About
I build products and businesses in markets that are hard to reach and harder to scale…
Articles by Chandler
Activity
4K followers
-
Chandler K. posted thisEveryone arguing about whether agents should pay for seats is solving the wrong problem. Seat pricing made sense when the thing consuming software was a human being with one login and limited hours in the day. The seat was a proxy for consumption, and as proxies go, it was reasonable. Agents don't map to that model. One agent can run around the clock, process thousands of records, operate across multiple systems at once. You can't count them like employees and call it a pricing strategy. Dan Shipper made a prediction worth repeating: we'll build software for humans and agents together, not just software that humans use and agents happen to tolerate. If that's right, the pricing model has to change too, because it was designed for a different kind of user. The companies that figure this out first won't be debating seat counts. They'll be picking units that actually track value — tasks, decisions, records processed, something that moves in proportion to what agents actually do. The seat was always a proxy for consumption. When the nature of that consumption changes, the proxy breaks. There's also a margin story here that doesn't get enough attention. Agents don't need onboarding or support, and they don't churn because the interface was confusing. If you price on value rather than headcount, the unit economics can get better. The per-seat model isn't going away because SaaS is dying. It's going away because agents broke the assumption it was built on. The vendors that see this early enough will redesign the model. The ones that don't will keep charging per seat and wonder why the math stopped working. Over time, does this change how we pay for humans too?
-
Chandler K. posted thisNobody gets promoted for stopping a report. We measure what we add. Prompts sent. Reports generated. Tools adopted. Monthly active users. These go up and to the right. They look great on a slide. Eric Porres, Chief AI Officer at Logitech, puts it cleanly: "AI adoption is not a training problem. It's a deletion problem. The models are good enough. People are curious enough. The real bottleneck is who has permission to stop doing the old thing." That reframe is correct. And there's a deeper problem underneath it. Usage dashboards measure curiosity. Transformation shows up in what disappears — which reports stopped being written, which workflows collapsed from days to hours, which artifacts vanished because an agent replaced them. But organizations don't have a mechanism for valuing deletion. We have performance reviews, project trackers, launch announcements. None of them capture "I stopped 47 people from spending 90 minutes every Friday on a status email nobody read." That's not a KPI. That's not a career move. That's not something the dashboard shows. The AI dashboard is lying to you because the incentive structure underneath it rewards addition and ignores subtraction. Until leaders can be evaluated on what they had the courage to stop, the usage numbers will keep going up and the work won't change.
-
Chandler K. posted thisUser research isn't listening if you don't change anything. Most companies do user research to confirm, not to learn. The tell is what happens after. If the roadmap doesn't move — if the findings get filed, discussed in a meeting, and quietly set aside because they were inconvenient — the research didn't matter. It was expensive validation theater. Real listening is uncomfortable. It surfaces things you didn't want to know. It forces decisions you weren't ready to make. It requires the conviction to act on what you heard even when what you heard contradicts what you built. We've industrialized the appearance of listening. NPS surveys, user interviews, quarterly research sprints. All the infrastructure of a customer-centric organization without the actual behavior change that makes it real. The gap between what users say and what users do is where product strategy lives. Users describe their past behavior and surface preferences. They can't describe their future needs or underlying motivations. They can tell you what frustrated them. They can't tell you what would delight them. The builders who get this right aren't the ones with the most interviews. They're the ones who know which signal to trust, which to discount, and which apparent contradiction is actually the insight. What's the last thing a user told you that you didn't want to hear — and what did you do with it?
