AI-Driven Insights For Market Research

AI’s ability to unlock insights from unstructured data is a massive breakthrough for businesses. I have been beating this drum for a while now. But the real magic? It happens when you combine structured and unstructured data. Here’s why. AI made it possible to ask questions of structured data, like company records, contact records and deal status, and get answers back in natural language. That was a breakthrough. Now, it is possible to ask evergreen questions of unstructured data, like emails, calls, video conferences, transcripts of meetings, and get real-time insights, also in natural language. That is another breakthrough. An even bigger one. But businesses don’t just need breakthroughs. They need results. And to get them, they need insights from both structured and unstructured data—working together. Let’s make it real with an example. Picture a sales leader getting a live feed of every time a competitor is mentioned in sales calls. Even better? AI identifies the salesperson who’s best at handling those objections. That’s unstructured data in action to deliver insights. But there are deeper questions they want to answer, like: Is there a competitor we consistently lose to? Is a new competitor suddenly appearing in deals in specific regions? To answer those questions, they need structured data. They need to cross-check their list of competitors with closed-lost and closed-won reports and pipeline trends by region. Now, they don’t just see what’s happening—they know which competitors to worry about and what messaging works best against them. That’s not just a useful insight—it’s a game-changing one. A smart sales leader won’t stop at knowing which competitor is a threat. They’ll turn that insight into action—launching targeted email campaigns, updating sales playbooks, and creating competitive content. But here’s the catch: AI-powered insights are only valuable if they’re accurate, governed, and respects permissions. AI has opened up a world of new possibilities. The question then becomes: How can businesses turn those possibilities into results? It is by unifying structured and unstructured data with the right context and governance to drive faster action. That's the key to unlocking AI's potential to help businesses grow! And that gets us excited everyday!

Happy Friday everyone, this week in #learnwithmz, if you are a Product manager learning about AI this post is for you. PMs looking to get hands-on with AI side projects don’t have to be expert in AI, just a curiosity and willingness to experiment. Here’s a step-by-step guide to help you get hands-on with AI side projects. 💡 Start small: Automate Regular Tasks Identify tasks you do frequently that AI can streamline, examples: - Feedback theme collection - Feature request prioritization - Market research automation 📌 Example project: AI-Powered Market Research Assistant What is it? A tool that uses AI to gather and analyze market data, customer reviews, competitor strategies, and trending topics, delivering actionable insights for product or feature development. Why build it? - Get near real-time insights into customer needs and competitor strategies. - Accelerate decision-making for market opportunities. - Ensure your product strategy stays aligned with industry trends. Step 1 - Define Scope Inputs: - Customer reviews and feedback. - News articles or blog posts about competitors. - Social media trends and hashtags. Outputs: - Key themes in customer sentiment. - Competitor summaries. - A list of emerging trends or gaps in the market. Step 2 - Choose Tech Stack Web Scraping: BeautifulSoup or Scrapy to gather data from review sites and blogs. Sentiment Analysis: OpenAI, Hugging Face, or #Azure AI Language. Trend Analysis: Google Trends API or Twitter API. Visualization: Power BI or Streamlit. Step 3 - Build and Iterate Start simple, test test test, and refine based on feedback. I’m working on a prototype for this assistant, stay tuned for updates after the holidays. What kind of market data do you find most valuable? Let’s discuss in the comments! #ProductManagement #AI #Innovation #marketresearch P.S. Image is generated via DALL·E

I’ve been seeing tons of Qs on how #AI can be used for qualitative #coding. We’ve run HUNDREDS of experiments on this at Looppanel. We’ve tested, re-tested & re-re-tested, so I thought I’d share the benefits of my learnings with all of you. 💁 1. AI is your research assistant. It can help you code faster, but it cannot replace you. I will say this over and over again, because setting expectations is key. AI is your helper. Don’t expect it to take over your job (thank god!) You should check any AI-generated data and expect it to be a great starting point—not blindly take it as the final answer. 💁 2. If you don’t have very, very good transcripts—you should. Quality of AI transcripts across accents has improved substantially. If you’re spending time checking, correcting transcripts—re-evaluate your tool stack. At Looppanel, we have transcripts with 95%+ accuracy—which means you can have them, too. 💁 3. Replacing a note-taker is now possible. Most researchers WANT a note-taker (because it makes analysis SO much easier), but finding a good note-taker for every call is a challenge. Luckily, note-taking is the kind of task AI is actually really good at. Remove your dependence on other people by adopting an AI-assisted note-taker (again, we already do this at Looppanel (https://bit.ly/4bBE1IO) ) 💁 4. AI-supported theming / analysis We’ve found through deep experimentation that it’s possible to auto-organize your notes by question in your discussion guide. It’s not at 100% accuracy, but let’s say 80-90%—pretty good. We’re currently testing if AI is good at identifying patterns outside of your discussion guide (e.g., identifying that 5 people talked about price being too high). To be honest, the jury's still out on that one—but I will report back with another post once we’ve tested the tech! I’ll keep posting my learnings as we figure out with hands-on testing,, just how good (or bad) AI is at different research tasks. 💁 5. The UX of any AI interaction is actually super important. Whenever we run betas with AI features, we’re partly testing the tech, but often the biggest insights are about UX and content. What tone do users expect? How long or short should a note be? When does it feel overwhelming? How do users discover and explore qualitative data? How do you build trust and traceability into the process? These are just some of the questions we’re constantly grappling with and uncovering via testing. If you want the complete guide on AI + qual coding / tagging, keep reading here: https://bit.ly/3SJbckW If you have specific questions on what AI can do wrt to research, please add them in the comments! I’ll tackle those ones next :) #Looppanel #UXResearch #AIinUX

