Persona Validation Processes

Earlier today, an AE less than a year in role told me this at the end of our first 45-min session. She said, "I wish I'd messaged you 6 months ago. I wouldn't be in this situation and probably would've made a lot more money." Let's turn the clocks back 2 weeks. She took on her first AE role less than a year ago. Reached out saying she wanted to level up her strategic outbound and run better discovery to uncover pain, not push product. So we focused our first session on Problem Enablement. Every seller I coach starts here. No exceptions. Doesn't matter if they have 2 months of experience, or 20 years. The results have been staggering. Here's what we did 👇 We shoved the product to the side, for the time being. And zeroed in on the problem. I came prepped with a detailed doc that mapped out: ➤ 1. Who are the PERSONAS → Who do you sell to → What does their day-to-day look like ➤ 2. What is the PROBLEM → key problem your offering solves → no buzzwords, product jargon or capabilities ➤ 3. How the PROBLEM affects each PERSONA → how is each person is affected by said problem → be specific, put yourself in the prospect shoes ➤ 4. How do they TACKLE the PROBLEM today → how do they usually deal with the problem → tools, workflows, or processes they might use ➤ 5. NEGATIVE IMPACT → impact of problem on the persona → negative consequence on them, team and org ➤ 6. METRICS / KPIs → metric or KPI persona is accountable for usually → how said metric is being affected by the problem ➤ 7. HOW YOU ARE UNIQUELY POSITIONED TO SOLVE → (now we sprinkle in a bit of product) → how is your offering uniquely positioned to solve ➤ 8. POSITIVE BUSINESS OUTCOMES → expected business outcomes they'll achieve → I call it the "so what" factor that I tell prospects We did this exercise for each problem her offering solves for. Went well beyond surface level. I told her this doc is the foundation. From it stems your strategic outbound, discovery, demo, and business case justification. And I'm confident the results will come, fast. Her words at the end, "So excited. I wish I’d messaged you 6 months ago. I wouldn’t be in this situation and probably would've made a lot more money. This is wonderful and exactly what I’ve needed. I’m super pumped.” We're just getting started. .................................................................. I believe this should be mandatory for every Sales onboarding. PROBLEM before Product.

I'll never forget the moment I realized we were doing persona research completely wrong. We were just presented this in a report: "Your ideal parents read Town and Country magazine." I stared at the report. Then asked the question that changed everything. "What are we supposed to do with that information?" Silence. We didn't have the budget for Town and Country ads. We didn't have the expertise to create them. We had zero ability to act on this insight. That's when it hit me. We'd spent months collecting data that looked impressive but was completely useless. Here's what most schools get wrong about persona development. They collect data for the sake of collecting data. They hire expensive firms. Download reports from fancy systems. Conduct elaborate surveys. Then they get insights they can't act on. The truth? You can build useful personas. You need to ask one simple question: Can we actually do something with this information? Here's the framework we developed instead: Start with your resources. What marketing channels do you control? Website, email, social media, events. Ask only questions that inform those channels. Where do they search online? What concerns keep them up at night? What objections do they have? Skip the impressive but useless stuff. Magazine subscriptions, luxury brand preferences, vacation habits. Focus on actionable insights. Keywords they search. Questions they ask. Problems you can solve. We rebuilt our persona process around this principle. Instead of "reads Town and Country," we identified "searches 'best private schools near me' at 11 PM on their phone." Instead of "drives a luxury SUV," we learned "worries their child isn't being challenged in public school." The difference? We could act on every single insight. We optimized our website for those keywords. We created content addressing those concerns. We crafted emails answering those questions. Within a year, our inquiries increased significantly. Not because we had better data. Because we had actionable data. Your persona research doesn't need to be exhaustive. It needs to be useful. Before you collect any information, ask yourself: What will I do differently once I know this? If you don't have a clear answer, don't collect it. Save yourself months of work. Focus only on insights you can act on with the resources you have. That's the difference between impressive research and effective marketing.

