, in order:

More innovation
More creativity
Less administrative work
Less quality checking
Less rinse-and-repeat activity

Priya's advice: do not wait for an agency or an employer to bring AI to you. Surface use cases proactively, reach out to agency partners to ask what they can already demonstrate, and volunteer to run proof-of-concept experiments within your organization.

Risks of AI That Researchers Can Actively Mitigate

A third poll asked where attendees want to focus their energy on managing AI risk. Results:

Complete AI outputs with context (most popular)
Review for hallucinations and errors
Identify viable research use cases

Other risks Priya named that researchers are positioned to address: smoothing and scaling AI adoption across the organization, and solving for security, privacy, and legal vulnerabilities. She drew a direct parallel to the early days of social listening, which was similarly overhyped as a universal solution before finding its proper place as one tool among many.

Advice for Early-Career Researchers

Priya was asked directly about the risk of working yourself out of a job by championing AI. Her answer: if the majority of your day sits in templated, repetitive tasks, that is a signal to upskill now. If most of your day involves collaborative, consultative, and context-heavy work, the near-term risk from AI is lower.

For those early in career, her specific recommendations:

Pursue prompt engineering training, certifications, and hands-on practice.
Raise your hand for AI QA and design roles within agencies or organizations that are building research automation capabilities.
Consider roles at AI startups or at research agencies actively building AI-enabled services.
Hone people skills alongside technical ones. Change management, influence, and EQ are the competencies AI cannot replicate.

We can either have AI be inflicted upon us, or we can be champions of taking AI and making it work for us. Backfooted versus frontfooted.

