Pranav Yadav

Thanks to the Professional Pricing Society for inviting me to participate on an expert panel (discussing survey-based pricing research methods) at #profitABLE25 in Las Vegas. I'm joined by Karthik R Balaji from Revenue Management Labs, and Fred Puech, PhD. Among other things, we discussed how survey questions that directly ask people how much they expect to pay (van Westendorp) or directly ask about whether they would buy a given product at a given price (e.g., Gabor-Granger) are often not as powerful or accurate as choice experiments (conjoint analysis). Moreover, with conjoint analysis you can disguise that you are after pricing information, because more than one attribute is varying along with varying prices. So, the respondent doesn't know whether it's mainly a brand, feature, or pricing study. This reduces the likelihood that they'll be able to game the system by low-balling you or stating a price that is socially desirable, etc. And, because you're varying more than one attribute in conjoint studies, you can simultaneously (in the same study) estimate price sensitivity as well as optimize product features across 1000s or millions of product variations. #sawtooth #pricing #conjoint #revenuemanagement

70

3 Comments

The old-school way of exploring #segmentation solutions is exhausting 😩 At Numerious Inc. we love running multiple algorithms to generate multiple 4-, 5-, 6-, 7-segment solutions for our clients. But let’s be honest—no client is going to wait for (or pay for) for us to personally evaluate every possible outcome. That’s why I loved this presentation on AI-Assisted Segmentation from Jackie Guthart and Curtis Frazier because it tackles the real pain point: 👉 Too many solutions, not enough time ⏳or brain space 🧠 In the past, we’ve relied on statistics to narrow down what to explore. But this approach? 💡 Uses an AI assistant to widdle down the multiple solutions, flagging which ones deserve deeper human review and which ones can be deprioritized. Real talk, after reviewing just two solutions, I struggle to remember what themes showed up where, and which segments overlapped! 📊 That’s where AI shines—it can scan through tables way faster than I can, compare consistency, spot patterns, and actually keep it all straight. Let’s be clear, this isn’t about replacing analysts. ✅ Pro: AI can drastically reduce the labor and cognitive overload of segmentation. Computers don’t forget which segments had which themes, where the overlaps are, or what changed across solutions. ❌ Con: AI still wants to please you—and might overemphasize patterns that aren’t really meaningful or actionable. Human validation is non-negotiable. Therefore, AI + Human = the win 🏅 Because the best solutions aren’t just smart, they’re strategically sound and client-ready #mrx Sawtooth #sawtoothconf

83

16 Comments

Data quality is an integral part of any research in order for it to be truly impactful. Recently, I signed the Data Quality Excellence Pledge, a statement of commitment to raising the data quality bar within the insights and analytics industry. It’s a set of practices and standards an organization commits to signal they are upholding the highest standards of data quality in their work. These commitments are to: - Uphold rigorous data quality standards throughout operations. - Provide transparency in these data quality standards. - Protect and elevate the experience and rights of research participants. - Enable education and collaboration among employees and partners. - Participate in and support GDQ initiatives. At Rare Patient Voice, we take data quality seriously. We follow a number of processes to ensure that our community members are authentic, starting with meeting patients and family caregivers at in person events. We prioritize gathering accurate and reliable information directly from them, ensuring that insights gathered reflect their genuine experiences and perspectives. This dedication not only enhances the integrity of research findings but also contributes to advancing understanding and improving outcomes for patients facing rare and non-rare diseases.

58

23 Comments

I am excited to introduce you to Drive Research technology, the latest and greatest in the market research industry. 🚀 It will catch bad players, eliminate bots from your survey data collection, and ensure you get pure and authentic responses for your panel survey that you and your insights team can believe in. Are you tired of not trusting your data or throwing out too many cases? 🚫 Here's a solution for you. It's called: Our humans. 🙋‍♀️ Unfortunately, nothing will ever be as good as having a human check your data file. And we have a fantastic team that catches it all. No advancements in tech, AI, or other space will fully replace a great research analyst reviewing responses for quality checks. Yet. Just a few of the things our researchers apply in our data quality process: 1️⃣ ReCaptcha integration to eliminate bot submissions 2️⃣ Articulation questions to review open-ended responses 3️⃣ Red herrings to identify those engaging genuinely 4️⃣ Duplicate IP addresses to eliminate duplicate responses 5️⃣ Straightlining checks to highlight repetitive answers 6️⃣ Time to complete to remove outliers of response times 7️⃣ Consistency in open-ended data to test contradictions As I've always said, ask for a data file if you work with our team. We're happy to share the work under the hood that helps our brands and clients arrive at an insightful and impactful report. 📊 No shortcuts behind the scenes. Unfortunately, so many firms skip these absolutely crucial steps that often go unnoticed by clients from the moment you sign off on your survey to the day you receive your final report. Your project might be less expensive, but at what cost? 💰 Shortcuts are usually driven by higher margins, speed, or shoving bad things under the rug to limit their time. And that's wrong. Because if that's the case, what are we all doing in our industry reporting on data we can't trust? 😔 You might have an excellent survey and a pretty report, but that falls apart if you don't have a market research partner you can trust when no one is watching. And not many brands are watching what's going on behind the scenes the way they should to hold their research partners accountable. -------------------------------------- Need MR advice? Message me. 📩 Visit Drive Research 💻  1200+ articles to help you. ✏️ --------------------------------------

71

1 Comment

I have published my AI predictions for 2025 at https://lnkd.in/ekA2jSs4 The post includes five predictions, and the first one is: 1) Amazon and Google will play a bigger role in AI in 2025. The launch of ChatGPT by OpenAI caught Google and Amazon off guard, and they had to move faster than they had planned. In terms of Amazon, watch out for Amazon Nova, which was launched in December 2024. Google plans to push Gemini hard, aiming to reach 500 million users (which would overtake ChatGPT); new products include Jules (a coding assistant), NotebookLM (a note-taking tool), and Project Astra (a universal AI assistant).

35

14 Comments