AI isn't just a tool; it's becoming a teammate. A major field experiment with 776 professionals at Procter & Gamble, led by researchers from Harvard, Wharton, and Warwick, revealed something remarkable: Generative AI can replicate and even outperform human teamwork. Read the recently published paper here: In a real-world new product development challenge, professionals were assigned to one of four conditions: 1. Control Individuals without AI 2. Human Team R&D + Commercial without AI (+0.24 SD) 3. Individual + AI Working alone with GPT-4 (+0.37 SD) 4. AI-Augmented Team Human team + GPT-4 (+0.39 SD) Key findings: ⭐ Individuals with AI matched the output quality of traditional teams, with 16% less time spent. ⭐ AI helped non-experts perform like seasoned product developers. ⭐ It flattened functional silos: R&D and Commercial employees produced more balanced, cross-functional solutions. ⭐ It made work feel better: AI users reported higher excitement and energy and lower anxiety, even more so than many working in human-only teams. What does this mean for organizations? 💡 Rethink team structures. One AI-empowered individual can do the work of two and do it faster. 💡 Democratize expertise. AI is a boundary-spanning engine that reduces reliance on deep specialization. 💡 Invest in AI fluency. Prompting and AI collaboration skills are the new competitive edge. 💡 Double down on innovation. AI + team = highest chance of top-tier breakthrough ideas. This is not just productivity software. This is a redefinition of how work happens. AI is no longer the intern or the assistant. It’s showing up as a cybernetic teammate, enhancing performance, dissolving silos, and lifting morale. The future of work isn’t human vs. AI. The next step is human + AI + new ways of collaborating. Are you ready?

📈 New Adobe data just dropped: generative AI traffic to retail sites surged 4,700% year-over-year in July 2025. But it's not the growth, while impressive, that has me thinking...it's what happens next. Here's what caught my attention and what I think marketing leaders need to pay attention to: AI-referred shoppers spend 32% more time on sites, view 10% more pages, and bounce 27% less. These aren't casual browsers. They're research-driven consumers who arrive knowing exactly what they're looking for. Three things brands need to start thinking about, because the shift in consumer behavior in the Agentic Web is happening fast: - We're optimizing for the wrong thing. 73% of AI users cite LLMs as their primary research source. Keywords won't cut it anymore. Brands need to be the authoritative source that AI systems reference when customers ask questions. - The attribution models are broken. AI traffic converts 23% less but generates 84% more revenue per visit than six months ago. These customers research through AI, then convert elsewhere. How are we tracking that journey? - The infrastructure shift is real. Consumer Electronics and Tech lead in AI visit share because complex purchases benefit most from AI research. But every category will follow. The question isn't if—it's when. For brands who have built great visibility in the current digital economy and are wondering what is happening to their metrics, It feels like the early days of digital all over again, equal parts terrifying and exhilarating. We're not just adding another channel. We're witnessing the emergence of the Agentic Era where AI agents become the new front door to discovery. The brands that recognize this shift and adapt their content, measurement, and customer journey strategies now will own the next decade. Read the full insights from our team at Adobe Digital Insights: https://lnkd.in/g2mGVGud What are you seeing in your data? How are you preparing for this shift? #MarketingStrategy #GenerativeAI #CustomerJourney #DigitalTransformation #AEO #GEO #AISearch #AdobeLLMOptimizer