I used Claude to role-play what enterprise buyers would say about our positioning, then tested it with real buyers — They 𝗱𝗶𝘀𝗮𝗴𝗿𝗲𝗲𝗱, and that's the point. Here is the full setup and where things got interesting. --------------- Step 1: 𝗦𝗲𝘁𝘂𝗽 𝗖𝗹𝗮𝘂𝗱𝗲 𝗽𝗿𝗼𝗷𝗲𝗰𝘁 𝘄𝗶𝘁𝗵 𝗱𝗲𝗲𝗽 𝗯𝘂𝘆𝗲𝗿 𝗽𝗲𝗿𝘀𝗼𝗻𝗮 𝘄𝗼𝗿𝗸. Upload ICP and persona research files. Without this foundation, every synthetic run is shaky. Step 2: 𝗨𝘀𝗲 𝗖𝗹𝗮𝘂𝗱𝗲 𝘁𝗼 𝗯𝘂𝗶𝗹𝗱 𝗵𝘆𝗽𝗼𝘁𝗵𝗲𝘀𝗲𝘀, 𝗡𝗢𝗧 𝗰𝗼𝗻𝗰𝗹𝘂𝘀𝗶𝗼𝗻𝘀.  I run synthetic buyer panels in a Claude project based on those personas. Cheap, fast, good for sharpening what to test next. ✍ Tip: ask it to role play with you and gives quotes Step 3: 𝗧𝗲𝘀𝘁 𝘄𝗶𝘁𝗵 𝗿𝗲𝗮𝗹 𝗯𝘂𝘆𝗲𝗿𝘀 𝗼𝗻 𝗪𝘆𝗻𝘁𝗲𝗿.  Panels of real targeted buyers telling us what they actually think about each options, in their own words. (Big win getting Wynter approved. it's so fast!) Step 4: 𝗖𝗼𝗻𝗰𝗹𝘂𝗱𝗲 𝘄𝗶𝘁𝗵 𝗵𝘂𝗺𝗮𝗻 𝗷𝘂𝗱𝗴𝗲𝗺𝗲𝗻𝘁 𝗮𝗻𝗱 𝗮𝗰𝗰𝗼𝘂𝗻𝘁𝗮𝗯𝗶𝗹𝗶𝘁𝘆.  Compare AI survey results vs. Wynter's human one: where they align, where they disagree, and why. Cross-tab by persona, code and pull verbatim, bring in organizational context that AI doesn’t have, and write the recommendations myself. --------------- Two contrasting examples: 🚩 Synthetic personas flagged a seniority difference on a key term. influenced the test design and ran the real panel. Confirmed. That insight shaped how we 𝘵𝘪𝘦𝘳 𝘮𝘦𝘴𝘴𝘢𝘨𝘪𝘯𝘨 𝘣𝘺 𝘢𝘶𝘥𝘪𝘦𝘯𝘤𝘦. ❌ Another round, synthetic personas were confident a buzzy AI term would win. Real buyers killed it. That reversal saved us from launching with language the 𝘮𝘢𝘳𝘬𝘦𝘵 𝘸𝘰𝘶𝘭𝘥'𝘷𝘦 𝘳𝘦𝘫𝘦𝘤𝘵𝘦𝘥. --------------- Claude made me days faster. But faster only matters because the research design, interpretation, and judgment calls are mine as a human PMM. That's the collaboration I believe in: AI as a force multiplier for the human who knows what to ask and what to do with the output.