-
Chandler K. posted thisSome businesses survive not because they are hard to replicate. Because they are hard to abandon. We spend a lot of time on the first question. Moats, switching costs, network effects, proprietary data. The asset durability test — does this business own something competitors can't replicate? That's a good question. But it's not the only one. The harder question is different: what human commitments make this organization difficult to walk away from? For me, three forces answer it. Community — belonging that forms around the organization over time. Not customers in a database. Participants in a shared outcome. CrossFit isn't a fitness brand. It's a distributed network of local communities where people learn each other's names and come back day after day. You can copy the workout. You can't manufacture the belonging. Trust — relationships built through repeated cooperation that no competitor can manufacture overnight. Toyota spent decades building supplier relationships, sharing operational knowledge, supporting partners through downturns. That network is a structural advantage that doesn't appear on a balance sheet. Conviction — people who refuse to abandon a mission when the economics get hard. Independent bookstores were supposed to be finished. First by chains, then by Amazon. The ones that survived did so because the owners refused to leave — and reoriented around curation, community, local identity. Things a fulfillment warehouse cannot replicate. The American Booksellers Association has grown its membership for several consecutive years. That recovery wasn't driven by a shift in unit economics but by conviction that held through the period when the economics were worst. AI is making it easier to start things. It is not making it easier to sustain them. When execution becomes cheap, the advantage shifts from speed to endurance. What are you building that would be hard to abandon — and why?
-
Chandler K. posted thisLinear doesn't run A/B tests. Decisions are driven by taste and conviction, not engagement metrics. When they launched, the conventional wisdom — validated by years of enterprise sales and user research — was that teams wanted customizable project management. Flexibility. The ability to configure their own workflows. Every data point said build for optionality. The research wasn't wrong. Teams did want flexibility. What it couldn't surface was that the flexibility itself was the problem. Teams didn't know what their workflow should be. They wanted someone who understood software development deeply enough to make that decision for them. The desire for customization was a symptom. The actual problem was the absence of a strong, opinionated answer. Linear built exactly that. One right way to manage work. No A/B tests. No consensus roadmap. Decisions made by people with deep enough expertise to know what good looks like before the data confirms it. $1.25B valuation. Profitable within two years. 17 people at Series C. $35K lifetime marketing spend. The simplicity of the product isn't a design achievement. It's a diagnostic one. It proves the founders understood the problem — the actual problem, not the stated one — completely enough to make decisions users couldn't make for themselves. Simplicity is proof you understood the problem. Complexity is proof you didn't finish diagnosing it. What's the most complex part of your product right now — and is it complex because the problem is hard, or because you haven't fully understood it yet?
-
Chandler K. posted thisI asked dozens of executives, founders and operators to walk me through their best and worst decisions. One finding came back cleaner than anything else — and nobody was prompted toward it. When asked what makes a decision significant, *every* single participant arrived at the same answer independently: how hard it is to undo. Not the size of the decision. Not the stakes. Not how long it took. Whether you can reverse it. One participant put it directly: "Significant decisions are the permanent ones. The ones that close a door behind you." That convergence matters. It happened across founders, operators, executives, individual contributors — different industries, different contexts, different decision styles. Reversibility is the universal definition of significance. Here's what that finding unlocks. When you look at the decisions people were actually proudest of — the ones they'd make again — every single one looked wrong on paper. Pay cuts. Career pivots away from prestige. Moves away from family. Closed businesses. In every case, the conventional inputs — salary, stability, status, proximity to support — pointed the other direction. The signal that made them feel right wasn't data. It was values alignment. Purpose over pay. Growth over comfort. Health over advancement. Data informed those calls. It didn't drive them. This is the pattern that keeps getting missed. We've spent decades building decision frameworks that optimize for the legible — the data-supported, the defensible, the things you can put in a deck. But the decisions that actually compound in people's lives were made on a different signal entirely. The most irreversible decisions — the ones that close a door — are the ones least likely to have clean analytical support at the moment you make them. If you led with values on your best decision and data on your worst, you're not alone. That was the dominant pattern in the data.
-
Chandler K. shared thisIMO, A lot of “failed” software categories weren’t failures. We just built them wrong. Stuff like Fundraising tools. Personal CRMs. Planning systems. We treated them like SaaS problems—persistent, compounding, system-of-record products. They’re not. They’re high-stakes, context-dependent, one-off decisions that never should have been persistent. They should have been ephemeral. Full write-up below.
