Using Data to Improve Training Programs

𝐖𝐞 𝐚𝐥𝐦𝐨𝐬𝐭 𝐥𝐨𝐬𝐭 𝐚 ₹6,00,000 𝐜𝐨𝐧𝐭𝐫𝐚𝐜𝐭 𝐛𝐞𝐜𝐚𝐮𝐬𝐞 𝐈 𝐫𝐞𝐟𝐮𝐬𝐞𝐝 𝐭𝐨 𝐫𝐞𝐦𝐨𝐯𝐞 𝐨𝐧𝐞 𝐬𝐥𝐢𝐝𝐞. The CEO loved the program. Her leadership team loved the design. Budget was approved. Then the CHRO said: “Can you just skip the pre and post assessments? The team finds evaluations stressful.” I said no. The room went quiet. Here’s what I knew that she didn’t want to hear, without assessment data, this program is just an expensive event. There’s no before, No after, No proof that anything changed. The Reality: 70% of training programs fail to show measurable behavior change, not because the content was bad, but because nobody measured anything to begin with. You can’t manage what you don’t measure. And you can’t defend your training budget in the next board meeting without numbers. They pushed back twice. I held the line twice. We got the contract. With the assessments intact. Here’s what that moment taught me: 1. Your non-negotiables are your credibility: The moment you dilute your methodology to close a deal, you’ve told the client your standards are negotiable. 2. Clients don’t always know what they need: Your job is to protect the outcome, not just the relationship. 3. The right clients respect the pushback: If they don’t, they were never the right fit. Protect your process. It’s the only thing that guarantees your results. 𝐖𝐨𝐮𝐥𝐝 𝐲𝐨𝐮 𝐡𝐚𝐯𝐞 𝐫𝐞𝐦𝐨𝐯𝐞𝐝 𝐭𝐡𝐞 𝐚𝐬𝐬𝐞𝐬𝐬𝐦𝐞𝐧𝐭𝐬 𝐨𝐞 𝐡𝐞𝐥𝐝 𝐭𝐡𝐞 𝐥𝐢𝐧𝐞? #training #leadership

Dogs and cats can't talk...yet. Diagnostics give them a voice. In our latest Shared Table discussion, I sat down with two of the most transformative leaders in our industry: Jay Mazelsky, CEO of IDEXX, and Jay Price, CEO of Mission Pet Health. From diagnostics and disease protection that enable the front line veterinary care, this dynamic sets the table for a great end-to-end conversation. We are entering an interesting window for Animal Health. Many adopted pets during the pandemic, as much as 4x higher than in the average year. Now, those pets are about 5-7 years old. We are entering a window where demand for senior pet care is likely to build significantly. The data tells a compelling story: - Only 1 in 5 clinical vet visits include bloodwork - Only 12% of wellness visits include diagnostics The opportunity is bigger than we can measure. As diagnostics expand and AI accelerates what we can learn, the market opportunity will expand. From cancer to diabetes to CKD and beyond - helping pets live longer, healthier, more active lives is a growing opportunity. The leaders who focus on delivering value - to the pet, pet owner, and veterinarian - will be the ones who grow with the industry and beyond. Watch the full Shared Table conversation here: https://lnkd.in/gmm_Fa8b #AnimalHealth #SharedTable

