AI Innovation Management

Want to start seeing value from your AI initiatives? Stop worrying about the ROI… At the WSJ Leadership Institute’s Technology Council Summit last week, CIOs and tech leaders starting speaking more sense about AI…. Traditional ROI metrics just aren’t cutting it and are starting to change their tune. Which is the smartest thing they can do. Here’s why... What were experiencing isn’t digital transformation. It’s innovation. And innovation just can’t be measured in the same way. 👉 Digital Transformation = Known tools, proven best practices, predictable ROI. Think, moving to cloud, implementing Salesforce, digitizing manual processes with proven tools 👉 AI Innovation = Emerging tech, experimental approaches, uncertain outcomes. Think, LLMs, Agentic AI, generative systems. Value from digital transformation comes from execution. Value from innovation only comes at scale.... and AI scale just isn’t there, yet. Leaders have two choices... Choice One: Treat AI like digital transformation ☑️ Expecting quarterly ROI like mature enterprise software ☑️ Abandoning projects when ChatGPT doesn’t work like Excel and assume it failed ☑️ Applying operations metrics (95%+ reliability) to innovation work Choice Two: Treating AI like innovation ☑️ Accepting that emerging tech often takes 5–10 years to mature ☑️ Expecting many experiments to “fail” (and that’s okay), it's what you learn from it that holds value ☑️ Measuring learning velocity, not immediate returns If you're one of the leaders starting to make this mind shift, you need to take a three tiered approach (enterprise led) 1️⃣ Top-Down: Strategic innovation portfolio --> Identify 2-3 bets that could be game changers --> Create long-term horizons, not just quarterly targets --> Invest in data, governance, and talent early 2️⃣ Bottom-Up: Innovation experimentation and enablement --> Allow controlled failure and iteration across the organization --> Separate innovation budgets from operations --> Enablement team driving postmortems, cross-team learning 3️⃣ Horizontal opportunities: Scaling what works --> The bottom-up experiments that consistently generate insight become enterprise-scale initiatives --> Centralizing enablement and oversight to avoid “AI for AI’s sake” while keeping enough freedom to innovate. The bottom line? Those organizations that are moving the need aren't showing huge ROI today they're; ☑️ Using innovation metrics: learning velocity, adoption, capability building ☑️ Planning with longer timelines: 5+ years, not just the next quarter ☑️ Fostering a mindset: “What can we learn?” vs. “What can we immediately deliver?” Digital transformation delivers predictable ROI in 12-18 months. AI innovation builds foundations for 2030, not just 2025. Article: in comments

💡 The AI honeymoon is over, and most organizations have little to show for it. After years of pilots, proof-of-concepts, and innovation theater, BCG reports only 26% of companies have deployed working AI products—and a mere 4% see meaningful returns. The problem isn't technology. It's the absence of disciplined strategy married to human purpose. I've spent three decades watching brilliant technologies fail not from technical shortcomings, but from organizational incoherence. AI is no different. What separates companies that generate real value from those burning resources on experiments that go nowhere? Two things: strategic discipline and portfolio thinking. In our recent Harvard Business Review articles, we explore how organizations can move beyond the chaos: First, balance innovation with governance using practical frameworks. Our OPEN and CARE framework provide structured ways to ask the right questions early — questions that align AI with genuine business priorities while protecting against risks that emerge when we automate without thinking. This isn't about slowing down or creating bureaucratic bottlenecks. It's about moving forward with intention, ensuring every AI initiative serves both business value and human purpose. Second, treat AI as a portfolio, not a collection of pet projects. Organizations like Northrop Grumman, PepsiCo, and Lloyds Banking Group have proven that structured portfolio management—complete with prioritization frameworks, resource allocation discipline, and clear buy/sell/hold decisions—transforms AI from cost center to strategic asset. When you combine these approaches, something fundamental shifts. AI stops being something bolted onto strategy and becomes inseparable from it. The result: better returns, less waste, and organizations that remain distinctly human even as they become more technologically capable. The question isn't whether to invest in AI. It's whether you're managing those investments with the same rigor you'd apply to any other strategic portfolio. 🔗 Read further @ 📍 "Two Frameworks for Balancing AI Innovation and Risk" → https://lnkd.in/edHnUzGK 📍 "Manage Your AI Investments Like a Portfolio" [with/ Tom Davenport, Paul Scade, PhD, Erik Nelson] → https://lnkd.in/gEJ_WnyM What's blocking your organization from moving AI from experiments to enterprise value? I'm curious what you're seeing.

