Gagan Biyani

Behind closed doors, the pressure to adopt AI is intense. Mostly, a CEO shouts “AI” every 20 minutes in an all-hands but provides no support, no definition of what good looks like, and no way to measure if it is working. The funny part is it's kind of working. Apply enough pressure and you will get results. But it’s slow, clunky, and stressing out your team. I keep coming back to the idea that there has to be a better way. A few things need to happen before AI adoption actually sticks. 1) Set a bar for what great looks like. If you don't have one internally, you need to borrow it from somewhere, whether that's stories on LinkedIn, blog posts, or teams that are a few steps ahead of you. 2) Change the system of work itself. Most companies underestimate this one. You need to connect data sources that have never talked to each other, redesign the workflows and job functions of entire teams, and write new policies that give people permission to experiment. 3) Invest in genuine upskilling. You cannot hand people a tool and expect them to figure it out on their own. Almost everyone just chats with it, and it takes months or years to go beyond chatting. This is hard to do, and why “AI Transformation/AI L&D” is one of the hottest new roles today. I think this transition takes a few years or more, and it is going to be messy and hard. Done well though, it can be a lot less painful than a CEO shouting "AI" in every meeting ;-) Maven is running a free speaker series on how to upskill your org on AI. We recruited 20+ practitioners who have done this effectively at scale and figured out incredible ways to build learning tools with AI. Here’s a sneak peek: - Mollie West Duffy, Head of L&D at Lattice - Ross Stevenson, Founder of Steal These Thoughts - Andrew Barry, Founder of Curious Lion, Former Program Dir at On Deck - Nick Shackleton-Jones, Author of How People Learn - Lavinia Mehedințu, Co-founder of Offbeat - Joe Leslie, AI L&D Leader at Golin Ketchum - Lori Niles-Hofmann, AI Transformation Strategist, ex-KPMG - Rishin Banker, COO at Maven + 16 more impressive leaders in AI transformation & L&D Check it out - it's free: bit.ly/ai-for-upskilling

I talk to so many people who find “self-promotion” icky but the truth is if you get over it, you’ll 2-3x your potential career upside. Here’s how to get over the fear: At Maven we work with world-class operators: PMs from OpenAI, designers from Airbnb, MLEs from GitHub. People who are genuinely the best in the world at what they do. And most of them avoid promoting a single thing they make. They think it is icky. Salesy. Beneath them. They worry their peers will quietly judge them. Meanwhile I scroll LinkedIn and watch people with far less experience build followings, land speaking gigs, get their dream roles, and shape the conversation in their industry.  If you're an expert who hates self-promotion, here’s a simple reframe: You're sharing what you know with people who need it. That's a generous act, not a vain one. Start small. Pick one platform and post twice a week about the real work you do. Be honest about what you have not figured out yet because the most credible voices online are the ones willing to admit uncertainty. You do not need to become an influencer, and you can be wildly successful without ever doing this. But it is just one of the easiest and most effective shortcuts available, once you build the mindset to get over the ick. PS Who is the most talented person you know who's terrible at talking about themselves? Tag them below 👇. We'd love to help them build a portfolio career on Maven.

150 practitioners have made over $100K on Maven (and 15 have made $1M+). Here’s how you build a successful portfolio career: I’ve watched thousands of people make hundreds or millions online (via Udemy and Maven) as instructors or creators. Here’s what separates those who succeed from those who fail: 1/ Persistence. Many instructors started small ($3k-5k/mo earnings) to test and learn. Once they learn the format, they often create a second or third course and skyrocket. Xinran Ma had only 5-8 students per cohort for 1 year, and when he launched his second course, he 10xed his earnings. 2/ Don’t wait. More than 50% of Maven’s top 10 instructors started without much of an audience or a super senior title. In fact, that made them relatable to the student and gave them the motivation to deliver an amazing experience. Colin Matthews started as an IC PM who simply loved teaching. He saw the AI wave and launched our first AI prototyping course and now is a top 10 earning instructor. 3/ Student-focused. It’s really hard to truly understand what your audience wants and set aside your own ego. Or more importantly, find the intersection between what you’re good at and what people want to learn. But you must. That’s the entire ball game. Sara Davison and Tyler Fisk make $1m+ on Maven. They are singularly obsessed with their students and have built a thriving community. 4/ Hustle. The #1 criteria we look for is whether someone is willing to hustle. These are high upside opportunities - $500K-$4M/year if you make it. To get that kind of upside, you gotta put some elbow grease in. If you do, the rewards are huge. Hamel Husain is one of those - the hustle and tenacity is unmatched. One of my favorite messages I get is when an instructor who never thought they would succeed says they bought a house or put their kid through college off of their earnings. Each time, they thank me for the opportunity. However, the reality is - they did it themselves. They worked hard, took ownership, and reaped the rewards.

How to build a team of A players: hire high potential people, and turn them into A players. For 99% of companies, this is far easier than hiring proven A players. Steve Jobs once said: “A players like to work with A players. B players hire C players.” This meme has made its rounds amongst anyone trying to build a high performing team. The problem is you're not Steve Jobs. For 99% of companies and especially startups, you cannot actually hire a bona fide A player. Those people are already successful, able to get jobs at the best companies, and aren’t going to take the risk on your untested idea. Instead, you have to bring on people who have high potential but haven’t yet proven it. It’s a certain type of person who is incredibly high agency and high hustle. For that high potential person, you need to: - Set a really high bar for excellence. - Cultivate cultural values that you reinforce regularly. - Provide “stretch” opportunities that are outside of their comfort zone. This bet doesn't always work. So then you need to part ways with the 30-50% of people who don’t fit in your system. Yes, this is hard to do, and the number does need to be ~30%, especially at startups, or you’re likely being too loose with your standards. Do that, and you can build a team of A players.

