[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"scribble-the-ai-roi-measurement-problem-b5c832c5":3},{"id":4,"title":5,"user_id":6,"is_anonymous":7,"tags":8,"created_at":15,"updated_at":15,"storage_path":16,"is_public":17,"linked_scribbles":18,"previous_scribble":19,"next_scribble":19,"is_draft":7,"related_scribbles":20,"slug":21,"author_name":22,"author_username":23,"body":24,"linked_articles":25,"related_articles":26,"reverse_relation_map":67},"b5c832c5-c537-46f9-aedf-62a4aaf12841","The AI ROI Measurement Problem","ff91778b-54c7-4d14-8f0c-9652e6b5021a",false,[9,10,11,12,13,14],"artificial intelligence","AI","AI ROI measurement","AI metrics","business value","enterprise AI","2025-12-31T06:52:02.594988+00:00","ff91778b-54c7-4d14-8f0c-9652e6b5021a/ec9b7808-a2fc-4ffa-8484-947fc2bc9aa3.md",true,[],null,[],"the-ai-roi-measurement-problem-b5c832c5","Itih","itih","I was reading about ROI measurement. The stat that stopped me: only 31 percent of business leaders think they'll be able to evaluate AI ROI within six months.\n\nThat's it. Three out of ten.\n\nWhich means seven out of ten companies are building AI initiatives with zero visibility on whether it's working. Not a long timeline. Not \"we'll measure in two years.\" Right now. Today. They don't know.\n\nWhy is this happening?\n\nThe problem is that AI benefits are weird. They're indirect. They're long-tail.\n\nLike, let's say you implement an AI chatbot for customer support. How do you measure the ROI?\n\nYou could measure reduction in response time. That's easy to measure.\n\nBut what about customer satisfaction improvement? Did satisfaction increase because of the chatbot, or because you hired better support staff? Hard to separate.\n\nWhat about the fact that the chatbot is learning from customer interactions, making future support better? That's value, but it's happening slowly. Over months. Over years maybe.\n\nWhat about the fact that your support team is less burned out because the chatbot handles routine tickets? That improves retention. That improves morale. That probably improves the quality of support overall. But how do you measure it?\n\nThis is why only 31 percent of leaders think they'll know within six months. Because the honest answer is: \"It's complicated.\"\n\nBut companies need answers. CFOs need to justify budgets. Boards need to see ROI. So what actually happens?\n\nCompanies make up metrics. They call things \"transformation\" without measuring them. They announce AI projects as wins even though they're still in pilot mode. They measure proxy metrics (like \"we deployed 5 AI use cases\") instead of actual impact metrics.\n\nThis is dangerous because it means the good AI investments are being funded the same way as the bad ones. The winners look the same as the losers on a spreadsheet.\n\nI think 2026 is when the reckoning happens. When companies stop being able to hide behind \"we're still measuring\" and actually have to prove value or kill the projects.\n\nThe companies that figure out ROI measurement first will have a huge advantage. They'll know which AI projects to double down on. Which ones to kill. Which ones to scale.\n\nEveryone else will be throwing money at AI and hoping it works out.",[],[27,31,35,39,43,47,51,55,59,63],{"id":28,"title":29,"slug":30},"7ed59681-19e9-47f4-8060-9e582da41b50","Companies Deployed AI Without Checking If It Works","companies-deployed-ai-without-checking-if-it-works-7ed59681",{"id":32,"title":33,"slug":34},"360b06c6-998e-4d43-9d6b-8d349428268b","Companies Can't Keep Up with AI Model Upgrades","companies-can-t-keep-up-with-ai-model-upgrades-360b06c6",{"id":36,"title":37,"slug":38},"362c71c9-30f7-4712-a038-96a688a401fc","Developers Prefer Claude for Coding","developers-prefer-claude-for-coding-362c71c9",{"id":40,"title":41,"slug":42},"ff32877c-fcb9-4e2b-97c4-25971a2c5665","Fewer AI Graduates Exist Than Companies Actually Need","fewer-ai-graduates-exist-than-companies-actually-need-ff32877c",{"id":44,"title":45,"slug":46},"de906776-1cb7-44b4-b981-b09ec5ee1e96","$244 Billion in AI Spending Can't Be Measured","-244-billion-in-ai-spending-can-t-be-measured-de906776",{"id":48,"title":49,"slug":50},"cce057da-947a-44db-a8af-dee0f295ec88","ChatGPT vs Claude vs Gemini 2025: The Competitive Shift","chatgpt-vs-claude-vs-gemini-2025-the-competitive-shift-cce057da",{"id":52,"title":53,"slug":54},"8ffb2e4a-0e0f-4e6c-9d1c-dfbadd30f05d","AI Engineer Shortage Cause Companies Can't Hire","ai-engineer-shortage-cause-companies-can-t-hire-8ffb2e4a",{"id":56,"title":57,"slug":58},"104ff8bb-d865-4eef-8380-63ea42b2f2d5","In AI, being cheaper will not help to Win","in-ai-being-cheaper-will-not-help-to-win-104ff8bb",{"id":60,"title":61,"slug":62},"0ff283a8-c0f6-43c4-8390-50a86abcdd30","Most of the AI Projects Never Leave Pilot Mode","most-of-the-ai-projects-never-leave-pilot-mode-0ff283a8",{"id":64,"title":65,"slug":66},"b18ab357-b4c2-4d4e-b23d-395a2c3d0100","The AI Adoption Gap","the-ai-adoption-gap-b18ab357",{}]