It's easy to get caught up in large-scale data and quantitative metrics. But to truly understand the experience of managers, we've found that you have to actively listen and hear their stories. At Google, we created the "Google Learning Advisors" program to do just that. A volunteer group of managers, recruited to be representative of the global Google population, engages in quick creative research activities every few weeks designed to uncover their lived experiences and challenges. The insights we learn from this research let us design programs that truly meet these managers where they are. A few of our key takeaways: 1️⃣Managers want to feel "seen." Our advisors highlighted the importance of engaging with them d in a way that conveys trust and partnership. This insight now influences everything from the design of manager communications to the names of our learning programs. 2️⃣Real life is ambiguous. While traditional training scenarios often have clear-cut answers, managers told us they need more help with the nuances and ambiguities of their roles.  We test our practice scenarios with advisors to ensure they reflect the real-world challenges managers face. 3️⃣Motivation is not one-size-fits-all. We learned that managers arrive in their roles for a variety of reasons, which impacts what they find motivating and challenging. This has helped us create learning products that incorporate a breadth of perspectives and approaches. Qualitative research like this acts as a compass, allowing us to co-create learning experiences  with managers  that meet their needs. Learn more about our process here: https://lnkd.in/eb6Zu8V9 #leadership #learninganddevelopment #TheGoogleSchoolForLeaders

Can we help you train for your next marathon? Maybe we've built something to get you started. Terra API, we analysed six months of real training data from 101 recreational to sub-elite marathon runners and created a predictive model that explains 79.3% of the variance in actual finish times (R² = 0.7933), with an average prediction error of about 18 minutes. I must admit, the accuracy is a long way off being useful. I think I could be closer to most people's time, without a fancy model! But we decided to publish it anyway to demonstrate the inherent difficulties of modelling human performance. The model uses five key inputs you can plug in yourself: •⁠ ⁠Your baseline running pace  •⁠ ⁠Total training volume over 6 months •⁠ ⁠Intensity distribution  •⁠ ⁠Training frequency It incorporates non-linear effects and interactions we observed in the data, such as: •⁠ ⁠Diminishing returns on extra volume (gains are bigger when you're at lower totals) •⁠ ⁠Accelerating benefits from easy miles — the higher the % of easy training, the bigger the payoff •⁠ ⁠High-intensity work best kept under ~20% of total time (beyond that, it often hurts more than helps) •⁠ ⁠Faster natural runners get disproportionately larger gains from volume and easy work This isn't magic or a guarantee; marathon performance is messy and multifactorial. The model misses race-day chaos (weather, nutrition, psychology, course quirks), incomplete GPS logs (not every session gets recorded), individual genetic/response differences (20–50% non-responders in some studies), and more. It's correlational, based on a modest sample, and probabilistic at best. But the patterns align with broader running science: easy-heavy pyramidal distributions dominate among faster runners, volume matters hugely, but plateaus, and personalisation beats one-size-fits-all. That's why we're sharing an interactive version for you to experiment with: input your own numbers, adjust sliders for different scenarios, see probabilistic predictions, and use it as a thought experiment for your training. It's a rough prototype, insightful for sparking ideas, but treat it lightly. Far more advanced versions (larger datasets, better handling of missing data, race-day factors, ML personalisation) are in development. Ready to play? Link in the comments below. Whether you're aiming for sub-4, sub-3, or just to finish strong, what's the one training change you're considering right now? Drop it in the comments; let's discuss! #MarathonTraining #RunningScience #EnduranceSports #DataDrivenFitness #PersonalisedTraining

To be a great sales manager, you have to be a great coach. But coaching often slips through the cracks — especially when there are big deals to close. Now, AI is making it possible for every manager to not just coach more, but coach better. When I was in sales, I saw many new managers fall into the same trap: putting on their superhero cape to rescue deals instead of coaching their reps through them. I was guilty, too! We all knew coaching was important — but we had no time and no tools to scale. With AI, sales managers can now deliver highly targeted coaching at scale. It’s now possible to analyze multiple call transcripts in minutes and pull in unstructured data to understand what happens between calls. You can: 1. Review each rep’s recent calls, emails, and conversations to get a complete picture of how they’re selling — and give them targeted recommendations for improvement. 2. Analyze calls from a specific segment and compare what’s working in closed-won versus closed-lost deals to pinpoint the messaging and strategies that perform best. 3. Generate summaries of how top performers handle objections, communicate ROI, and build a business case — and share those insights with new reps as they ramp. Many HubSpot customers (and our own sales managers) are already using these insights to send regular, personalized coaching to reps — and improve the productivity of their teams. Being a sales manager used to feel like you’re a “super rep” — jumping between calls, rescuing deals, trying to fit in some coaching along the way. Now, it feels like you’re a “super coach” — spotting trends, sharing insights, and helping your whole team scale their impact. Exciting times!