Transcript

Read the full transcript

hi everyone um I'm going to pull up my screen here and share my presentation in just a second so Bill if you could let me know when you can see everything we'll get started I can see everything awesome awesome all right so hi everyone my name is priia mothur and I am a senior customer intelligence manager working at Google Cloud so the topic that I wanted to walk through today is very near and dear to my heart because as you may have heard um Ai and specifically generative AI is a very hot topic for the tech sector um this past year and uh you know Google is sort of sitting very much um right next to Microsoft at kind of the Forefront of this conversation so Ai and and the role that AI plays in our lives is sort of this everpresent topic this year um and really the way that the lens that I wanted to look at um the topic of AI and generative AI for this uh presentation is really what does this mean for the modern Market researcher because I suspect that a lot of us on this call right now have probably asked ourselves like well like you know what does AI going to mean for me right every profession on the planet is having to reconsider what is AI going to mean for them and we are no different so before I get into some of my content for today um just to introduce uh myself um I wanted to share that I am a 20-year market research veteran I've been in this category or rather in this profession for for a good bit of time um but I will say that I see myself first and foremost as an avid customer Advocate um so the way that I position myself in every interview and every you know sort of organization that I work in I see myself genuinely in my purpose as a market researcher to really be that advocate for the customer um Within especially within you know sort of the client side of the business um beyond that I also see myself sort of a little bit differently from the other marketers that I work with in the sense that I am data driven measurement grounded and customer obsessed um as a marketer versus many of my other marketing counterparts and so in many of the roles that you'll see me talk about on this chart like my my um application of customer and market research skills has generally been in the marketing space um I graduated a little too long ago to mention um from the from the University of Texas at Austin so go long horns um in and with a degree in economics um I very quickly from there moved to sort of really learning my market research chops like how to design a great survey how to write a solid guide on the agency side at milward Brown which is a canar company and a wpp company and this was back in Austin as well um shortly thereafter I made my first move to the client side supporting ad sales research um in Florida with Comcast and specifically the ad sales arm of Comcast which is called Comcast Spotlight um and then after that I ended up moving to Seattle um where I joined Microsoft and specifically within Microsoft I joined the Xbox again ad sales team uh within Xbox uh so in a research capacity working with our ad sales teams selling ad slots on Microsoft's advertising platforms um and then sort of that kind of um wrapped up the BTO C sort of spectrum of my career so far and I moved uh very intentionally into the B2B space which I was far less familiar with and I did so working on Azure so I spent a good I'd say probably the last eight to nine years um working on working within the B2B space or a total of8 to nine years working supporting azure Microsoft Azure and that really just sort of placed me in this Plum spot to really see the cloud category being born which uh suffice it to say we learned a lot sort of seeing this tectonic shift in technology come to life grow and to the point now where it's mature um and then sort of just in time for this next tectonic shift in the tech sector uh about two years ago I moved to Google Cloud where I stepped in here in the role of senior customer intelligence manager just in time to see the AI category explode so seeing a lot of corollaries to the birth of cloud and the birth of well not really the birth of ai ai has been around a while but really generative AI making AI as a as a topic far more relevant and within reach of every single person on the planet um sort of that is sort of this new chapter of AI so we've really seen the geni specifically category uh or piece of of AI explode within the last nine months and I've been on that ride for the last nine months so the reason why you know sort of that's a really great segue to what I wanted to talk to you about today is that when we think about ourselves as researchers we all know that there are aspects of our job that could very easily and probably should be um usurped by AI um and so when I think about this right and we think about the standard research process of you know starting with scoping moving to design going into fieldwork and data collection then going through data processing and ultimately to reporting and socialization you've heard some of the other speakers today talk about socialization um and so uh you know just sort of looking and mapping out this journey both on the agency side as well as the client side there's you know I've taken a sort of smattering of tasks this is by no means comprehensive but really was just meant to give me a couple of data points by which to really assess which of these pieces or parts of the journey is AI really well positioned to help us as Market researchers and I want to pause there to actually think about this right like many of us have and and and I've been doing the research to see that a lot of a lot of customers have concerns about AI essentially taking over their role or taking over their job while a good chunk of other folks in the same functions have ALS Al had a little bit more of a of a lean towards how is AI going to empower me as you know as a market researcher as a marketer as a operations specialist so suffice it to say that it's really important for us as Market researchers to really start to tease out the line in all these tasks that we're responsible for whether you sit on the agency side or whether you sit on the client side it really is incumbent upon us to evaluate our day-to-day with a lens of what pieces of this could I like could I offload or automate in some way and what does that free me up to spend more time and energy delivering value through so again none of the tasks on this list should be surprising to the folks invol um generating an RFP um I think everybody on the agency side has probably been exposed to that part of the process of scoping right uh proposing methodologies for a project seeking efficiencies to drive project profitability things crafting questions for your draft questionnaire um