-
Chandler K. posted thisA few weeks ago I wrote about conviction versus confidence — acting before the signal is unambiguous. I've been very curious about this topic and decided to run a study to understand it more. I asked dozens of senior operators and founders to walk me through their best and worst decisions (thanks to those that participated!) — five modules of structured reflection across work and life. One pattern came back cleaner than anything else. The decisions people were proudest of looked wrong on paper. Pay cuts. Career pivots away from prestige. Moves away from family. Data informed those calls. It didn't drive them. The signal that made them feel right was values alignment — which is just another way of saying conviction. The decisions that broke down shared one failure mode: the decision-maker assumed alignment that didn't actually exist. They thought people were bought in. They weren't. And the most expensive part of the whole process — across the board — wasn't analysis. It was people. Coordinating, aligning, bringing stakeholders along after the call was already clear. One participant said it exactly: "The cost isn't in the decision — it's in alignment." This points to something that gets underbuilt. Decision support isn't just a thinking problem. Most capable people know how to think through a call. Where they want help is twofold: getting feedback from smart people to build the right context, and moving faster on alignment — cleaner communication, better pressure-testing, without scheduling six meetings. And most wanted some way to track how decisions played out — not just make them and move on. The opportunity isn't "make the decision for me." That's human. That's us. But there's lots of space around both "help me get to clarity faster" AND "bring the right people along without so much friction."
-
Chandler K. posted thisIn response to a recent post of mine, Matt Holden suggested skills.md files are the React components of the modern technology stack. He's right and really smart on this topic. The depth he's gone AI native on building modern software is more than anyone else I've chatted with, and even he is underselling this. Skills.md files are the React components of the AI native organization. They're the atomic units of action, instruction, and organizational direction — the place where a DRI's judgment gets distributed without a meeting, without a Slack thread, without a one-on-one. Right now, enablement is a distribution failure wearing a management failure's clothes. Someone learns something. They share it. The team forgets. You repeat the cycle. This isn't about better content (and never has been). Skills files break the cycle. When instructions live in a versioned, pushable, machine-readable file — enablement becomes infrastructure. You don't train people on the new process. You push an update. The organizational intelligence adopts it. This changes what leaders actually produce. An AI-native founder / CEO told me recently that their CTO had set the company loose to build whatever needed to be built — with one guardrail: don't touch my context files. Those eng-context.md files hold the CTO's decision posture, their guardrails, their non-negotiables. Not a doc. Not a wiki. A versioned file the whole system executes against. That's the model. A CPO does the same for product bets and prioritization logic. A CFO for capital allocation rules. A CEO for culture and performance standards. This also radically extends founders reach early as the play many of those roles - build the context files with advisors and smart input, and let the system run. Organizational strategy stops being a slide deck. It becomes a composition of context files — authored by the people most accountable, distributed to every agent and human executing against them. The constraint here isn't the file format. It's articulation. This only works if executives can actually enumerate their thinking — not just the output, but the reasoning behind the output. The decision logic. The context that produced it. Most leaders have never been asked to do that. They've been asked to decide. Writing down how they decide is a different discipline entirely. This is also the only way your agentic workers can live into your values, mission, and expectations — instead of inventing their own. The leaders and founders who build this muscle first will have a structural advantage — not an executional one. And structural advantages compound in ways executional ones never do. What barriers are showing up to doing this? One I've identified is that there simply is not a good system for maintaining skills.md and org_context.md files today - versioned, trustable, shareable, etc. Ramp's internal system is the closest I've seen. Anything better out there?
Chandler K. commented on a post
1w
Yup. They’re ahead (this week). As you said, will be interesting to see how it all evolves!
Chandler K. commented on a post
3w
Some of this is making it into the Guild solution, but my version of this is here: https://github.com/ckoglmeier/classroomThe question of "Given I built a capability (skill) or context, how might I distribute it without drift" is such a human-org problem, re-manifested through AI.