When a potential hire says they are “good at strategy” 🚩🚩🚩 Even an executive who is just good at “strategy” is simply just not that useful. The thing is - strategy and execution are a tightly connected feedback loop. You need to understand the details and be able to manipulate them to successfully make a strategy come to life. And along the way, you’ll get feedback from the execution that will help update and shape the strategy. So, my hiring advice is to stay away from strategy-only employees. And my advice for new hires is to remember that you are unlikely to be wildly successful if you are a thinker but not a doer. Both are required, and if you can do both, you’ll be invaluable to any great team.

The supply & demand curves you learn in college missed a key detail: post-purchase value perception. I studied Econ in college and think that there’s something deep that’s missing in most Econ theory. I’ll use Maven as an example, since I’ve spent 15 years in and around the education space so I know a lot about course pricing. Why do we tell Maven instructors to charge >$1000 for their courses? Why not $250 or $5000? This was a highly intentional decision with lots of iteration… Two reasons: 1) Maximizing revenue. Many instructors/creators get this wrong because they don’t realize that in-depth courses are inelastic goods. There’s barely 1-3 courses in the world that teach Problems-first Agentic AI Applications, and only one is taught by Aishwarya Naresh Reganti. If the product is good and there’s demand, then people are willing to pay. So you usually want to charge higher than you feel comfortable. 2) Maximizing value perception. Once you realize demand for your product is inelastic, you just charge to the moon! $5,000. $10,000. Why not? This is what scammy course creators do (not all creators). The problem is you destroy word of mouth. Value perception matters, and reviews reflect the mood of the buyer. Many course creators think the only way people find their course is when they do proactive marketing. In reality, 40%+ (the plurality) of users report they find courses through word of mouth. To price well, understand the demand curve AND the value curve. For in-depth, cohort-based courses taught by a bona fide expert (aka a Maven course), the ideal price is roughly $1,200-$2,500 depending on the topic, the course, and the target ICP. Maven courses tend to be for mid-career/senior professionals that make $100k+, so for a different ICP, you need to charge a different price. Note: some of you might be thinking: “Well, that isn’t good for me as the buyer.” It may be counter-intuitive, but semi-high prices are actually good for consumers too when the product is scarce. You WANT your instructors to make really good money. If they don’t, they won’t teach, and the education wouldn’t even exist in the first place. Their hourly rate is enormous and are weighing the opportunity cost of teaching you versus getting a consulting gig. The financial incentive motivates high quality instructors to put high effort into creating great learning experiences. The broader lesson applies to any product. You have to balance the maximum price you can charge today with the price at which a customer still believes the value is high. Too low and the word of mouth is through the roof, but you don’t make enough money. Too high and your customers reluctantly buy and it kills your growth engine.

L&D died after COVID. Now’s the time for a resurgence. It’s been a rollercoaster decade for Learning and Development in Enterprise. L&D grew steadily during the second tech boom (2008-2020), then had utter chaos during COVID. Recently, the function is suffering and L&D budgets have been slashed. I interviewed 20+ product and 20+ L&D leaders at top companies. Not a single product leader mentioned L&D as part of the solution to their AI upskilling problems. Consequently, most L&D leaders were struggling to find their place in the AI era. The Problem: I’m proud to have co-founded Udemy, and popularized video-based learning (“MOOCs”) that offered ubiquitous low-cost access to education. Hundreds of millions of people upskilled on technical skills via Udemy, Coursera, LinkedIn Learning in the 2008-2020 era. Everyone knows that completion rates for MOOCs are an abysmal 4-11%. But the bigger problem is utilization rate: most employees don’t even start courses in the first place. I’ve heard numbers from 10-20% at best. Multiply 4% completion rate x 15% utilization rate and you get 0.6%. Executives see low utilization and cut budgets, creating a vicious cycle. AI is a Major L&D Opportunity: The AI boom will be bigger than the tech boom of 2008-2020. The majority of knowledge workers are faced with a serious problem: their skills are getting outdated. Rapidly. It’s time for L&D to become the heroes of the AI era. After speaking with both CPOs and L&D leaders, I believe there’s a clear solution: 1/ We need L&D solutions that have high utilization AND engagement. Some employees may engage with videos, skills-based workshops or an AI chatbot. Others will want live learning in cohorts. L&D needs to provide multimodal learning solutions for diverse employees and focus heavily on utilization and engagement rates. 2/ Fresh sources of content. Instead of buying from the bigger players, L&D will need to find the niche companies that actually know AI natively, and have instructors who have their fingers on the pulse of what’s changing week-to-week within AI tools. 3/ Specific AND broad. In addition to providing AI 101 training, L&D needs to offer training that is specific to every role (for example: Claude Code for PMs or AI-Powered User Research). This is a huge opportunity for L&D, but the moment is not guaranteed. If L&D professionals and L&D companies can meet the moment, then once again learning and development can grow its position in the Enterprise. P.S. Maven just started 2 enterprise pilots with 90% utilization rates (!) and engagement rates of 80-90% (!!!). We are starting to open up our next round of pilot customers - if you’re interested in learning more, DM me.