We are excited to share a research paper co-authored by WFP India's Paramjyoti Chattopadhyay, Head of the Research, Assessment, Monitoring, and Evaluation Unit, and Vijay Avinandan, Monitoring and Evaluation Officer, published in the Asia Pacific Journal of Evaluation on how evaluations can drive inclusive policymaking. Some the questions the paper, looks at are: 1. Can instilling the dimension of equity and inclusivity in evaluative choices, approaches, and frameworks, nudge social security schemes and policymakers to be instinctively responsive to the needs of vulnerable populations? 2. Can evaluations nudge policy makers to pay more attention to equity and inclusion considerations? Read https://lnkd.in/gGXtrRRY #nudge-effect #evaluation #systems #equity #evaluations #inclusion

🚨 Public Private Data Partnerships? It’s Time to Scale 👉 Today’s Development Data Partnership Day 2025 was more than a showcase—it was a call to action. Hosted at Google’s NYC office, the event gathered multilateral institutions, tech companies, and researchers to explore how third-party data can drive public good. 👉 But beneath the many success stories, one theme kept surfacing: we’re still stuck in pilots. To truly harness data for development, we need systems, not one-offs. My 10 takeaways from today: 1️⃣ Go beyond pilots—institutionalize workflows so data collaboration becomes routine, not exceptional. 2️⃣ Match demand and supply – Align real-world policy needs with available datasets, proactively. 3️⃣ Think national and local – Focus on building data ecosystems in countries and cities, not just global HQs. 4️⃣ Standardize and streamline – Develop templates and interoperability frameworks to lower transaction costs. 5️⃣ Embed the workflow – Make data access for re-use a structured, repeatable process—not ad hoc heroics. 6️⃣ Support data stewards – Invest in the people who can bridge domain specific, policy, technology and data communities. 7️⃣ Align incentives – Collaboration shouldn’t rely on goodwill alone; it needs to be in everyone’s interest (and recognize each others interdependencies). 8️⃣ Prepare data for AI – Data must be representative, high-quality, and responsibly governed to fuel ethical and public interest AI. 9️⃣ Earn a social license – Legal agreements aren’t enough—public legitimacy matters. 🔟 Close the data divide – Data access and infrastructure are core to addressing global inequality. 🗂️ Featured datasets included Ookla (connectivity), LinkedIn (energy/digital jobs), Planet (Remote sensing), Google (building data), ESRI (AmazoniaForever360), JBA Consulting (flood prevention), and more. 🌍 Other initiatives like Disha (UN Global Pulse: https://lnkd.in/evteQ3-t) and ImpactAI (World Bank: https://lnkd.in/eSSQNNwJ) were also mentioned to show what’s possible when data is made accessible for re-use. 👉 Bottom line: Bridging the data divide must be central to development strategies. Doing so requires long-term investment in institutions, infrastructure, positions and governance—not just one-off solutions. #DataForDevelopment #AIReadiness #SocialLicense #DataStewardship #PublicInterestTech #DevelopmentData #DataCommons #GovLab #DevelopmentDataPartnership