In many ways, I think this is more solvable in AI land. AI harnesses listen (most of the time) where we all know how hard it is to cascade intel or training into people in an org!
-
Chandler K. reacted on thisChandler K. reacted on thisToday, I’m thrilled to share that Natoma has entered into an agreement to join Snowflake. Since we started building the company in early 2024, we’ve gone through ups and downs, built an incredible team, served fantastic customers and seen a rapid increase in the pace of enterprise AI adoption. While much about the agentic future is uncertain, it’s clear that companies require easy, safe access to data and tools to fully see productivity wins. Natoma + Snowflake, together, are positioned to provide a world-class experience for organizations to securely enable their agents to access their most sensitive enterprise data. Starting Natoma has been one of the hardest and most rewarding journeys of my life. I’ve learned more than I ever thought I would in 2.5 years, worked with an incredible team and partnered with the best cofounders in the world. While this is a great milestone, the hard work really begins in earnest now. Our hour is now, and this opportunity is massive. Joining Snowflake will help us bring our vision to the best enterprises around the world. First off, thank you to Pratyus Patnaik, Paresh Bhaya, and Zachary Hart for embarking on this journey with me. Thank you to our amazing team for getting us to this point today- I can’t wait for the work we’ll do over the next few years. Thank you to our fantastic lead investors for believing in our team and our vision - Saam Motamedi @ Greylock Partners Partners, Shardul Shah @ Index Ventures Ventures, and Dave Zilberman @ Norwest In particular, thank you to Abby Hart, Dwane Hamilton, and Glen Evans for helping us land so many early key teammates. To my family, thank you for your support- without you, I wouldn’t still be in this role and we wouldn't be doing this today. Most importantly, thank you to my wife, Rosie, for believing in me and our team. You’ve sacrificed so much personally to let me live out this dream. You supported me on long nights, work trips, interrupted dinners and missed family events and have been there for me every step of the way. I love you! Onwards and upwards, and for my Natomies, let’s have some fun! 🚀 https://lnkd.in/dX2RMU3TSnowflake to Acquire Natoma to Bring Governed Agentic Access to the EnterpriseSnowflake to Acquire Natoma to Bring Governed Agentic Access to the Enterprise
-
Chandler K. liked thisChandler K. liked thisAnthropic shipped Live Artifacts in Claude about a month ago, and it hasn't gotten the attention it deserves. The pitch: build an artifact once, and it pulls fresh data from your connectors every time you open it. A personal MRR ticker streaming from Stripe. A custom HubSpot pipeline view that beats the native UI. A read-only Linear dashboard your manager can actually use. What's interesting is Live Artifacts turn connectors into the live backend for vibe-coded products. The artifact is the front-end you wrote in a single prompt. The connector is the data layer that keeps it useful tomorrow. Without persistence, vibe-coded apps die the moment you close the chat. With it, they become tools. (I'd bet ChatGPT ships a version of this within a quarter. It's too obvious not to.) If you've been on the fence about building a Claude connector, the surface area just expanded.