I've spent 12 months ruthlessly testing AI tools for market research at Impact Theory. What did we learn? We identified market opportunities 6 months before competitors. We 10X'd our research capabilities. We turned market analysis from guesswork into science. But most people get AI market research completely wrong. They're passive. They wait for the perfect prompt. They expect AI to do the work. Those who are killing it with AI take a different approach. I use what I call the "Market Intelligence System": Step 1: Problem Verification Use this prompt: "List the top 5 urgent and painful problems faced by [your target market] with supporting evidence from Reddit, Amazon, Facebook, or other real sources." Step 2: Competitive Gap Analysis "Identify primary competitors and evaluate their strengths, weaknesses. Highlight clear opportunities to meaningfully differentiate my product." Step 3: Market Demand Assessment "Assess current market size and potential for growth. Evaluate key trends indicating increasing or declining demand with evidence from search volumes, surveys, industry data." Step 4: Pricing Intelligence "Suggest realistic pricing strategies and benchmarks. Analyze customer willingness to pay based on real data." Step 5: Validation Framework "Recommend actionable validation experiments to verify all base assumptions. List early warning signs of potential product-market misfit." The nuclear question: "What do people who disagree with these trends say? What are their best arguments?" This process takes me from zero market knowledge to expert-level intelligence in hours, not months. In a world where everyone has access to data, the advantage goes to those who know exactly how to extract insights from it. Most are drowning in information. Be the one who turns data into decisions. I built a free GPT that walks you through the whole process in 30 minutes. It will give you a step-by-step roadmap to launch your business. Try it out here: https://buff.ly/WQHxGFU

I turned 50+ pages of insights on 10 accounts into a 15 minute podcast using NotebookLM and ZoomInfo AccountAI. Here's how I was able to do it: The problem hit me last week. I was headed to dinner with a bunch of customers. Like many GTM leaders, I was staring at a mountain of account insights for dinner. Years ago, this meant frantically scanning Salesforce mobile in an Uber (which never worked). But the world has changed. The explosion of GTM data has created a new challenge. Between CRM records, ZoomInfo insights, earnings transcripts, intent signals and support tickets—we're drowning in information. The best reps somehow find time to synthesize all of this. The rest just wing it. At ZI Labs, we've been experimenting with using AccountAI to reimagine account summaries and meeting prep. The goal? Transform mountains of data into personalized briefings you can consume anywhere. Here's what we did: First, we aggregated everything:  📊 Account summaries  📝 Recent earnings call transcripts  🎯 Buyer intent signals  🔄 Customer support interactions  💼 CRM opportunity history  📧 Historical email/meeting notes Then we let AccountAI do the heavy lifting:  🧠 Extract strategic priorities  🔍 Surface competitive insights  ❗ Identify shared pain points  🎯 Map product-market fit signals  📈 Highlight recent org changes  🔢 Calculate propensity scores Finally, NotebookLM transformed this into audio:  🎧 Natural conversational flow  ⭐ Prioritized by relevance  🔄 Context preserved  💬 Key quotes included  📝 Clear narrative structure The result? A 15-minute personalized podcast covering everything I needed to know about all 16 accounts attending dinner. I listened to it on the way to dinner. No prep required. This feels like the future of GTM intelligence. The days of "winging it" are over. Every rep should have their daily meetings, accounts and opportunities summarized into audio briefs they can consume anywhere (this extends to every profession, but sales will be first). Think about it—your calendar automatically generating custom podcasts with everything you need to know about upcoming meetings. Your opportunities summarized with competitive insights and next steps. While you drive to work. 🤯 This isn't science fiction. We're doing it today. PS - Huge thanks to Millie Beetham who helped architect this. And to Henry Schuck for always pushing us to reimagine what's possible with AI + GTM data. DM me if you want the full podcast. It’s incredible. 