AI agents are reshaping how enterprises innovate, organize work, and experience disruption. In the latest episode of “What’s the BUZZ?”, Christian Muehlroth, CEO of ITONICS, shares how agentic AI will redefine innovation management and why many organizations are still structurally unprepared for it. Here are four key insights from the conversation: 1. Invention is not innovation AI is not “new” as an invention. The mathematical foundations and core ideas have been around for decades. What changed is innovation through scalable infrastructure, powerful interfaces, and new delivery models that made AI usable at scale. Leaders who confuse invention with innovation often miss the real inflection points. 2. Innovation waves are compressing AI is the latest long-term “innovation waves” (such as steam, electricity, and the internet). Each wave now builds on previous ones, compressing time and increasing the sense of acceleration. This creates a dangerous gap where technology accelerates exponentially while large organizations slow down due to processes, politics, and policies. 3. Agents drive an “abundance of labor” AI agents today resemble tireless digital interns that can reach expert-level performance on specific tasks but still require oversight. As costs for this kind of digital labor trend toward near-zero, the constraint shifts from headcount to ideas and throughput. The real leverage lies in using agentic AI to amplify people with initiative, creativity, and ownership, not in blanket rollouts that dilute impact. 4. Avoid AI tourism: fix foundations first Many enterprises try to “put AI on top” of legacy processes and public LLMs. The result is AI tourism: experiments that look impressive but lack strategic value. The real work is often less glamorous. Redesign processes instead of automating inefficient ones. Build a clean enterprise data foundation (customer insights, patents, portfolio, pipeline, competitive data). Use secure, enterprise-grade AI setups where data governance and context are under control. Leaders need to take disruption seriously, double down on strategic intelligence, empower the people who want change, and invest in data and platform foundations before scaling agents. Listen to the full episode of “What’s the BUZZ?” tonight at 8pm ET to dive deeper into how agentic AI will shape the next wave of business innovation, and subscribe on your preferred podcast platform to stay ahead of the curve. Is Agentic AI already disrupting businesses (or can we just not see it yet)? #ArtificialIntelligence #Innovation #IntelligenceBriefing

We’ve been working closely with organizations to explore how AI is fundamentally reshaping their approach to innovation. In our view, this evolution unfolds in three distinct stages: 1. Conversational Innovation — where early adopters use AI to enhance individual productivity and creativity. 2. Facilitated Innovation — where leading enterprises embed AI into structured innovation across the enterprise, driving scale, efficiency, and stronger returns. 3. Autonomous Innovation: A future phase in which systems will begin to independently identify, prioritize, and act on opportunities. While this is still nearly a bit away, the foundation is being laid today. Organizations building the right infrastructure — AI-enabled decision systems, feedback loops, and integration with core operations — are already seeing meaningful gains in speed, capability, and commercialization success. This article introduces the Innovation Intelligence Curve — a model for understanding this progression and why the most disciplined companies are best positioned to lead in the age of autonomous innovation. Curious how others are approaching this shift inside their organizations — and where they see it heading.