-
Chandler K. reacted on thisChandler K. reacted on thisToday, we’re announcing Humanly’s $25M Series B. When we started Humanly, we believed high volume hiring teams needed much more than another system to manage applicants. They needed a way to engage at scale, hire the right employees faster and build pipeline for the next requistion. While creating a better experience for candidates along the way. That belief feels even more urgent today. Employers are under enormous pressure to stay fully staffed with the best talent, while managing high application volume, recruiter burnout, and rising candidate expectations. Most hiring systems still weren’t built for that reality. Meanwhile companies are spending $100B+ per year to attract applicants with job ads and recruitment marketing, but only have the human time to engage with 5% of that demand. At Humanly, we’ve spent the last several years building conversational AI across chat, phone, and video to help companies engage every candidate, build talent pipelines, qualify talent faster, and move people from application to interview in seconds. We've also engaged with millions of job seekers and built relationships with them over time. And this is why I'm so excited about where we’re headed next. We’re expanding toward a Service-as-a-Software model where customers don’t just buy recruiting tools; they get hiring outcomes. That means a consistent flow of qualified candidates from Humanly, faster hiring, and fully staffed teams. Tools with candidate pipeline already built in! Huge thanks to our customers, team, and investors, including SEEK Investments, Drive Capital, Zeal Capital Partners, and MassMutual Ventures, for believing in this vision. And thank you to the millions of job seekers and candidates who’s interacted with Humanly along the way. We believe hiring should feel faster, more human, and more accessible for everyone involved. Read the full announcement here: https://lnkd.in/gVhcXJRSHumanly Raises $25M Series B to Scale AI Hiring for Frontline and High-Volume Recruiting | HumanlyHumanly Raises $25M Series B to Scale AI Hiring for Frontline and High-Volume Recruiting | Humanly
-
Chandler K. reacted on thisChandler K. reacted on thisWhen you have some of the smartest people in your country creating and selling AI while warning you about the dangers of their own technology, that message resonates, and it’s scary. Conversely, when you have some of the smartest people in your country creating and selling AI while actively promoting its upside, that message does not resonate; it looks like an advertisement. People in both camps continuously disparage the other, and AI sentiment is incredibly low in the US. The tech community continues to fail to communicate AI’s upside and pragmatic potential. There’s a lot to look forward to, and there will be bumps. Below are my high-level thoughts about the future of jobs and opportunities as we enter the AI era.When Life Gives You Lemons and AI: The Future of JobsWhen Life Gives You Lemons and AI: The Future of Jobs
-
Chandler K. reacted on thisChandler K. reacted on thisExcited to share that I recently joined Aura as their Chief Product Officer. Aura is building something genuinely interesting- an AI-powered platform that protects individuals and families across their entire digital lives. The product challenge is real and the market need is clear as individuals, parents, kids, businesses, and schools all face threats to safety and wellbeing. After speaking with CEO Hari Ravichandran and COO (and HBS classmate) Thomas Clayton about the company and role, I was drawn to the opportunity to help lead Aura's transformation into an AI-first product and engineering organization. We're early in deploying agentic development practices at scale, and I expect to learn as much as I contribute, but that combination of a product vision and a front-row seat to how AI is reshaping the way software and products get built was too much to pass up. Aura recently announced the acquisition of Australia-based Qoria and I got to a chance in the interview process to meet Tim Levy who has built an incredible mission-aligned business focused on digital wellbeing for schools and parents. The combination of Aura and Qoria is going to deliver an amazing platform for users worldwide. I'm also excited to have a chance to reunite with some of my favorite Tripadvisor engineering alums Rekha Singh, Erdinç Yılmazel, Bill Langenberg, Jay A. and Raksik Kim as well as get to know an entire group of new colleagues. Looking forward to sharing more in the weeks ahead. https://lnkd.in/ez8nz8Ad
-
