AI Is Rewriting the Market Research Job—Not Erasing It

A market research director posted on Reddit a few months ago under the title "Worried about job." The post was three paragraphs of raw professional panic. He had a blue-chip client list—YouTube was on it—and in the span of a single month, all his active clients had gone quiet. Attempts to work with their bosses were going nowhere. He'd been "constantly anxious about job security," he wrote, describing it as "a rant and a super messy view of the disjointed conversations that go on in my brain." His most vulnerable line: "I feel like I don't have transferable skills for any other jobs." He used AI occasionally—to draft emails.

That post could have been written by thousands of market research analysts right now. The anxiety is rational. NielsenIQ "cleaned house throughout 2025," according to a thread in r/Marketresearch. "Staff were impacted at every level, from entry." This is industry-wide structural disruption, not one person's bad luck.

The Bureau of Labor Statistics projects 87,200 market research analyst job openings per year through 2034. The profession isn't disappearing. But the work inside it is changing fast enough that standing still is its own kind of risk. What distinguishes the analysts who are navigating this from those who aren't isn't seniority or technical genius. It's a clear-eyed read on which parts of the job are already going to AI—and which parts just became more valuable.

What's Actually Automating, and When

The scary headline—that market research analysts face a 90% risk of AI displacement—has been circulating in career forums and generating genuine dread. That number deserves a direct response: it's based on theoretical task exposure, not observed employment outcomes, and the distinction changes everything.

Here's what the actual data shows. Gartner surveyed over 400 CMOs in early 2026 and found that marketing leaders expect AI-driven automation of their work to more than double, from 16% today to 36% by 2028. That's the most direct near-term signal for analyst task portfolios—roughly a third of current marketing and research work going to AI within two years. Perspective AI reports that AI-moderated interviews now cost $8 to $15 per completed interview, versus $150 to $300 for traditional interviews. That's not efficiency. That's a price collapse that reaches your client's budget conversation.

But the 90% replacement claim conflates task-level automation with job elimination, and the empirical evidence won't support it. Anthropic published labor market research in March 2026 finding no systematic increase in unemployment for AI-exposed occupations, and that 57% of AI usage involves augmentation rather than automation. The Yale Budget Lab concluded that theoretical AI exposure scores have "only a limited correlation" with actual employment outcomes. The World Economic Forum ranks market research analysts as the 10th fastest-growing role globally, projecting 53% growth by 2030.

The honest picture: roughly 40 to 50% of the current analyst task portfolio faces high automation risk within two years. The tasks automating fastest are the most time-consuming and, frankly, least interesting ones—data cleaning, secondary research synthesis, basic report formatting, survey coding. What isn't automating, and what the data consistently shows growing in premium, is the work that requires knowing why a number matters to a specific stakeholder in a specific business context.

This same pressure applies across roles adjacent to market research—UX researchers, marketing strategists, customer insights managers all face the same task-triage calculus. Anywhere that analysis has historically been bundled with interpretation, AI is splitting them apart.

Two Veterans, Two Bets

Knowing that disruption is task-specific and survivable is one thing. Being persuaded that adaptation is actually realistic for someone deep into their career is another. Two veterans with over 70 combined years in this field made very different bets when AI began automating their core work. Both are still standing.

Soumya Mohanty is Managing Director at Kantar with 25 years in market research. When AI started simulating the synthesis and pattern-matching he'd spent decades perfecting, his response was to write a book about the Indian consumer. His argument: AI "can't place the micro moment inside the macro cultural sweep." He's betting that lived cultural depth—the kind that comes from decades of field work across industries and contexts—beats processing speed. He's doubling down on expertise that AI hasn't lived and can't train on.

Now I am told AI will replace me. That a language model can do what I do — synthesise, pattern-match, generate insight.
— Soumya Mohanty, Managing Director, Kantar

Ray Poynter has spent 47 years at the intersection of research, technology, and thought. In July 2023, rather than defend against the tool, he co-founded ResearchWiseAI to automate the survey analysis he'd spent his career perfecting. His bet: the fastest path to survival is building the thing that threatens you. He repositioned from practitioner to AI operator—commoditizing his own craft before someone else did it for him.