evaluating you know sort of quality sample and and creating sample checks identifying the key data views that you want um to you know generate your cross tabs with crafting a report wireframe and building charts and slides all the way to presenting findings and making edits and additions and Corrections these are all some of the things that kinds of tasks that I think we've undertaken on the agency side and then on the client side you know sort of similar um similar functions on the other side of the fence are like evaluating rfps for fit to business context managing costs looking for efficiencies reviewing the research instrument overseeing a collaborative effort to lock um you know to lock a survey with your internal stakeholders um looking at partial data you know re-evaluating the research scope and design based on you know those first couple of interviews that you complete or that first set of completes that you see coming into your survey um co-creating and you know reporting um uh Co co-creating a report wireframe or crafting research recommendations and ultimately even things like presenting findings and connecting research to uh a broader array of research you might have so connecting the dots so to speak these are all tasks that either on the agency side or on the client side we've undertaken I'm going to pause here for a quick moment Bill do we have any questions or comments before I move on to the next slide uh nothing yet okay fantastic so just to keep going so when we look at that same list and the thing that I tried to do here is sort of pull apart some of the things that I can imagine right off the bat that AI could help us out with right generating an RFP we can load up RFP templates and you know sort of put in a set of parameters and have a tool generate uh an RFP like a a sample or a draft RFP for us we can have an AI um capability that crafts you know questions for us if we type in the business objectives of a given piece of research or say that hey we're looking to do a segmentation or we're looking to do uh a you know Persona like a a Persona piece of research or we're trying to do you know sort of um Market opportunity analysis what are some standard questions within these types of research projects that I should you know sort of put put on the put on the drawing board we can have ai help us conduct um sample checks right um monitor very closely soft launch and point out any unnatural variances we could have ai identify um some key data views that given the questions that are already sort of pre-loaded into our survey what are some standard charts and and data views that would be really meaningful um for us to you know for for the AI to automatically generate for us that AI that AI capability can also generate charts and slides create a report wireframe um now how close that is to fit for purpose is going to be a big question mark and something that we'll iterate you know with an AI tool over time but the point of of Bolding these particular bullets is just to say that right out the gate we could probably have an AI product or service get us at least 30 or 40% of the way there um today right and the reason why I did this exercise is really so I can start you know trying to tease apart what are the kinds of tasks here that AI is particularly helpful for and what are the kinds of tasks that I really want to reserve more of that human sensibility and human um you know sort of context building for even on the client side right inspecting partial data quality we can assign that or automate that in some way um co-creating a report wireframe ident ifying stories coming up with initial recommendations customizing those recommendations to individual audiences these are all some things that AI could help us in large part or at least in small part with um and again these are subjective so I am by no means presenting myself as a subject matter expert saying definitively that these are things AI can do today it's merely observations and then the reason why I named this session exploring the line between human and AI value is because it is just that at this point in the sort of um introduction of generative Ai and AI into our world it is I think a little unfair to expect anyone to purport themselves as a subject matter expert at this point this is more ideation and observation so um you know feel free to disagree with what I've you know sort of bolded here in terms of bullets that AI could do for us that's not the purpose of this conversation I think the purpose of this conversation is really to provoke thought and ask us to evaluate our daytoday and look at maybe for you some of these bullet points might be bolded a little bit differently but look at what is your day-to-day chart out those various tasks and really consider and ask yourself where is it necessary for me to intervene and where is it perhaps helpful and empowering to have some of these functions automated or or relegated to AI so some of the kinds of observations that I came up with was um and I did this in the wrong order so I'm going to I'm going to just blow this out but couple of the things that I observed right out of the gate of where AI could really do the types of things that I personally don't want to do right number one where AI can help is really with those standardized and more templated tasks the things that we would categorize is rinse repeat right and those are things like standard questions right from a question Bank um sample evaluation or sample monitoring um with many of our large scale trackers um both on the client side as well as on the um as well as on the agency side I've generated what we call Smoke jump templates which is where you know you look at all of your sample metrics as your data is coming in and you have these very complex and you know sort of uh to some degree overwhelming spreadsheets of data points where you're looking wave over wave how much is my sample for this wave's results deviating from previous waves and we're looking at that variance right we're looking at that data variance very closely we've typically done it using conditional formatting today but there's no reason why AI couldn't spot some of those red flags for us and surface them for us real time right generating things like a report template I mean especially if you're on the agency side and you have a standard way that you build out segmentation decks a standard way that you build out value or concept value Testing Research right have an AI tool generate and populate some of that for you um without you having to sort of stare at a blank screen first other things would be like repetitive tests that are typically prone to human error so like I mentioned sample quality checks data variance checks sometimes when you're looking at these