How to prove AI's value to CFOs with hard and squishy ROI

There are a lot of questions swirling about the value that AI transformation is having on the enterprise and whether or not this can or even should be measured. At UserTesting, we separate our value measurements into two categories: *Hard ROI - prove the math by measuring the impact of specific AI deployments in the company. *Squishy ROI - evaluate the culture change that makes the math repeatable. Many thanks to Greg Shove and the Section team for spotlighting this at AI:ROI. You can read more in Greg's post below and in Sections latest blog post (link in comments).

📊 "We're spending $500/month on AI—what are we actually getting for this?" If you're leading AI adoption inside an SMB, this moment will come. The CFO email. The budget review. The ROI reckoning. You know AI is helping: → Proposals move faster → Customers get better service → Your team breathes easier But if you're flying on vibes instead of numbers, it's time to fix that. Here’s a 3-step playbook to measure what actually matters: 1. Audit Your Usage Start a #ai-show-and-tell channel. Ask your team: → What AI tools do you use? → What tasks do they help with? → How much time do they save? 2. Build Your ROI Jig Create a GPT, Claude Project, or Gemini Gem that analyzes: → 💸 Immediate impact (time saved, rev accelerated) → 🧠 Strategic readiness (team capability, agility, brand) 3. Create a Dashboard Your jig should produce a simple artifact that shows: → ROI % → Time/money saved → Strategic lifts worth doubling down on Your goal? Prove the $500/month investment is driving $50k+ in returns—and make the case to double down. 👉 Full guide: https://lnkd.in/eZsF7vXy ✨ ✌🏻 ✨ #generativeAI #SMB #AIroi #growthops #budget #audit #getbetter #smb #moment #doubledown #strategic #finance #usage #customers #makersgonnamake #aiadoption #showandtell #knowai #breatheeasy #movefaster #email #customer #aitools #getgood #number #project #task #start #goodservice #proposals #month #breatheeasier #cfo #betterservice

I spent the week in DC with Pavilion digging into how AI is actually being applied in GTM. Lots of smart operators, investors, and enablement leaders shared what’s working and what’s not. Here are my Top 10 AI Takeaways for GTM Teams: 1. AI should enhance top performers and lift low performers, not replace them. It will not replace the need for sales and CS training on core skills to land, expand, and renew 2. AI adoption takes time. Map where you are on a maturity curve (piloting, scaling, or high-value) and set next-step goals. 3. Start with proven use cases: signal capture, account research, high intent outbound, and account prioritization. 4. Push for clear ROI metrics. Track often and iterate from learnings. 5. Get your playbooks right first with messaging, objections, and handoffs, then use AI to accelerate execution. 6. Data quality and ownership matter more than the tools. Weak foundations stall adoption. 7. Remove blockers like process ambiguity, governance gaps, and lack of time. They drag everything else down. 8. Boosters such as training and template libraries only work once blockers are gone. 9. Achieving high value from AI takes time. Only 1 in 5 companies are there today. 10. The risk of AI backfiring is low. Only 6 percent of companies reported issues. 🤖Bottom line: Let AI handle the busy work so your reps can focus on selling. #GTM #AI

Quick Wins vs. Strategic Bets When it comes to AI in GTM, not every use case should be treated the same. You need both quick wins and strategic bets. But knowing which is which and when to pursue them can be the difference between momentum and stall. I learned this the hard way. The Big Bet (too early): We rolled out AI-driven forecasting. The goal was better predictability, but leadership didn’t trust it. Instead of anchoring to the AI output, we ended up treating it as a scenario while still relying on our traditional forecasting methodology to drive the actual forecast. The lesson? Without trust, even the smartest AI model becomes background noise. The Quick Win (what worked): On the other hand, introducing AI deep research into early stage discovery completely. With deeper research and sharper prompts, reps uncovered value drivers faster and built stronger business cases. That wasn’t just incremental it reshaped how customers engaged with us. The reps trusted it, the customers felt it, and leadership saw the impact immediately. Here’s the approach I follow now: Quick Wins: Low friction, fast adoption, visible ROI. Example: AI-augmented discovery conversations that lead to bigger business cases. Strategic Bets: High potential, but adoption requires time, trust, and cultural buy-in. Example: AI-driven forecasting. Push the big bets too early, and you risk credibility. Stack quick wins, and you create momentum for the bigger plays.