How Orthopedic Surgeons can employ the 'Six Thinking Hats,' develop patient personas, and attract more ideal patients. When FEED. The Agency is engaged to help orthopedic and spine surgeons grow their brand and avoid the money-pit of traditional marketing, we create ideal patient personas. Each hat represents a different perspective or type of thinking: White (information), Red (emotions), Black (cautious), Yellow (optimistic), Green (creative), and Blue (process). White Hat (Factual, Neutral Thinking) 1. What are the common characteristics (demographic, psychographic, and clinical) of our current patient base, and how can these data inform the creation of detailed patient personas? 2. What trends in orthopedic conditions are emerging from recent healthcare data, and how might these impact the types of patients seeking treatment? Red Hat (Emotions, Feelings, and Intuition) 3. How do our ideal patients feel about their orthopedic health, and what emotional journeys do they undergo from recognizing a problem to seeking treatment? 4. What fears or hesitations might patients have about orthopedic surgery, and how can we address these in our patient personas to ensure they feel understood and supported? Black Hat (Critical, Cautious Thinking) 5. What are the potential barriers or challenges in reaching our ideal patients, and how can we anticipate and mitigate these in our persona development? 6. Considering the diversity of conditions treated by orthopedic surgery, how can we ensure our patient personas do not oversimplify or ignore key patient segments? Yellow Hat (Optimistic, Positive Thinking) 7. What opportunities can we identify for engaging with ideal patients more effectively through targeted messaging and services, based on our patient personas? 8. How can developing detailed patient personas improve patient outcomes and satisfaction, leading to a stronger practice reputation and patient referrals? Green Hat (Creative, Lateral Thinking) 9. In what innovative ways can we gather insights and feedback from current and prospective patients to refine and validate our patient personas? 10. How can we use patient personas to personalize and enhance the patient experience, from the initial consultation through to recovery and follow-up care? Blue Hat (Process, Control) 11. What process will we follow to create, review, and update our patient personas to ensure they remain accurate and relevant over time? 12. How will we measure the effectiveness of using patient personas in our marketing and patient engagement strategies to attract more ideal patients? These questions encourage a multi-dimensional approach to developing patient personas, ensuring a thorough analysis that encompasses data, emotions, potential challenges, opportunities for positive engagement, creative strategies for persona utilization, and a structured process for implementation and evaluation.

I co-authored the two Persona Lifecycle books with John Pruitt. They are really all about looking holistically at persona efforts through our somewhat-cute analogy of a human lifecycle: 🌡 Family planning, where you figure out why you need personas, whether your org will benefit from them, which obstacles you need to overcome, and most importantly (IMO) the measurable goals for your persona effort (most people skip this, which is why so many people end up hating personas -- if you don't identify the real problem you are trying to solve, and you end up using personas and they are the wrong tool, then of course you end up thinking personas are useless.) 😍 Conception & Gestation, where you create the personas. Our books focus on creating personas from data. I NEVER start this way anymore and haven't since 2005. I start by creating alignment personas to identify and name the assumptions of the key stakeholders. THEN AND ONLY THEN collected data to validate or invalidate. Otherwise the assumptions of powerful people in your org will always show up again and clobber your data-driven personas. Which is another reason people say 'personas don't work.' 👶 Birth and maturation, in which you do smart things to introduce your personas into your org and set them up to do their jobs 🤺 Adulthood, in which personas DO their job--which requires that you plan for how and when to bring personas into your design and dev process with actual tools that we actually show you in the book 👵 Retirement and lifetime achievement, where you go back to your measurable goals for the persona effort and MEASURE. And then plan for what's next. The book I'd write today (and have sort of half-written) would be about Alignment Personas--what they are and how to create them. It's a method I've developed over almost 20 years of consulting in the real world with real organization that have real politics and real power dynamics. It's a series of five conversations that helps (aka forces) key people to: 1. Create and agree on measurable, prioritized goals, usually around usage metrics for a product (Increase or decrease <important metric> by <actual amount with a number> within <defined time period> 2. List all the ways they currently talk about users and customers 3. Group these users and customers based on statements that start with 'i want' or 'i need' 4. Create persona candidates 5. Prioritize the candidates based on the goals (our goal is X--so if we don't make <persona> ridiculously happy, what the heck are we actually doing?) THEN you can gather data and it has a place to land. It has a chance of changing minds. Or, you can take the new alignment you've created and make some really great decisions for your project, because alignment around SOMETHING is often more helpful than data about the 'right' thing. Given that I can't get myself to write the book, would any of you be interested in a video series? I'm creating it now. I'd love feedback. You in? #personas #ux