💡 Are Compliance Standards Killing Innovation, or Are We Framing Them Wrong?💡 Compliance standards are often viewed as barriers to creativity, especially in fields like artificial intelligence (AI). But frameworks like ISO42001 are not obstacles as much as they are enablers. They provide the structure needed to innovate responsibly, ensuring organizations can offer accountability, trust, and scalability. For leaders implementing an Artificial Intelligence Management System (AIMS), conformance to the standard can help establish a foundation for trustworthy AI systems, reducing risks and enabling sustainable innovation that also aligns with the OECD.AI’s Principles. ➡️ How ISO42001 Drives AI Innovation 1. Clarity Creates Confidence 🔹 Challenge: Teams hesitate to deploy AI when risks like bias or privacy breaches remain unresolved. 🔹ISO42001 Solution: Establishes clear processes for risk management, documentation, and decision traceability. 🔸Impact: Developers can innovate confidently within a framework that reduces uncertainty. 2. Risk Management Enables Bold Ideas 🔹Challenge: AI development involves unpredictable outcomes and operational risks. 🔹ISO42001 Solution: Provides structured tools to identify, mitigate, and monitor risks throughout the AI lifecycle. 🔸Impact: Teams can pursue ambitious ideas with safeguards in place, balancing creativity with accountability. 3. Accountability Builds Trust 🔹Challenge: Stakeholders demand transparency and fairness in AI decision-making. 🔹ISO42001 Solution: Embeds accountability mechanisms, ensuring decisions are traceable and ethical. 🔸Impact: Encourages collaboration and risk-taking, knowing ethical considerations are part of the process. 4. Collaboration Fuels Innovation 🔹Challenge: Innovation often stalls when teams operate in silos. 🔹ISO42001 Solution: Defines clear roles and responsibilities, enabling cross-functional alignment. 🔸Impact: Teams work together more effectively, addressing risks early and accelerating progress. ➡️ AIMS as a Platform for Innovation ISO42001 creates the environment where AI innovation thrives. By integrating ethical considerations, risk management, and lifecycle monitoring, you can scale your AI solutions responsibly while fostering creativity. 🔹Example: AIMS ensures challenges like bias or transparency are proactively addressed, allowing developers to focus on building impactful AI systems. 🔸Long-term Value: Innovations are not just scalable but also aligned with societal and organizational goals. ➡️ Rethinking Compliance Governance/Management frameworks like ISO42001 are not roadblocks, they are opportunities. They establish trust, reduce uncertainty, and provide the structure you need to innovate responsibly. 🔸Key Takeaway: Success in AI isn’t defined by how quickly systems are built, but by how effectively they deliver ethical, sustainable value. A-LIGN #TheBusinessofCompliance #ComplianceAlignedtoYou ISO/IEC Artificial Intelligence (AI)

The AI org chart in post-sale is flipping upside down, and many leaders are about to become the problem. What I am seeing across multiple companies is that the most meaningful AI innovation is not coming from strategy decks or centralized AI teams. It is coming from what I call "Frontline Innovators". Over the last few months, I have spoken with many leaders who all described the same dynamic. Front-line team members are building AI-powered agents and workflows that are already running in production and materially changing how work gets done. In one case, a Frontline Innovator CSM created an automated weekly customer health digest that pulls signals from CS platforms, support interactions, and call notes, then delivers an executive-ready summary into Slack highlighting risk, expansion opportunities, and operational bottlenecks. At another company, a Frontline Innovator implementation consultant built a workflow that automatically creates a personalized onboarding plan the moment a deal closes. In a third example, a Frontline Innovator eliminated manual account handoffs by generating transition documents from Sales to Implementation and between CSMs using real CRM data and call transcripts. This is what the inverted AI org chart looks like in practice. Historically, innovation flowed top-down. Leaders defined the strategy, selected the tools, and directed execution. With AI, innovation is increasingly bottom-up. People closest to customers and the day-to-day friction see opportunities first and move quickly to improve those processes. They are not waiting for permission or formal resourcing. They are just doing it. This is a wake-up call for leaders. If leadership does not step up, teams will not stop innovating. Instead, leadership risks becoming the bottleneck. In the worst case, leaders become a barrier rather than an enabler of progress. Leaders need to: - Know who their Frontline Innovators are - Give them time, space, and permission to experiment - Put clear but lightweight guardrails in place - Actively surface, validate, and scale what works - Role model behaviors that encourage AI-fluency AI leadership in post-sale is no longer about having all the answers. It is about creating the conditions where the best answers can emerge. Who are the Frontline Innovators in your organization? If they are building, automating, or experimenting with AI, recognize them here. Show them that you see them and appreciate their innovation.