There’s plenty of AI ROI - but only if you’re prepared to put in the work. I joined Section’s AI:ROI Conference last week, led by Section CEO Greg Shove, and what really stood out was the honesty. Real leaders, sharing the reality: AI ROI exists, but it doesn’t fall into place by itself. The big takeaways for me: 👉You need basic AI proficiency in place before chasing advanced workflows. 👉Track every project for time saved, costs reduced, or value added. 👉Recognise the difference between “hard ROI” (numbers your CFO will love) and “squishy ROI” (strategic bets you believe in). 👉Change management is half the battle, adoption doesn’t happen with one memo, it needs constant communication and champions. The reality: AI ROI isn’t a quick destination, it’s an ongoing practice. The companies seeing returns are rolling up their sleeves, measuring, and building adoption step by step. I’ve linked the Section playbook from the event below, it’s full of practical examples and well worth a read and well worth following Section! For my part, I help businesses work through this with a Roadmap and Framework that makes the first steps less overwhelming and more structured. DM me if you would like to have a chat!

Quotations 📚 “AI ROI isn’t achieved through hype—it’s earned through disciplined measurement, behavioral adoption, and workflow redesign.” 📚 “Most AI projects can’t predict returns accurately upfront. Build measurement discipline gradually rather than demanding precision.” 📚 “The fundamental question isn’t financial return—it’s whether you trust and empower your people with tools that obviously provide value.” 📚 “Federated governance balances speed with safety—centralized control without killing innovation.” 📚 “Stop cloning ChatGPT. Buy proven tools off-the-shelf and innovate where your data gives you an edge.” 📚 “AI slop forces human accountability back into focus—machines amplify judgment, not replace it.” 📚 “The application layer, not infrastructure, will determine the long-term winners.” 📚 “ROI comes in three currencies: time saved, quality improved, and new revenue discovered.” Key Points 📚 AI ROI evolves across three phases: pilot ROI → scaled ROI → post-implementation ROI frameworks. 📚 Four hidden barriers prevent ROI realization: 1. No “Why of AI” – lack of strategic purpose. 2. Limited access – only 29% of employees have usable tools. 3. Over-restrictive policy – fear and unclear guidance stall experimentation. 4. No use cases – lack of applied, measurable impact in daily work. 📚 Enterprise tactics: – Novo Nordisk: Map psychological barriers first; build AI adoption through ambassador networks. – Mars: Use federated governance; measure ROI via probabilistic portfolios, not rigid cost savings. – AbbVie: Apply the “Stop, Easy, Stick” litmus test—what behaviors change, how easy is adoption, and will it sustain? – Autodesk: Build champions inside teams; pick three workflows and measure ruthlessly. – Logitech: Use outtake forms to quantify project results—ROI = efficiency + quality + revenue. – Moderna: Treat AI as infrastructure—focus on trust, empowerment, and new human capabilities. 📚 Investor insight (Goldman Sachs, IBM): The application layer will produce outsized returns; vertical AI startups outperform generalized tools. 📚 Measurement insight: Effective AI programs focus on behavioral pull (employees demand access), not push (mandated adoption). 📚 Cultural insight: AI success depends on psychological safety, internal marketing, and empowerment, not technical deployment alone. Headlines 📚 “AI ROI Is Behavioral, Not Just Financial” 📚 “From Pilots to Platforms: How Enterprises Move Beyond AI Experiments” 📚 “The 4 Hidden Barriers Sabotaging AI ROI” 📚 “The Future Belongs to the Application Layer, Not Infrastructure” #AIROI #ExecutiveStrategy #AIGovernance #DigitalTransformation #FutureOfWork #AIAdoption #AILeadership #ResponsibleAI #InnovationEconomy

Your AI won’t scale until your data does. You win by activating your data. I’ve taught 1,000's leaders what actually works: action before hype and the order to follow to get it right. Here’s the 6-step 𝗗𝗮𝘁𝗮 𝗔𝗰𝘁𝗶𝘃𝗮𝘁𝗶𝗼𝗻 flow (the baseline for durable AI ROI): 𝟭) 𝗡𝗲𝗲𝗱𝘀 — Define business outcomes and the KPIs that matter. 𝟮) 𝗔𝗰𝗾𝘂𝗶𝗿𝗲 — Identify the exact data sources required to hit those outcomes. 𝟯) 𝗣𝗿𝗲𝗽𝗮𝗿𝗲 — Turn raw data into reusable, reliable tables/pipelines. 𝟰) 𝗠𝗼𝗱𝗲𝗹 — Choose/fit models, tools, agents needed to leverage the data. 𝟱) 𝗣𝗿𝗲𝗱𝗶𝗰𝘁 — Leverage the models to predict and generate insights. 𝟲) 𝗔𝗰𝘁𝗶𝘃𝗮𝘁𝗲 — Automate + integrate predictions to create compounding value. Do it in order → teams build capacity, not just prototypes. Masters of this process create new value consistently. Masters of this process can create new value consistently and quickly. I vibe coded an interactive version to share. Interestingly, I built it live, on stage, in front of ~1000 people. Companies have created 𝗺𝗶𝗹𝗹𝗶𝗼𝗻𝘀 in value by running this playbook. Comment “𝗔𝗖𝗧𝗜𝗩𝗔𝗧𝗘” + follow—I’ll DM you the link.