Chandler K. reacted on thisChandler K. reacted on thisDuring a period of worker strife, it can feel borderline insensitive to talk about something going well. I hesitated to post about this for that reason. But I also think it’s important to recognize the bright spots. Highlight that there are companies, leaders, and workplaces where employees feel genuinely rewarded for their loyalty, commitment, and impact. Places where policies exist not just to attract talent, but to support and sustain workers over the long term. At Guild, employees become eligible for a five-week paid sabbatical after five years with the company. Maybe more importantly, it is a benefit people actually use. It’s not treated as performative or quietly discouraged - in fact, over the past five years I’ve watched colleagues across all teams and levels fully step away, recharge, and return. Sabbaticals have been modeled for me as a real and respected benefit at Guild. So in mid-April, I began my own five-week sabbatical. I spent the entire time in Europe, mostly in France, visiting museums, sitting at cafés, reading books, seeing friends, and even taking a pastry-making course at Le Cordon Bleu Paris (photos below). I didn’t bring my laptop. I took Slack off my phone. And I can proudly say I responded to only three emails (I'm a work in progress!). But before I left, I had a surprising amount of anticipatory anxiety. Like many of us, my identity is tightly wrapped up in my work: the impact I’m making, the leader I’m becoming, the sense of purpose and self-worth I derive from it. I was worried I’d be bored, or lonely, or maybe worst of all... that I’d actually miss work? But after my first few days in Paris, I found myself slowly relinquishing control and settling into truly being off. I trusted my teammates, especially Bijal Shah, Sarah Sawatzky, Beth Myer, Lisa Boyd, Zoe (Weintraub) Barrett, Kathy Qu, Matthew J. Daniel, Brooke LaRue, and others to step in while I was away, and slowly I let myself be more and more present in the moment instead of mentally tethered to work. And in that space, I found an overwhelming sense of gratitude. Much of it was gratitude for landing at Guild in 2021 (h/t Allison Dulin Salisbury), and for working at a company that genuinely walks the talk when it comes to valuing employees and investing in their well-being. Now, just three days into my first week back, I’m heading straight into Guild Gathers Live in Denver, our annual all-company event, which I’ve been joking is basically my welcome home party. So to anyone still waiting on a response from me, thank you for your patience. I’m still adjusting back from the slower pace of Parisian life :) #EmployeeSentiment #EmployeeBenefits #Sabbatical #WorkCulture #WorkLifeBalance #TimeOff #CorporateAmerica
-
Chandler K. reacted on thisChandler K. reacted on thisToday, The Institute of Evidence-based Policymaking (IEBP), a board I’m lucky to serve on. Announced that we are being renamed the Romer Institute of Evidence-based Policy in honor of my Grandparents, Roy and Bea Romer. We are also formalizing an affiliation with Metropolitan State University of Denver (MSU Denver) - the university that Granpa Roy championed into existence as a young state legislator more than six decades ago. As he once told me, it was one of the most As part of this affiliation, Granpa Roy will donate his papers to Metropolitan State University of Denver archives. The collection will be a valuable resource for research and study, spanning my grandpa's legislative career, three terms as Governor of Colorado, his tenure as Superintendent of the Los Angeles Unified School District, and his national work on education reform and democratic engagement. Grandpa built & lead institutions on the conviction that good decisions require honest evidence, and that common ground is more available than politics suggests, and that the people who govern communities deserve the best available information. The Romer Institute aims to carry that conviction forward. I'm grateful to see two great leaders, Shepard Nevel and Janine Davidson, Ph.D. leading this great partnership! https://lnkd.in/gv5V7ZaH
Experience & Education
-
Guild Education
View Chandler’s full experience
See their title, tenure and more.
or
Recommendations received
8 people have recommended Chandler
Join now to viewOther similar profiles
-
Ravi Mehta
Ravi Mehta
Senior product executive turned hands-on Product Advisor to high-growth consumer and AI startups.<br><br>I've built and scaled product organizations at brands reaching hundreds of millions of users—CPO at Tinder, Product Director at Facebook (Gen Z Engagement), VP Consumer Product at TripAdvisor, and early Xbox Live team at Microsoft. Across these roles, I've built teams of 40–70 PMs while driving billion-dollar business growth.<br><br>Today I selectively partner with founders to unlock product-market fit, durable growth, and AI advantage. Recent highlight: as interim CPO/Design lead at Sesame (direct-to-patient health marketplace), I drove a complete redesign, launched their mobile app in under 9 months, helped secure a successful Costco partnership, and shifted 60% of revenue to subscriptions—all with a lean six-person team.<br><br>I also co-created Reforge's AI Strategy program (with Brian Balfour) and created their Product Leadership program. My frameworks on goal setting, product competency development, strategy, and AI are used by thousands of product leaders globally.<br><br>Active angel investor focused on consumer, marketplace, and applied-AI opportunities. Always open to strategic conversations with ambitious founders building the next generation of products people love.
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content