very overwhelming charts and slides it can be easy for the human to overlook some small things right like Overlook small variants do you know quick calculation that you might have done incorrectly right or or just you know sort of at times we just don't have full focus in some of these repetitive repetitive tasks so here as well AI can play an outsize role in helping us um just sort of deliver better quality um in some of these repetitive tasks and then the third space where AI really should be taking more of The reigns is even taking that first taking a run at that first draft of even some creative tasks like drafting my survey or drafting my guide knowing the kind of research that I'm doing just give me sort of the standard question bank right already generated into a dock for me or already you know generated into um you know uh into uh an Excel spreadsheet or or Google Sheets right so these are all things that that first draft can absolutely be done with with the use of AI simple things like a draft analysis and Report I've actually been thoroughly impressed by Ai and specifically generative ai's abilities to craft recommendations based off of the data that are you know very cogent that that actually sound like a human root them right like they're not dry in sort of this you know very you know botlike um you know sort of rendering of a sentence these are conversational these are you know these are great first runs however what those first runs leave a ton of room for is context right what they lack is an understanding of the business an understanding of the humans that are actually going to be reading some of these decks so that first run though can absolutely be taken by an AI um and then even things like drafting recommendations right these are all things that I would love to not have to stare at a blank screen but have uh but have an AI take a first uh an AI capability or service take a first run at the things where we as as humans become far more um sort of capable of delivering unique value and this is I come I I raised these simply by looking at like the previous two slides right and looking at patterns of what did I see were things that got bolded and things that didn't get bolded the things that didn't get bolded were were you know sort of activities or tasks that required collaborating with others right and AI is not quite at the point yet where it can attend a meeting understand business context understand urgency understand political landscape right and help you with things like ideation adaptation based on the changing needs within the business or coordination which might have some political wrangling elements of it right anybody who's worked in a mid to large siiz organization knows that there is some amount of political angling that takes place even with research right how you're going to land recommendations what you're going to say even the word choice of your recommendations and your data points right like your your slide headlines have to be sometimes carefully crafted um being mindful of different org you know relationships um who takes ownership of what right where is something given equal accountability versus singular accountability these are all the types of things where knowing people right and understanding how to work with people and that EQ element of the research that we do is critical the second space where we see um AI not really up to the Mark or able to quite you know sort of serve the the human function is adapting to change right and AI is not going an AI tool is not going to understand that hey there was so and so org change right and now so and so teams roll up to this other team we need to change how we've crafted our recommendations or we need to you know sort of respond to changing Team Dynamics in roles and responsibilities or we need to adapt to changing timelines um at this point that is still something that requires a human to notice a human to recognize has implications for the research and a human to then sort of mitigate um you know based off of the changes that are taking place um in the research so again adapting the research based on changing priorities changing Team Dynamics and changing timelines is something that really a human can probably uh a human is still needed for today and then the last thing sort of I observed across the bullet points in the previous two slides is really docking to the bigger picture um AI could give us that first pass but it can't quite fully understand the context of you know sensitivities uh within a particular industry or within a particular audience or within a particular organization right so understanding the business understanding in the category and understanding your customer is still very much uh within the realm of of human value right where creating connections um is something that humans are uniquely capable of doing that that an AI function couldn't possibly do today because unless you're hooking up every piece of knowledge you have of your customer um and that includes qualitative research quantitative research Telemetry social data unless you're taking all of these pieces that you're using to know your customer and that includes like account team interactions right all of these pieces where you touch your customer loading them all up to a singular system and having the AI be able to read all of those customer interactions and all of that customer Insight collectively it's impossible for an AI to to have that bigger picture right where are those disparities to other research where do we need to potentially partner with another team who is doing an adjacent piece of research and connect the dots and AI capability is not going to be able to do that today another thing is extending your research so thinking about other use cases for the set of you know insights or or recommendations that you've come up with right you generated this report for the sales team how does that apply to marketing or how does that affect operations is not something that that sort of business context that rich understanding of the business is something that AI can't replicate today and then the last piece of that is relevance right um making sure that these insights are surfaced to the right people at the right point in time to affect real-time decision making again this is all context timeliness that AI can't fully serve but what we can say is that if AI can be leveraged to do these top three things right like help us with more standard ized and templated tasks help us with repetitive tasks that are prone to human error and help us with you know sort of that first draft of some of these creative tasks then that gives us as researchers a lot more time and energy and mind share to apply towards where we really have the ability to add