𝟑 𝐦𝐢𝐬𝐭𝐚𝐤𝐞𝐬 𝐈 𝐦𝐚𝐝𝐞 𝐚𝐬 𝐚 𝐒𝐩𝐨𝐫𝐭 𝐒𝐜𝐢𝐞𝐧𝐭𝐢𝐬𝐭 𝐰𝐢𝐭𝐡 𝐃𝐚𝐭𝐚 📊 Reflecting on my journey as a sport scientist, from my time at QPR FC to managing the London Gaelic football senior team and working as a Fitness Coach with Kerala Blasters FC, I’ve learned a lot from my mistakes. Here are three big ones that stand out: [This pic brings back great memories of working with a bunch of players from all over the world in the ISL in 2015, all over India. Here I am with the great Mohammed Rafi...who happened to have one of the best 'leap & hang' jumps I've seen...and a truly humble guy also] 1️⃣ 𝐎𝐯𝐞𝐫𝐥𝐨𝐚𝐝𝐢𝐧𝐠 𝐒𝐭𝐚𝐤𝐞𝐡𝐨𝐥𝐝𝐞𝐫𝐬 𝐰𝐢𝐭𝐡 𝐃𝐚𝐭𝐚 While working as a Sport Scientist at 𝐐𝐏𝐑 𝐅𝐂 here in London, I often presented too many metrics to coaches and players. The result? Confusion and disengagement. 𝐋𝐞𝐬𝐬𝐨𝐧 𝐋𝐞𝐚𝐫𝐧𝐞𝐝: Focus on the key insights that matter most to your audience. Less is more when it comes to impactful reporting. 2️⃣ 𝐍𝐨𝐭 𝐏𝐥𝐚𝐧𝐧𝐢𝐧𝐠 𝐖𝐨𝐫𝐤𝐟𝐥𝐨𝐰𝐬 𝐟𝐨𝐫 𝐄𝐟𝐟𝐢𝐜𝐢𝐞𝐧𝐜𝐲 When I was Manager of the 𝐋𝐨𝐧𝐝𝐨𝐧 𝐒𝐞𝐧𝐢𝐨𝐫 𝐆𝐚𝐞𝐥𝐢𝐜 𝐟𝐨𝐨𝐭𝐛𝐚𝐥𝐥 𝐭𝐞𝐚𝐦, I relied heavily on manual processes in Excel to track training loads and wellness data. This led to wasted time and errors. 𝐋𝐞𝐬𝐬𝐨𝐧 𝐋𝐞𝐚𝐫𝐧𝐞𝐝: Automate wherever possible. Tools like Power BI could have saved me hours and improved accuracy. (Check out all our CPD courses and professional services over on the Sport Horizon UK website) 3️⃣ 𝐍𝐨𝐭 𝐒𝐭𝐚𝐧𝐝𝐢𝐧𝐠 𝐛𝐲 𝐌𝐲 𝐂𝐨𝐧𝐯𝐢𝐜𝐭𝐢𝐨𝐧𝐬 𝐨𝐧 𝐃𝐚𝐭𝐚 𝐍𝐞𝐞𝐝𝐬 When I was working as a Fitness Coach with 𝐊𝐞𝐫𝐚𝐥𝐚 𝐁𝐥𝐚𝐬𝐭𝐞𝐫𝐬 𝐅𝐂 𝐢𝐧 𝐭𝐡𝐞 𝐈𝐧𝐝𝐢𝐚𝐧 𝐒𝐮𝐩𝐞𝐫 𝐋𝐞𝐚𝐠𝐮𝐞 (𝐈��𝐋), I gave in to pressure to reduce the level of data we were gathering. This meant we lacked critical performance data later in the season, which could have been used to tailor training programmes and identify players at risk of overtraining or undertraining. 𝐋𝐞𝐬𝐬𝐨𝐧 𝐋𝐞𝐚𝐫𝐧𝐞𝐝: Be clear and confident about the long-term importance of data collection. It’s better to gather the right data upfront than regret not having it later. Mistakes like these are opportunities to grow. Looking back, these lessons have helped me improve how I communicate, process, and use data effectively in sport. 💬 What are the biggest lessons you’ve learned working with data? #SportHorizon #SportScience #PerformanceAnalysis #DataVisualisation #DataAnalytics #PowerBI #Tableau #Scouting #Football #Soccer #GAA #GaelicFootball #BespokeInsights

