Rob Duffy

Admin Cost vs. Benefit Amount — the Medicaid version A few days ago, Evan Brociner posted about admin costs growing faster than the rest of hospitals' budgets (link: https://lnkd.in/eaTYnx-x). Made me wonder: does the same pattern show up somewhere else in healthcare? For example, Medicaid? We just pulled in CMS's MBES Financial Management Report data (the CMS-64 report every state files quarterly) and ingested those in mimilabs — three tables: benefit payments (mbes_fmr_map), administrative costs (mbes_fmr_adm), and CHIP (mbes_fmr_chip). Each row gives you state, fiscal_year, and total_computable spending by category, going back years. It lets you separate "money that paid for care" from "money that ran the program." So I tracked all 50 states + DC, FY2019 to FY2024, indexed to each state's own 2019 baseline (not absolute dollars — benefit spending is naturally many multiples larger than admin spending everywhere). Turns out every state tells a different story: → New York: admin costs jumped +72% in a single year, FY2023 to FY2024 — right as pandemic-era continuous enrollment protections ended and the state had to re-determine eligibility for millions of people at once. → Texas: the odd one out. Benefit payments actually declined in FY2024 (fewer people enrolled, post-unwinding) — but admin costs kept climbing. The cost of processing eligibility didn't fall even as the caseload did. → North Carolina: not an admin story at all. Benefit spending more than doubled in FY2024 because Medicaid expansion took effect there in December 2023. (I believe...) Same federal reporting system, same six years, three completely different narratives — one driven by a compliance surge (unwinding), one by sticky fixed costs, one by a coverage policy change, not admin bloat at all. I think the nuance from Evan's post applies here too: a rising admin line mixes real waste (redundant eligibility paperwork, duplicate systems) with legitimate cost (staffing up to actually re-verify tens of millions of people correctly, so eligible people don't lose coverage). Before assuming "admin bloat," it's worth asking what drove the spike in each state. Which means that you can dig into these datasets in mimilabs!! Data: mimi_ws_1.datamedicaidgov (mbes_fmr_map, mbes_fmr_adm) — CMS-64/MBES Financial Management Report, FY2019-2024.

27

8 Comments

Here’s why Claude’s recent progress has unlocked the next leap forward for Canvas Medical. Claude’s tools, combined with our Deep Unified Architecture and SDK, are enabling new use cases for automation while simultaneously making us exponentially faster in shipping customer specific enhancements. It's a double win. Said another way, Canvas is an EMR built to provide context and tools for automation. And now we can enable AI driven development for any user or customer, whether technical or non technical, to build their own custom workflows in Canvas. Just in the past couple of days our team has shipped: For Brigade Health - a customized panel management dashboard for high-risk patients For Vida Health - ability to parse unstructured lab results from a fax into the chart as structured clinical data For Radial - improvements to the provider onboarding experience to accelerate the rapid scaling of their medical group operations For Doctronic - in-workflow evals for clinicians to submit structured feedback on the performance of their AI-driven note generation The best part? All of this work was done without software engineers 🤯 We used the Canvas Plugin Assistant, powered by Claude Code and the Canvas SDK. We’re living in the future. Here’s a graphic from Anthropic’s presentation yesterday with a couple of annotations to explain. Sometimes a picture is worth a few hundred words.

66

7 Comments

Fixing GTM can unlock 50% faster SaaS growth in 2026, without adding headcount or ad spend. Teams that get GTM right scale faster with the same resources. Teams that don’t stay busy and stuck. This 2026 GTM Readiness Checklist helps you fix that before you scale. Quick GTM check: • One clear ICP (not 3) • Value explained without a demo • One dominant GTM motion • Sales that’s repeatable • Retention and expansion prioritized Avoid these 2026 mistakes: • Chasing multiple channels • Scaling sales before GTM fit • Confusing activity with traction Fix GTM first. Growth compounds after. Image Credit: Alexander Estner Comment “GTM” and I’ll send you the checklist link.