The Strategic Imperative: Build Your AI GTM Moat Before Competitors Do GTM teams slow to leverage AI's content generation and data synthesis capabilities will be systematically outmaneuvered by competitors in their market space that do. Is your competitors' use of AI keeping you up at night? Are they building unfair advantage: Sales reps armed with POV battle cards for discovery calls, Customer Success teams with real-time Customer Account health alerts highlighting likelihood to churn before the customer signals an issue, Marketing generating personalized campaigns highly curated to Target ICP and Personas, while your team debates single campaign messaging. They're not just working faster—they're playing a completely different game where they see opportunities, patterns, and solutions invisible to traditional approaches. Competitors outmaneuvering you aren't just using AI tools—they're combining AI's content and data capabilities with their proprietary customer data, industry insights, and process knowledge to increase the quality of Outreach motions, Discovery Calls, and Customer QBR's, creating defensible competitive advantages that cannot be replicated. They're not automating existing processes; they're inventing entirely new categories of delivering customer value to differentiate themselves from you in sales cycles. Your 90-Day Action Plan: Audit Data Assets: What unique customer insights, market intelligence, and operational data do you possess that competitors cannot access? This is your AI differentiation foundation. Implement Dual-Engine AI Strategy: Deploy content generation for scale (personalized outreach, health scores, curated proposals, real-time competitive positioning) AND data synthesis for intelligence (predictive qualification, account prioritization, churn prevention). Create AI-Native Customer Experiences: Design interactions that would be impossible without AI—real-time deal coaching, predictive customer success interventions, and dynamic pricing optimization. The Competitive Reality Check: Are you up at night, worried that your sales team is flying blind or spending valuable time trying to get to the data needed to be effective in sales cycles, while competitors have synthesized content enriched in real-time? Are your AE's and SDR's guessing at pain points while AI-powered competitors arrive armed with data-driven insights about each persona's specific challenges, decision-making patterns, and preferred communication styles? Are your Customer Success managers surprised by churn notifications while your competitors deliver dynamically generated QBRs that speak directly to usage health, value delivered, and new use cases that align with stakeholders' priorities? Modernize core GTM processes and motions with AI. Competitive advantage depends on how quickly you can combine AI's dual capabilities with existing documented processes, data-driven insights, and market position to create defensible differentiation.

Essential AI Tip researchers: Like many of my fellow market research and customer insights friends, I have been experimenting with AI for analyzing unstructured data (so far, mostly text from IDIs and survey OEs). Very interesting, stunningly fast, usually about 80% "correct" on thematic analysis, but wow you have to be precise. >>> ⏰ Key lesson for those just getting started--let me save you some heartache 💔 Always tell the AI (ChatGPT, Claude, whatever you use): "Using only the provided data, ....". For example: "Using only the provided data, identify six themes related to product X purchase deterrents, with three supporting quotes per theme." Otherwise it will "inform" its analysis by additional data sources. And asking for supporting quotes makes it easy for me to go and spot check that A) those quotes do exist and B) that I agree with how they were used to identify a theme. 🤖 🤖 🤖 #AI #marketresearch #cxresearch #consumerinsights #mrx

Just 30 days post launch, Sparks ✨ is transforming how Crayon customers compete, and now you can try it too! We launched Sparks a month ago 🚀 , and it’s been amazing to watch our customers harness the power of LLMs for competitive analysis 💪 . In just 30 days, we’ve had over 4,000 Sparks created from over 350,000 insights analyzed, and our customers have been creative and ingenious in applying this first-of-its-kind AI to their compete programs. Now some of you have seen too many “AI launches” and think Sparks is hype. But customers, prospects, and partners that have actually used it know it’s transformative. So we decided the best way to cut through the marketing BS was to simply let you try it! Starting today, you can head to crayon.co/sparks and sign up to try Sparks yourself. So how are customers using Sparks? Here are 5 interesting real-world applications from our customers: 1. Detect new competitive objections & arm sellers Data = [call clips, win-loss from interviews/CRM, seller-submitted intel] Prompt = “Based on recent Gong call clips, win-loss and intel submitted by our sellers, what are the top objections our sellers are getting on competitor x? For each objection, also explain how our best sellers are responding” 2. Identify emerging competitors Data = [news, press, social posts, forum threads, online reviews, call clips, win-loss, seller-submitted intel] Prompt = “Here is who we consider our Tier 1 competitors [insert list] and Tier 2 competitors [insert list]. Based on all the data provided, what other competitors should we be paying attention to and why?” 3. Update a Battlecard tile for the sales team Data = [call clips, win-loss, seller-submitted intel, online reviews] Prompt = “This Battlecard tile says “how we win against competitor x”: [insert tile]. Based on recent sales calls, win-loss, intel submitted by sellers and online reviews, how should we update this tile? For each recommendation, explain your reasoning and provide links to the specific data points that back up the recommendation.” 4. Do a hiring trends analysis for inclusion in a strategy deck Data = [job postings] Prompt = “Summarize the hiring patterns for this competitor. For each item, suggest/infer what this competitor’s company strategy might be related to this hiring pattern.” 5. Using Sparks to automate a weekly competitor digest email for the revenue team Data = [all data from the past week across 100+ insight categories] Prompt = “Summarize the trends, events or stories for these competitors based on all of this information. Emphasize things that would be most helpful for a sales team who is competing against these companies.” There are thousands of ways our customers are using Sparks, but hopefully these 5 examples give you an idea of what’s possible. If you want to try Sparks, head to crayon.co/sparks and let’s get you set up. This is compete’s renaissance and we want you to be part of it!