To build a solid Change Management Framework for AI Transformation, enterprises must go beyond technology adoption and address the people and process side of change. AI introduces new ways of working, decision-making, and collaboration, requiring deliberate planning to ensure successful adoption, sustained engagement, and measurable impact. Here are the main components of a robust AI-focused Change Management framework: 🔷 1. Organizational Readiness & Impact Assessment AI Maturity Assessment: Evaluate current capabilities across people, data, and systems. Change Impact Analysis: Identify how AI will affect roles, workflows, and decision rights. Readiness Mapping: Segment the organization by readiness levels and tailor interventions accordingly. 🔷 2. Stakeholder Engagement & Alignment Executive Alignment: Ensure leadership champions the change and visibly supports it. Middle Management Enablement: Equip managers with the knowledge and tools to lead their teams through the change. End-User Involvement: Involve frontline users early to co-design workflows and increase adoption. 🔷 3. Process Reengineering & Role Redefinition AI-Augmented Process Design: Redesign tasks and workflows to integrate human-machine collaboration. Job Role Evolution: Clarify how roles change (e.g., oversight, validation, decision support). Governance Embedding: Update SOPs, risk controls, and approval workflows for AI-infused operations. 🔷 4. Communication & Education Strategy Change Narrative: Define and share a compelling story—why AI, why now, and what’s in it for each role. Multi-Channel Communication Plan: Use town halls, demos, and internal platforms to reinforce messages. Myth Busting & FAQs: Address fear and uncertainty (e.g., “AI will replace me”) with transparent answers. 🔷 5. Training, Upskilling & Support Role-Specific Training: Tailor content for business users, analysts, and technical teams. AI Literacy Programs: Provide foundational understanding of AI concepts, risks, and limitations. Just-in-Time Learning: Embed help and guidance within new tools and workflows. 🔷 6. Adoption Tracking & Feedback Loops Adoption KPIs: Monitor usage, satisfaction, process adherence, and business impact. Feedback Mechanisms: Create forums and channels to capture real-time user feedback. Change Iteration: Use insights to refine tools, workflows, and communications. 🔷 7. Cultural Integration & Long-Term Reinforcement Celebrate Quick Wins: Showcase early success stories to build momentum. Align Incentives: Adjust performance metrics and rewards to reinforce new behaviors. Embed into Culture: Integrate AI adoption into values, rituals, and leadership routines. 💡 In every AI transformation I’ve been part of, one thing has remained constant: If people don’t engage, the transformation doesn’t stick. #AITransformation #ChangeManagement #DigitalTransformation #ArtificialIntelligence

GTM Systems teams CANNOT be order-takers in the AI era. Innovation won’t come from requirements - it will come from experiments. And these teams must evolve into true Product orgs. I see this as the most overlooked challenge with adopting AI in GTM Systems. A successful strategy means you need to move FAST. Experiment. Prototype. Iterate. This is the default standard in Product culture. The Problem: this approach runs counter to Biz Tech culture. Salesforce & Internal Tools experts will hear this and say I’m crazy. “You need strict governance & careful planning to scale systems infrastructure.” And previously, I would completely agree. But the AI era is a different beast for a few reasons. 1) Teams don’t know what’s possible or what they want from AI. • Success is judged by behavior change, not completion of a backlog item. • The value of AI will emerges through usage and iteration • New features will not result from traditional requirements gathering. 2) AI has completely shifted the delivery timetable. • Historically, the goal is to craft a long-term GTM Systems roadmap. • Then, you break key initiatives into months long implementation cycles. • But AI innovation is moving too fast to only ship 1x in 3 months. • Companies need to adopt a rapid experimentation mindset. 3) You CAN move fast by investing in composability. • An API-first approach allows you to ship outside core infrastructure. • Previously, all new feature build happened in tools like Salesforce. • You’re constrained by technical debt, dependencies, and more. • Now, you can deploy AI solutions in isolation. • An app that communicates to other systems via API is relatively low risk. Realistically, this approach will make most Biz Tech teams uncomfortable. Rapid experimentation historically led directly to scalability issues. But this is the default way of operating for core Product teams. A few ways they get it right without leaving a wake of technical debt: 1) Use MVPs with clear scope • Ship measurable slices of value to learn, not solve a whole problem up front. 2) Invest in composability • Every test is built with future modularity in mind - winning ideas can scale. 3) Leverage Users for Research • Stakeholders & Users are a source of insights, not requests. • It’s the old Henry Ford quote: “If I asked people what they wanted, they would have said faster horses.” 4) Document Assumptions • Experiments have clear hypotheses - learn from every test, even if it fails. GTM Systems teams have the opportunity to lead innovation like never before. AI is delivering the much-needed attention and investment in this function. And for the first time, they are less constrained by stakeholder requests. These teams can finally DRIVE strategy, not just support it. But success will depend on their ability to embrace this new approach. __ #AI #GTM #CRM