11

2 Comments

AI Is Rewriting the Rules of Go-to-Market (GTM) As someone who’s spent years aligning strategy, systems, and outcomes across product, engineering, and GTM teams — this article by Dave Birckhead really hits home. In his latest Full-Stack Growth post, “Why AI Requires a New Kind of GTM Role and Org Structure,” Birckhead explains why traditional functional silos — marketing ops, sales ops, customer success ops — can’t keep pace in the AI era. AI doesn’t thrive in isolation. It thrives on shared context, unified data, and connected workflows. Most enterprise AI projects fail not because the models are weak, but because the systems they live in are fragmented. The takeaway: AI value is created when organizations reimagine entire workflows that span the customer journey — using AI as connective tissue, not as point solutions. Birckhead makes a compelling case for a new leadership role: Head of GTM Systems — the orchestrator of data, tools, and AI workflows across marketing, sales, and customer success. The payoff? ✅ Faster innovation ✅ Higher ROI on AI investments ✅ Seamless customer experience ✅ Shared measurement across the funnel Just as SaaS created Marketing Ops and RevOps, the rise of AI now demands GTM Systems Leadership — a discipline that connects, aligns, and scales the way growth truly happens. 📖 Full article here: https://lnkd.in/gU2fa5Jn By Dave Birckhead, Full-Stack Growth (Oct 14, 2025) #AI #GoToMarket #ProductOperations #GTMSystems #RevOps #DigitalTransformation #Leadership #Strategy #Innovation #FutureOfWork #OperationalExcellence #ProductManagement #SalesOps #MarketingOps #CustomerSuccess #B2BLeadership

11

2 Comments

From thousands of conversations, most outbound teams don't lack tools. They lack a single layer that ties everything together. We fixed that this week by connecting Claude to the entire Forge Stack via the Salesforge MCP server. Here's what that actually means: Claude now has live access to: → Every campaign, sequence, and sending account → Inbox activity, warm-up behavior, deliverability signals → Lead segments, enrichment data, infra config → Performance gaps across the whole system Instead of jumping between dashboards, you ask a question. Claude organizes the data, surfaces what matters, and tells you exactly what to fix. Simply put, Salesforge MCP + Claude = More 🥓 made We built 5 core workflows around this: 1. Campaign Usage Monitor 2. Performance Analyst 3. Campaign Health Dashboard 4. Infrastructure & Inbox Monitor 5. Full Outbound Audit Report Each one is a single prompt. Each one replaces hours of manual review. This isn't reporting. It's your outbound operating system. Full guide with workflows, prompts, and setup guide prepped for you. Comment “Claude” below and I’ll send you the guide. P.S Repost for priority access 🥓

51

34 Comments

If you feel like the STR industry is moving faster than guests with the option for an early check-in, you’re not wrong. Here’s the major headlines your google alerts didn't send you. (source links in comments) 🔴 With Sonder’s bankruptcy making headlines, Marriott is now alleging that the company put guest safety at risk while seeking financial support, leading to a swift end to their partnership amid a chaotic wind-down. (source: Business Insider) 🤝 Beyond and Boom and have joined forces to create an AI-powered, fully connected ecosystem that brings intelligent automation, real-time data, and dynamic pricing together to redefine the future of property management and revenue growth. (source: Beyond) ⛳ East Lothian is considering temporary short-term rental licenses during major events like the Genesis Scottish Open, a move designed to boost accommodation supply and support visitor demand during peak periods. (source: Golfweek) ⚽ And in Kansas City, a similar proposal is taking shape: the city is weighing a temporary STR permit (just $50) for the 2026 World Cup (and other major local events - read: when TSwift watches the Chiefs play) to help residents participate economically and meet the surge in lodging demand. Check it out Lindsey Branding (source: The Kansas City Star) 🏘️ Airbnb reported a record-breaking Q3 with $4.1B in revenue and $1.4B in profit, alongside major updates for hosts, including “Reserve Now, Pay Later,” expanded AI tools, 65 product improvements, and a pilot program for hotel listings in key cities. (source: The Host Report)   📸 Vrbo now lets guests upload their own photos (appearing in reviews, main galleries, and AI review summaries), adding more transparency and real-world context to the booking experience. (source: Skift) 🇬🇧 Airbnb has launched a £1 million “Best of British” fund, backed by VisitBritain, offering grants of up to £100,000 to help communities turn local traditions into cultural experiences and attract more evenly distributed tourism across the UK. (source: ShortTermRentalz) #shorttermrentals #vacationrentals