I often felt that user research and creating personas were a guessing game and a waste of time. I was wrong. Here is how to ensure the research brings great results: It can indeed feel like a pointless exercise when you're doing research just to check a box, or when your personas end up being a slide nobody ever opens again. The truth is, only good research drives good decisions. So, why isn't it always good? 1) You interview too few people, or only those easy to reach Talking to just five people from your internal network or friends of friends rarely gives you a full picture. If you don't capture a range of motivations and use cases, you're likely building for a narrow and biased segment. 2) You ask leading questions When people sense what you want to hear, they try to be nice. This results in empty validation that hides the real frictions they face. 3) You stop at surface-level insights If the notes are a collection of generic statements like "I want it to be easy to use," you’re not learning anything actionable. Real insights come from digging into stories, context, and behavior. 4) Your findings aren't actionable Insights without a direct impact on what you're building tend to fade into the background. If you can't point to how research shaped a feature or decision, it's just noise. 𝗪𝗵𝗮𝘁 𝗺𝗮𝗸𝗲𝘀 𝗶𝘁 𝗿𝗲𝗹𝗶𝗮𝗯𝗹𝗲 • Focus on behavior, not opinion: Asking people to describe what they did in a specific situation reveals more truth than asking them what they want. • Pattern recognition for the win: It’s tempting to anchor on one powerful quote, but decisions based on isolated comments are dangerous. The goal is to spot repeated patterns across interviews and use those to inform the product direction. • Co-create personas with your team: This way, they use them, not ignore them. Personas made in isolation often fail because they don’t feel real or relevant. Involving designers, engineers, and even sales in creating personas helps ensure they are grounded in actual experience and get referenced often.    𝗧𝗼𝗼𝗹𝘀 𝘁𝗵𝗮𝘁 𝗰𝗮𝗻 𝗵𝗲𝗹𝗽 • Maze makes it easy to run user tests without scheduling interviews. It’s great for testing flows, copy, and concepts with actual users at scale. • WhiteBridge.ai helps you to identify similar people or talk to completely fresh prospects. • Dovetail allows you to tag and synthesize interview data efficiently. You can quickly identify themes and build a research repository that your team can access anytime.    Remember, if you can't do it right, you shouldn't do it at all. There are other ways to make the best product bets possible. Do you trust in your user research? Sound off in the comments! #productmanagement #productmanager #userresearch P.S. To become a Product Manager who can perform good research, be sure to check out my courses on www. drbartpm. com :)

You don’t need $100K to define your ICP—here’s how to do it smarter At OneLogin, we spent $100K on ICP research—quantitative data analysis, stakeholder, customer, and competitor interviews. The result? A 100-page report that largely confirmed what we already knew: ✅ Mid-market segment ✅ Technology-first industries (FinTech, EduTech, Telcos) + manufacturing The real insight? Buyers wanted more value at a competitive price. Competing with Okta and Microsoft in IAM meant we had to be strategic—positioning OneLogin as the #1 Value Leader was the right move. Good news: You don’t need $100K to do this today. AI can help. 🚀 AI-powered market analysis, TAM, TRM, industry and persona research Google Gemini Pro + Deep Research → Industry analysis in minutes (I use it every time when onboarding a new client). Perplexity, ChatGPT → Persona development, interview guides, messaging validation. Next: Analyze and segment your CRM data. Once you’ve identified your target segments and personas, you need to validate them against your actual customer data. This is where most companies struggle—their CRM is a mess. ❌ No data normalization ❌ Inconsistent industry tagging ❌ Missing roles, functions, firmographics At some point, you have two choices: 1️⃣ Clean up and enrich your CRM (tools like ZoomInfo and LI Sales Navigator can help) 2️⃣ Scrap it and rebuild from scratch If your data is somewhat usable, run segmentation analysis to uncover: Churn & lost opportunity reasons New and expansion revenue by segment Segments with the highest and lowest utilization ICP mistakes companies make (and how to fix them) ❌ Don't assume your current and future ICPs are the same. ✅ Build a moat—Your best customers today may not be your best customers tomorrow. ❌ Focus only on firmographics. ✅ Look at technographics—Are prospects using complementary tech? Should you block accounts where certain tech stacks lead to lost deals? ❌ They define personas based on assumptions. ✅ Validate with real customers. Listen for the exact language buyers use to describe pain points. 🔹 Pro tip: Use ChatGPT to create an interview guide, summarize call recordings, and extract insights to refine your ICP. How do you operationalize ICP? Knowing your ICP helps marketing with broad targeting. BDRs use it for qualification. But sellers need a subset—target accounts. 🔥 Jon Miller’s FIRE framework helps prioritize accounts based on: Fit (firmographics, technographics, industry, geo) Intent (buying signals) Relationships (leverage execs, board, investors for warm intros) Engagement (activity with your brand) Final Thought: Your ICP isn’t static—it should evolve with market shifts, competitor moves, new products, and GTM strategies. If you serve multiple industries or have a multi-product offering, you need multiple ICPs. If you’re not revisiting it, you’re flying blind. 📌 More on ICP process? Check out GTM Partners’ resource—link in comments. #ICP #B2BMarketing #AI #BoldGTM