Maximizing ROI on invested time for health Geoff Yang (GY): Dustin Nabhan, people investing time in their health goals but not always in the right places. When you work with elite athletes, how do you maximize their ROI? Dustin Nabhan (DN): It starts with quantified goals & rigorous measurement. In professional sports, we don't guess. We assess relevant systems and performance inputs: strength, power, nutrition, recovery, body composition, etc. Then we allocate time and resources to the areas with the biggest gaps. The same logic applies to anyone serious about performing at the highest level. GY: Most people aren't getting that kind of assessment? DN: Right, and that's the issue. You need to set goals, assess where you are, build a plan, and measure your progress. Without that, you're guessing. You might spend 5 hours a week on cardio, but if your aerobic fitness is already in the 85th percentile for your age/gender while your muscle fitness is in the 40th, you're overinvesting in a strength and underinvesting in a weakness. That imbalance may show up as injury, lower energy, lower performance, or accelerated aging in the systems they've been neglecting. GY: That's essentially the idea behind our Healthspan Domains™ model. DN: Instead of treating "health" as one thing, we break it into eight measurable domains: aerobic fitness, muscle fitness, body composition, bone, balance, movement quality, cognitive health, and blood biomarkers. Each domain is scored on a percentile basis for your age and gender. So we’re not comparing a 25-year-old female triathlete to a 55-year old male C-Suite executive. GY: Why does that matter? DN: We’ve seen conceptual curves showing healthspan vs longevity. But the question is: where are you on that curve? How do you go from a subjective assessment, like "I'm in pretty good shape," to something predictive of how you’ll perform and how you’ll age?  When you see you’re in the 83rd percentile for bone density but the 41st for body composition, the conversation shifts immediately. You focus on "how do I move this specific number?" That's a much more productive mindset. GY: So how does this change a time-strapped executive's approach? DN: It becomes a resource allocation problem, which is something executives understand. If you only have 3 hours, invest it where you get the highest ROI for your goals and health. That might be changes in training, nutrition, or sleep.  The domain scores act as a filter. They tell you: here are the 1 or 2 areas where investment will generate the highest return. The Apeiron Life team then builds protocols around those gaps — specific, measurable, time-efficient. Then add optimal frequency and sequencing and multiply it all using technology, supplements, biohacks. GY: Focus moves outcomes. DN: Exactly. In professional sports, we set the goal, measure what matters, focus effort where it counts, and let the data do the prioritizing.  That's how you get the most out of limited time.

The headline that caught my eye this week was "People Are Uploading Their Medical Records to A.I. Chatbots." Here's my take: I recently fed my medical records, lab results, some exercise testing I had done, and lots of wearable device data into an AI engine to optimize my exercise and health routine. The analysis was remarkably sophisticated, identifying patterns between my heart rate variability, vo2 max, sleep quality, and workout intensity that helped me adjust my daily schedule — including getting rid of situps as part of the morning routine I’ve been doing for decades. Millions are doing something similar, turning to AI for health insights with a fascinating mix of desperation and pragmatism.    The New York Times reports that while some receive dangerously wrong diagnoses (a woman's suspected pituitary tumor that wasn't there), others discover life-threatening blockages their doctors initially dismissed as manageable. The privacy trade-offs are stark. HIPAA doesn't apply to ChatGPT. Your medical data could theoretically leak into future training sets. Yet Robert Gebhardt, 88, speaks for many when he shrugs: "My cellphone is following me wherever I go. Anybody that wants to know anything about me can find out, including my medical data. It’s a fact of life, and I’ve reconciled myself to that.” A big question in all of this is the liability that the AI companies may ultimately bear either for the cybersecurity issues around the protection of health data or for the consequences to people who follow the advice given. (Don't worry: I won't ever sue over replacing sit-ups with more squats and plank time.)    In any case, my own experience underscores that we are entering a new world. We're witnessing the emergence of a parallel health consultation system, unregulated but responsive, risky but accessible, where it's not yet fully clear whether accuracy matches availability. https://lnkd.in/esxUWZvR