81

5 Comments

If you run a business outside of software, you’ve likely missed how completely the software development field has been disrupted. Over the past year, I’ve spoken with dozens of leaders across independent and family-owned businesses in New England. One pattern keeps coming up: Many smart, experienced executives haven’t seen how dramatically the world of knowledge work has ALREADY changed because of AI. The first profession to be radically transformed by AI was software engineering. Much of what used to take a team weeks - designing, writing, and testing code - specialized AI can now do (mostly) reliably in literal minutes or hours, with humans increasingly acting as orchestrators & reviewers. Today, even at giant longstanding companies like Microsoft and Google, ~30% of ALL NEW CODE is now AI-generated and rising quickly according to Microsoft CEO Satya Nadella. And they have thousands of engineers... at earlier stage startups with less legacy code, this percentage can be far higher 🤯 Let that sink in. For decades, learning to code was considered one of the most durable career paths in the world. Within ~3 years, AI began writing, designing, and reviewing a large portion of that code. Tech job postings are still down 36% from 2020 levels (Indeed Hiring Lab), and hiring of new graduates in tech has fallen more than 50% (SignalFire). Senior engineers are increasingly orchestrating teams of specialized AI systems to do far more with far fewer people. So why did this disruption happen to software first? Because improving coding ability is the highest-ROI capability for frontier AI labs (Anthropic, OpenAI). If AI can write better code, it helps engineers build better AI models. Those better systems then improve themselves - a powerful feedback loop. But they are not stopping there. OpenAI has reportedly hired 100+ former Wall Street investment bankers to train models on financial analysis. Across the industry, teams of lawyers, analysts, economists, and domain experts are now being recruited to train AI in their fields. Their goal is simple: Teach AI to perform every category of knowledge work done on a computer. Which is… most white-collar work. Meanwhile, many businesses outside tech still treat AI like: • a future discussion • a curiosity • or a productivity tool for interns That is a dangerous misunderstanding. Large corporations and Private Equity firms are already investing tens of billions into AI systems, data infrastructure, and automation across their companies. They are not experimenting. They are rebuilding how companies operate❗ If independent and family-owned businesses wait years to respond, they will be competing against organizations that have been compounding AI & data advantages the entire time. The AI revolution is not a future conversation. It is already meaningfully restructuring the workforce. And many businesses outside software still don’t realize it. ~85% of New Hampshire businesses are family-owned!

24

For years, SaaS meant control. Teams would delay launch for every round of QA and debate, pushing live only when everything felt predictable. You needed proof in hand before anyone saw your work, because reputation, retention, and revenue depended on minimizing every unknown. That old instinct to launch only when everything’s perfect? In AI, it’s the main reason teams fall behind. The classic SaaS approach worked because customers bought consistency. Sales pitches highlighted reliability, support played defense, and new features waited for a green light from every internal gatekeeper. Teams measured maturity by what they hid. Bugs, doubts and surprises. A good launch felt like closure. AI flips the table. Now, every release happens in public. Models change, data shifts, and product value doesn’t show up on launch day. It emerges after weeks of live interaction and blunt feedback. AI products stop pretending they’re finished. The ones that lead put their rough edges on display and treat every flaw as tomorrow’s advantage. Every SaaS team was trained to avoid mistakes in public. The teams thriving now are the ones showing the messy middle, letting users witness how the product changes under real pressure. They launch an early version, watch what users break, publish their limits, and ask for help naming what doesn’t work. While others polish and hesitate, these teams collect four cycles of hard-earned learning for every one release that the old guard ships. They turn error messages into feature requests and respond to friction faster than anyone else can copy them. Teams that hesitate for certainty fall further behind with every iteration. In AI, there’s no finish line. Only a faster cycle to discover what breaks next. The hardest part is unlearning the need for control. What really separates the winners now? They replace every instinct to control with the habit of chasing their own blind spots. You build momentum in AI by giving users a front row seat to your product’s evolution. Share the limits, welcome the awkward feedback, and treat every launch as an open invitation for your market to teach you faster. Teams who delay for certainty miss the next cycle of learning and let others set the standard. The old goal was certainty before launch. The new reality is you launch to earn it.

24

9 Comments