Ever been handed a vague project like "We need better personas" and a crazy deadline? A simple framework can turn that chaos into clear action: The key? Start with the END GOAL in mind and work backwards. This is because only when you’re clear on the outcome can you create a process that’s realistic, effective, and aligned with business goals. Let’s break it down with the example: "We need better personas." 🎯 Step 1: Define the end goal Ask: Why do we need better personas? What’s the real business metric we’re trying to move? Example: Increase win rates by 9% over the next 6 months. In this case, it’s clear the project isn’t just about creating personas, it’s about using those personas to sharpen messaging and drive more sales. 🎯 Step 2: Align stakeholders & set milestones Before jumping into deliverables, align with key stakeholders. Ensure everyone agrees on the goals, timelines, and success metrics. Kickoff meeting: Confirm the end goal, scope, and key deliverables. Milestone check-ins: Schedule  updates to ensure alignment and course-correct if needed. 🎯 Step 3: Get specific on deliverables If the focus is on increasing win rates, what’s needed beyond just personas? - > Persona profiles: Core buyer personas, pain points, triggers, buying journey maps, and content preferences. - > Messaging guide: Value propositions, key messaging themes with proof points, objection handling, and specific talking points. - > Sales enablement toolkit: Persona-specific pitch decks, talk tracks, one-pagers, FAQs, and objection-handling guides. 🎯 Step 4: Gather data Given the timeline and goals, what’s realistic for research? Examples could be: - > Deploy a customer survey to 200 customers to refine and segment personas. - > Analyze 10 closed sales deals within ICP. - > Conduct 5 in-depth customer interviews for qualitative insights. 🎯 Step 5: Build, test, and iterate Once stakeholders agree on the research plan and deliverables, start building and validating. - > Develop personas and associated messaging. - > A/B test messaging to validate impact (e.g. using emails) -> Collect sales team feedback on persona usability and messaging effectiveness. Key takeaway: Working backwards forces clarity and also makes it easier for you to counter unrealistic times.  I have been working through this process with dozens of clients to help them get more clarity. I’d love to hear from you! How do you approach vague project requests? #productmarketing #coaching #GTM #productivity #career

This weekend, over dinner, a friend told me their team had started using #syntheticusers in the very early stages of innovation and ideation to explore ideas quickly and discard the less promising ones. We agreed there can be value if the scope is well defined. But user research needs real users. Without that, AI-generated data, however convincing it may sound, risks influencing decisions it is neither prepared nor reliable enough to inform. Based on my experiences, on the literature review, and on the self-reported best practices from some vendors (enriched by recent academic findings) here is a checklist for the responsible use of synthetic users: 1. State the purpose: Define clearly what the simulation can and cannot answer. Treat it as a hypothesis-generation tool, not a truth-discovery mechanism. 2. Transparency first: Record prompts, datasets, model versions, and parameters so results can be traced, questioned, and replicated. 3. Measure variance: Check how varied the outputs are. If results are unusually consistent, treat them as a weak signal. Harvard’s Häusler et al. warn that over-consistency can mask missing perspectives. 4. Diversity by design: Vary profiles, contexts, and prompts systematically. Go beyond demographic variety. Make a narrative-aware evaluation to detect stereotyping, erasure, or reductive portrayals. 5. Always validate with humans: No product or service decision should be made without proportional real-world input. Develop community-centred validation, especially when personas represent vulnerable or under-represented groups. 6. Ethics and impact: Map who could be harmed if the simulation is wrong. Audit synthetic personas for bias, exoticism, and omission before using them in decision-making. When used with a tightly defined scope and robust validation, synthetic users may help speed up early-stage prototyping. Outside those boundaries, their outputs risk being misleading and should not be treated as substitutes for human insight. What about your experience with this topic? Where do you see them fitting responsibly into your process? _________ Sources (links on the comments): a) Nielsen Norman Group b) Bain & Company c) NielsenIQ d) Venkit et al., Harvard, 2025 e) Häusler et al., 2025