Is the future of AI management more decentralized than we think? I've been wondering for some time if a new kind of role could emerge in the business world: the AI Manager. I usually don’t like corporate title, but this function could reshape how we approach and leverage AI at both personal and organizational levels. Here are 6 points on why I think this role is interesting to explore: 1. We've all see it, AI have hallucinations. An AI Manager would be our first line of defense, ensuring AI outputs align with reality and business needs. 2. This role would require a deep understanding of human psychology, business domains, and AI capabilities to effectively translate between human intent and machine execution. 3. As ontologies and knowledge graphs become the way to link AI models to business data, AI Managers will need to be especially resilient at structuring and maintaining these knowledge frameworks. 4. A background in AI engineering, data engineering, and software development would be crucial for optimizing AI systems and integrating them into existing infrastructures. 5. This role could augment millions of people’s job to manage specialized AI systems, tailoring them to specific needs or aggregating them for enterprise-level solutions. 6. This decentralized model of AI management stands in contrast to the pursuit of AGI, focusing instead on practical, targeted applications of AI. What do you think? Are we heading towards a future where AI Managers become as common and necessary as project or product managers are today? This decentralized approach to AI management ultimately represents a 'mixed martial arts' philosophy for data: blending diverse skills into an adaptive, cross-disciplinary managing style, much like my friend Joe Reis champions.

A recent straw poll of HBS alumni—spanning five generations—revealed something fascinating: most leaders are AI curious... but also feel their companies are somewhat or completely unprepared to take action. That’s happening while AI innovation is growing exponentially—faster than any other tech wave in modern history (see second chart). Clearly the cost of inaction is exponential. So how can business leaders bridge the gap? After compiling thoughts from dozens of convos with operators, execs and HBS professors this past week, the answer is surprisingly simple to start—and increasingly complex to scale. Here’s the 3-step playbook we heard again and again: 1️⃣ 𝗔𝘀𝗸 𝘁𝗵𝗶𝘀 𝗢𝗡𝗘 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻 𝗲𝘃𝗲𝗿𝘆 𝗺𝗼𝗿𝗻𝗶𝗻𝗴: 💡 “What can GPT help me do today?” That’s it. This daily prompt puts you on your own learning flywheel—helping you spot AI use cases organically. Depending on where you are in the learning cycle, push yourself to go beyond basic utilitarian use cases (e.g., drafting emails) to using AI as a thought partner (e.g., deep dive research on market expansion, managing board of directors or key employees). 2️⃣ 𝗥𝗮𝗹𝗹𝘆 𝗮 𝗺𝗶𝗻𝗶 𝗔𝗜 𝗦𝗪𝗔𝗧 𝘁𝗲𝗮𝗺. Pick a small group with whom you can measure progress and hold accountable. Get them experimenting too. Share what’s working. Let them report back with practical, cross-functional use cases. Think of this as a small cyborg team with real and measurable ROI, executing at a micro scale of the org where the cost of failure is minimal with substantial upside from insights created throughout the learning process. 3️⃣ 𝗜𝗱𝗲𝗻𝘁𝗶𝗳𝘆 𝗵𝗶𝗴𝗵 𝗥𝗢𝗜 𝗔𝗜 𝘂𝘀𝗲 𝗰𝗮𝘀𝗲𝘀 𝗮𝗻𝗱 𝗽𝗿𝗲𝗽𝗮𝗿𝗲 𝗳𝗼𝗿 𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗮𝗹 𝗰𝗵𝗮𝗻𝗴𝗲. Org-wide AI adoption is where things get tricky. What works for a 3-person team may break at scale. Tech is evolving so quickly that today’s best tools may be obsolete in 6–9 months. The key? 🌀 Organizational and technological agility. If you’re too rigid, you won’t survive. Change management isn’t optional—it’s existential. ⚠️ If you're a business leader, the first two steps aren’t optional—they’re table stakes. They’ll give you conviction from the top-down, and momentum from the bottom-up, in order to drive organizational change. The AI wave is here. Ride it—or risk being swept away. #AI #Leadership #DigitalTransformation #FutureOfWork #ChangeManagement #HBS #BusinessStrategy