Mohanty's path is available to anyone with deep domain expertise and the discipline to make it explicit: industry-specific pattern recognition, cultural knowledge, institutional relationships built over years. Poynter's path is available to anyone with the appetite to learn the technical layer and position themselves as the person who runs the machine. Both paths work. Neither requires starting over. The question is whether you have irreplaceable expertise to deepen, or the temperament to build alongside the tool that's reshaping your field.

The Task Triage

The director who posted on Reddit knew what he should do. What he didn't have was a map for where to start. Here's that map.

Three tiers, organized by what the data actually shows about automation speed and value trajectory.

The first tier is automating now, within the next 18 months: data cleaning, secondary research synthesis, basic report formatting, survey coding, competitive benchmarking from published sources. These tasks are going to AI regardless of preference. Contesting the timeline wastes energy better spent elsewhere. The weekly action here is simple but requires honesty: identify one task from this category you do every week, find the AI workflow that handles it, spend two hours learning it, and stop doing that task manually. Not eventually. This week.

The second tier is shifting value, over the next 12 to 24 months: survey design, qualitative coding, AI-moderated interview interpretation, thematic analysis. AI now handles roughly 70% of this work. Your value lives in the 30% where it reliably fails—catching cultural misreads, flagging methodological overconfidence, identifying the question the AI didn't know to ask. The weekly action: the next time you use an AI tool on one of these tasks, write three sentences about what it missed. That critique is becoming your job description.

AI is not replacing market researchers. It is quietly rewriting the research stack.
— Eva Guterres, Director of Market Research and Intelligence, Great Minds

The third tier is growing premium, and this is where investment pays off most immediately: research problem framing, stakeholder communication, strategic storytelling, ethical oversight, primary research design. This is where AI is structurally weakest. Gartner and Anthropic both point to human judgment in ambiguous, high-stakes contexts as the durable source of value. The weekly action: take your most recent research project and write a one-page strategic interpretation memo—not a summary of findings, but a recommendation for a specific business decision. Practice owning the room, not just the data.

This triage applies regardless of seniority. The director with 15 years and the analyst with 3 face the same three tiers. The proportions of each in their weekly work differ; the structure doesn't.

The More Radical Option

For readers whose current portfolio is weighted heavily toward tiers one and two, there's a harder question worth asking: what does it look like to build with AI rather than adapt to it?

Avi Yashchin came from quantitative finance—high-frequency trading, then Analytics Lead at Two Sigma—and founded Subconscious AI to replace traditional survey panels with synthetic respondents modeled on behavioral data. He didn't adapt to the disruption. He became it. His path wasn't linear from market research; it was lateral from quantitative analysis and a willingness to apply it somewhere entirely new.

The WEF's 53% growth projection and ESOMAR's finding that the $140 billion industry is growing at 6.4% year-over-year—with AI-native methods as the only category posting double-digit growth—tell a specific story: the profession is expanding precisely where AI is most involved, not despite it. Analysts with deep domain expertise in consumer psychology, survey methodology, and cultural interpretation have something Yashchin had to acquire: substantive knowledge of what market research actually needs to do. That's a real advantage in the AI-native layer of the industry, for those willing to bring it there.

For most readers, the triage in the previous section is the right next step. For some, the more aggressive pivot toward building is worth considering seriously. The distinction comes down to honest assessment of your portfolio, your risk tolerance, and your appetite for construction over adaptation.

Start Here

The director on Reddit wasn't asking the wrong question. He just didn't have a map for the answer. The analysts navigating this successfully aren't the ones with less to lose—they're the ones who got specific about where to put their energy and started with the smallest possible concrete step.

Three things to do this week, in order of importance. First, audit your last two weeks of work tasks and sort them into the three tiers from this article—automating now, shifting value, growing premium. Be ruthless about which tier each task actually belongs to. Second, take one recent research project and write a one-page strategic interpretation memo—not a summary, a recommendation. This is practice for tier three, and the only way to build that muscle is to use it. Third, find one job posting for a senior research role in your sector and read the required skills carefully. Note which ones have appeared in the last 18 months. That list is your upskilling map.

The craft isn't going away. It's just finally becoming the whole job.

Recommended Tools & Resources