that value which is collaborating with others adapting the research to evolving to the evolving needs of the business and and in in imbuing it with Rich business context again I'll pause here um before I move on Bill any comments questions um no it looks like we're still clear okay so one of the other things that I wanted to kind of um point out um before we kind of go to a little bit more of the Q&A portion of of this of this presentation is is the notion that all methodologies are not going to be affected equally by AI um and again this is my you know sort of I've worked across a number of these methodologies and um in my sort of 20-year career span and just sort of stepping back and looking at where do I really think AI is going to play an outsized role versus a relatively smaller role and being mindful of really sort of these blue boxes versus these red boxes right like where is there more intrinsic human value in Creative you know creative functions context building functions you know sort of change management functions and people engagement functions right it becomes a little bit clear that not all of our methodologies are going to be impacted by AI equally and so this is more again just food for thought or a different angle at really evaluating our day because in a given day I might be doing Idis in one piece of research I might be sending out a survey for my second meeting today um as part of my third meeting today I had a conversation about social benchmarking and you know my last meeting today is going to be about a competitive intelligence piece of research AI is going to have a different role to play in each of those and so I didn't necessarily go as far as you know starting to like break down and evaluate how is how and where can AI play A Part a bigger part or a smaller part in each of these methodologies but that is perhaps something for us to consider next next right is where is the hum the unique human value in these different methodologies and how can we uplevel and really Empower ourselves with technology as researchers for these methodologies differently and with that I wanted to actually turn it over to the room a little bit and ask you all a couple of questions because these are more sort of things that I'd like this group to consider either immediately now on this call or certainly in you know in the time uh that you sort of perhaps ruminate on on the meaning of AI for our profession over the next couple of weeks and months but question number one is what skills will make us AI proof right let's really think about that right and based off of what I've talked about I put a couple of things on here but this is Again by no means an exhaustive list so I'd love to invite folks to you know sort of add responses to this list or or you know potentially um you know sort of weigh in with really where do you feel you're going to want to maybe invest in some of these skills to AI proof yourself but a couple of the ones that immediately came to mind for me based off of what I've just shared with you today are things like change management influencing for impact storytelling organizational behavior prompt engineering prompt engineering is a big one my world right now is centered around um prompt engineering because this is something that the entire sort of like well especially in some of the social data I've been looking at recently prompt engineering is becoming something that every person who engages with technology today is trying to figure out how do I get AI to work for me and work well for me as we as Market researchers know so many times the quality the quality of the answer comes down to the quality of the question question right and that is what prompt engineering is all about the output you get from the AI is only going to be as good as the prompt you put in so if I was to pick a particular favorite in here I would certainly choose this one um and then AI QA and design because even as a number of these tasks that I pointed out that I bolded in the previous slides become automated or you know sort of market research companies find ways to offer products and services that automate these functions somebody is going to have to Pivot their day-to-day job from being a market research analyst or a data analyst to being the person who who does the quality assessment on the AI doing the same thing right who designs the AI functionality and writes the requirements for the AI functionality that enables us to do those things in a more automated fashion but here again I'll pause Bill do we were we able to get the poll question up and do we have any answers we did and we have let's see most popular is change management followed in a tie for second with uh let's see influencing for impact and storytelling yeah I mean if just going back to the sort of the blue boxes that I highlighted previously those are kind of the big ones right like those are spaces where AI can't possibly compete right AI is at least today day AI can't possibly compete because AI could give you storytelling but it can't read the room right like AI can write a compelling set of bullets but it can't like I said read body language it can't read the room it can't you know sort of um change and evolve as the conversation in the room progresses right so storytelling is a big one reading the room really that EQ of understanding and knowing your audience those are all skills that are going to become increasingly important for us as Market researchers and set us apart from something that can be automated let's go to our second question so the next question that I want you to ask yourselves is let's really sort of as I've done in this exercise so far let's reconsider what we want to do versus what we have to do right every single one of us has parts of our day that we despise and parts of our day that we thoroughly enjoy right so what pieces of our day do do we really want AI to serve right and what pieces of it do we want to reserve for ourselves so again just based off of my personal experience I listed out a few right I want to do less administrative things I want to do less quality checking I want to do fewer of those sort of rinse repeat kind of activities and do less oversight and monitoring and then personally I I I've personally come alive or really experience joy in my in my job when I get to do more innovating I get to you know opportunities to be more creative more ideation and brainstorming and driving more impact but let's pose this question to you and see where do folks feel you all really feel that AI could do more of versus less of I'm just curious what do you want to sort of like what is it that you want to see AI enable for you again Bill I'm going to pause here and and turn it over to you do we do we see any results coming