McKinsey’s latest on generative AI captures what we’re seeing across organizations today: AI’s real value isn’t in experimentation — it’s in rewiring how a business operates to create measurable advantage. That requires moving beyond isolated pilots to address the fundamentals: - Building organizational and technical architectures that enable scale, not fragmentation. - Upskilling teams in gen-AI-specific skills like context engineering, prompt design, and applied data science. - Establishing centralized governance for responsible scaling — ensuring standards, security, and model quality. - Strengthening data foundations, especially around unstructured knowledge that fuels real-world use cases. - And embedding trust, transparency, and reusability, so every AI initiative becomes part of a scalable system. At FEAT, this is the strategy we build and execute with our clients. We don’t just deploy AI tools — we rethink the wiring of how people, independent expertise, and business models connect. That means shifting from valuing inputs (hours worked) to outputs (impact created), and realigning compensation, performance, and leadership incentives around measurable productivity and value. And we do it at one-third the cost of traditional consulting, with operators who execute, not just advise. --- Discover The FEAT’s approach to building the infrastructure of modern work: https://lnkd.in/gditqC-j #GenerativeAI #DigitalTransformation #OrganizationalDesign #AIIntegration #FutureOfWork #PrivateEquity #StrategyExecution #FractionalLeadership #TheFEAT #BusinessTransformation

Your proposal sounds perfect. That's the problem. We've all seen it happen: A confident presenter walks into a meeting with what seems like a brilliant proposal. Ten minutes later, it's falling apart. ===== AIAndYou Tip 11: The Devil's Advocate 😈 Ask AI to argue against your ideas before you pitch them. Picture this scenario: Someone spends hours perfecting their pitch, covering all the benefits, building the business case. They walk in confident. Then the questions start flying: "What about budget constraints?" "Have you considered the timeline risks?" "How does this align with our priorities?" The problem: They'd only argued FOR their idea. They'd never argued AGAINST it. ===== Here's what changes everything: Using AI as your harshest critic before any important pitch. Try this approach: "I think we should implement this new customer onboarding system. What are the strongest arguments against this approach?" AI might reveal: - Implementation costs you hadn't fully calculated - Staff training time you'd underestimated - Integration risks with existing systems - Potential customer confusion during transition 🌱 The transformation: Instead of being blindsided by these objections, you prepare for them. You have data, mitigation plans, and alternative approaches ready. When those questions come up: "Great question - here's exactly how we'd handle that..." The principle: Your best ideas need to survive their worst critics. AI can be that critic - safely, privately, before it matters. Try this: Before your next big pitch, ask AI to tear it apart. Your proposals will thank you. ===== #AIAndYou #RiskManagement #ArgumentBuilding #StrongerProposals #CriticalThinking #PresentationSkills #SmartAIHabits #ProfessionalDevelopment

We’ve talked about avoiding copy-paste plans and fixing broken budget logic. But there’s a bigger shift for 2026, one everyone keeps talking about, but very few are planning for with ambition. The question isn’t “should we use AI?” anymore. It’s “how does this stuff actually change what we can do?” Most messy-middle companies are trying to bolt AI onto old systems. It’s like putting a jet engine on a shopping trolley...sure, it moves faster, but not necessarily in a good way. They’re planning around the tools instead of rebuilding for them. That’s the trap. AI doesn’t just make the same work faster, it changes the playing field. A few examples I’ve seen lately: - Content that once needed an agency can now happen in-house, if you plan for it. - Personalisation that used to be enterprise-only is suddenly accessible, if you design for it. - Reporting that required analysts is now automated, if you structure for it. Here’s a shift for 2026: Stop planning headcount around workload, start planning capability around potential. That designer you were about to hire? Maybe you need a strategist who knows how to direct AI tools instead. That content agency producing 20 pieces a month? You could be doing 100 yourself with the right workflow. The winners next year won’t have the biggest teams. They’ll have the smartest human-AI systems, people who know how to use leverage, not just labour. Here’s the question: Are you planning to use AI... or planning around what it truly enables? Tomorrow: before you finalise any 2026 plans, there’s one diagnostic step most companies skip (and it’s costing them). #MarketingStrategy #BusinessGrowth #ScalingUp #MessyMiddle #2026Planning #Leadership #AI #MarketingInnovation