in we do in fact all right so we've got at the top more Innovation and more creativity followed by less administrative and then a few votes a piece for less quality checking and less rinse and repeat great and again here like I would really sort of challenge all of you to to sit back after after you know after your day is done today and think about you know what are those spots that AI can you know sort of Empower you can serve you as a market researcher and be the person who Champions that right reach out to your agencies and say Hey how do we build something like this um or be the person within your organization that says I want to do less of this I want to bring in an AI tool and experiment with an AI tool that allows us to do more of this or less of this right be that change agent within your organization and get in front of this right really embrace the technology that is coming at us at the speed of light because we can either sort of have ai be inflicted upon us or we can be champions of of really sort of taking Ai and making it work for us right so backf footed versus frontf footed right I would really challenge everyone on this call to to take a stab at being more frontf footed and then the last question that I have for you is what are some of the risks of AI that we can mitigate because like everything else I mean I remember you know as a market researcher when social listening was this new kid on the Block right it was this brand new methodology and it was being purported as the thing that's going to solve everything right it's going to have all the answers and it didn't right it has some great answers for some very important functions but but it is a tool in the researchers toolbox similarly AI is not golden right AI doesn't have the answers to everything AI has its pros and it definitely has its cons right so I'm curious from your perspective what are some of the cons or risks of AI that we as humans can sort of really take charge of right we can as I was just mentioning identify research use cases we can smooth and scale adoption of AI across our organizations we can review and be sort of that quality assurance and design kind of contact UM within our organizations to watch and sort of help evolve AI into something that is actually productive for our business by reviewing it and correcting it of its hallucinations and errors we can complete it with context right which is again something that uniquely we're in a position to do and we can can solve for security privacy and legal vulnerabilities again this is by no means an exhaustive list of the risks or the challenges that come with AI like with any other new technology but I'm curious where does this group sort of feel like is your first thing that you want to do with AI right do you want to road test a couple of these Solutions and review them for hallucinations and errors do you want to be sort of at the Forefront of smoothing and scaling adoption of AI for your business where do you really want to put your energy if you were to pick one of these things to focus on within the next let's say within the next 3 to six months again bill I'll lean on you to tell us what folks are how folks are feeling about this all right we got coming in at number one complete with context followed by review for hallucina hallucinations and errors and then third most popular being identify research use cases fantastic I love love the fact that this group is sort of really embracing this topic because you know like I said we as researchers can choose to kind of let this happen to us right and and let the chips fall as they may which is a very unintentional and sort of accidental relationship to have with AI or we can choose to get in front of it and I love the fact that you're all engaging and hopefully starting to really think about the role that AI could play in your in your life and how it can help you Excel at and sort of really make yourself stand out as a market researcher so with that that's all that I had for all of you but I wanted to open up the floor for any questions comments and like I said I am by no means a subject matter expert on this topic so feel free to you know disagree feel free to come in and chime in with your own thoughts but I wanted to open the floor up for others any comments or questions or disagreements Bill and while we're waiting for some of those questions to come in I have a couple of my own um and and for you priia how much of the bolded uh material on your on your uh intro slide have you uh either implemented or planning to in the in the coming we'll say year um so I am currently in the stage where I'm ident ifying all of these as use cases for our business right so um like every other company on the planet Google and Google Cloud are rapidly sort of asking the question of hey everyone what ideas do we collectively have as marketing Organization for how we can leverage AI in our own business and in our own day-to-day so I've been surfacing all of these as use cases for research at this point where I'm where I'm sort of um where I'm at in terms of progress is these are all being evaluated by our team to figure out where can we get funding and support like specifically engineering support right to actually build out some of these capabilities the other thing that I'm doing is engaging a lot of our research agencies to say what pieces of this do they feel that they can do or they can deliver within relatively short order so not only are we trying to build some of these capabilities for ourselves on the client side I'm also sort of actively and avidly reaching out to my agency saying hey show me show me if any of these are possible with you today and the ones that are going to stand out are the ones that are going to say hey you know what we've been playing around with this we've got something to show you so again like benefits of being frontf footed and I want to say like I spent a good chunk of time at Microsoft have a lot of love for Microsoft um despite the fact that I work at Google today so want to be very clear about that my loyalties are very clear but I have a lot of love for Microsoft and one of the things that I really enjoyed about my tenure of about ten years working at Microsoft is Saia nadella's growth mindset um sort of mentality that he impressed upon the entire company several times is this idea of let's not be afraid to experiment right I'd rather have a a mult a multitude of failed experiments and be able to show that sort and be able to show celebrate that failure but show progress ra rather than be the ones who say like we're not going to start something we're g not we're not going to try something until it's perfect so I happen to have the privilege of working in two companies both Microsoft and Google where experimentation is how we get things done and so I am blessed honestly to have uh teams and teammates and leaders that are pushing us to actually um take these these um use cases and turn them into capabilities I'd say realistically the timeline on being able to see some of these in action is probably going to be closer to 9 to 12 months but we're we're actively discussing [Music] them well and it's you know it's that time of year I guess use the analogy of all right next year planning is underway right you know we're figuring out we're creating those big spreadsheets and matrices of you know resource allocation right what do we have upcoming what's the likelihood that it's going to come along you take a swag at what the cost might be and what's going to be stacked up against it is it a actual human or is it something more electronic and you know I feel like that's the stage where we are with with with all of this you know we're kind of figuring it out but if we're not at least thinking about it and you know creating agendas for it um we do risk being being potentially left behind you know and I've been asked by I Mentor a couple of students from the University of Washington and I've been asked by a couple of them you know sort of aren't you afraid that by volunteering for some of these proof of Concepts and experiments that you're working yourself out of a job and it's a good question it was a painful question but it was a really good question because I do often find myself right in these positions wanting to raise my hand and saying I'm going to I'm happy to be the you know sort of be the guinea pig for this generative AI use case or I'm happy to be the person who writes this wreck that is eventually going to take you know 20% of my day off my plate right and I'd say to that and and what I told them was that if 100% of my day is spent Within These red boxes one it's not a great great day and two you know it's it's something that I should then actively consider sort of how do I upskill myself right whereas I know today 80% of my day is spent in the blue boxes I'm I don't feel concerned that I'm going to work myself out of a job because I know that a lot of what I do today can't be replicated anytime in the near future and I by near future I say within the next two years by AI I haven't seen any evidence that of any AI tools and I've been playing around with a bunch right like I haven't seen any evidence from any AI tool that they can do what we do um especially in these blue boxes so a big sort of challenge I I sort of um make to the folks on this call is that evaluate you know what proportion of your day sits in these red boxes and if it's a large proportion then start thinking about where and how you upskill and really s that brings us to this question that i' posed of what are those skills that will make us AI proof I know a lot of the blue box material um is you know it's things that folks who are more seasoned in the business or are you know kind of falling into this territory right yes you know for those of us who have you know made our way to the more consultative side of research um probably there's a little bit more security to be felt but you know for younger analyst level folks who you know they live and breathe and red right um what advice would you give for you know how to I guess approach the the career path as it were yeah I mean it's a great question and a really important one um I would say that first off you know hone the skills like hone the people skills right like hone the change management types of skills and so again when I come back to you know sort of what are those skills that will make us AI proof right like this whole list applies to people that are earlier in career in market research but perhaps none more than the last two right prompt engineering right generative AI certifications right um and trainings right like uh things around uh AI QA and design because AI is nowhere near like we don't have the volume of use cases at this point um to really say that hey these tasks are no longer important we are firmly in the era of where AI capabilities are being designed shaped and and you know sort of these prompts are being written to make humans successful so if you're coming out of college right now or if you're still early in career learning the skills around how to you know how to write great prompts to get AI to work for you how to look at at you know various AI um products and services go work for an AI company right and um and there are a number of startups right in the AI space right now that are just exploding right like work roles in those um or even like even at agencies that are just now starting to onboard and build some of their own AI functionality right put your like raise your hand for a QA and design roles where you are actively ort of checking AI product and and running the proof of concepts of these services and capabilities within your organization because these use these use cases are only going to multiply right when you're done with one there's going to be three more when you're done with those three there's going to be 15 more use cases for AI in the research Journey so I'd say for folks that are just coming in net new into our profession prompt engineering and and you know AI QA and design are probably two spaces that I would over over index on um both getting like certifications training and practical experience in and it's something that some of us more seasoned um Market researchers are probably going to be less prone to go towards indeed um but even you know the more seasoned we're obviously tasked with you know training the Next Generation so we we need to be just as aware and just as involved in the process indeed well I'm not seeing any other questions coming through which I believe could be a good test case for uh our audien's propensity to take part in quantitative research and polls not so much in the qualitative I don't know take that for what it is for what it's worth it's all good all right well priia this was fascinating I think this is a great topic that you know we're we're going to keep talking about this over and over for the for the next several years and you know who knows how long into the future from there um so yeah my encouragement to everyone is if you're not if it's not top of mind right now it needs to be because it's it's happening and whe whether we like it or not right indeed and with that um you can all find me on LinkedIn if you just type in um I can certainly I don't know if you have my LinkedIn profile handy Bill and can put it in the chat but if you don't you can simply search for preauth or Google cloud and you'll find me um so if if you'd like to connect on LinkedIn come find me I'd love to continue the conversation