Digiday: AI use outpaces skills

Marketing teams are buying AI faster than they’re learning how to use it. That’s the real story here. Digiday’s latest marketer survey shows adoption still climbing hard, but the human side — training, technical skills, operating discipline — is not keeping pace. So the news is not just “more AI.” It’s that the gap between tool rollout and actual capability is getting wider. ### What actually moved? The biggest shift is simple: AI investment among marketers kept rising in Digiday’s survey, reaching 86% of companies in 2025. That is up from 71% in 2024, 57% in 2023, and 44% in 2022. In other words, this is no longer early-adopter behavior or a pilot-program phase. It’s a mainstream budget line. ### Why is that a bigger deal than it sounds? Because adoption numbers can hide a mess underneath. A company can say it is “investing in AI” when one team uses a chatbot for copy drafts, or when the whole marketing org is rebuilding workflows around automation. Those are where teams scramble to figure out training, governance, and who actually knows enough to run them well. ### Where is the gap showing up? Mostly in skills. Marketers are clearly willing to plug AI into loyalty programs, chatbots, search work, media planning, and creative production. But knowing how to prompt a model is not the same thing as knowing how to evaluate outputs, manage risk, and it tends to lag because it takes process change, not just software spend. ### Why does training lag so badly? Because training is slower, less glamorous, and harder to measure than procurement. Buying software looks like progress. Teaching hundreds of employees how to use it safely and consistently feels like overhead — until something breaks. There’s also a moving-target problem. AI products keep changing models that may look different in six months. That creates a weird middle state where everyone is using AI, but not many people feel fully fluent. ### What does that mean for vendors? It favors products that do more of the thinking for the customer. If buyers do not have deep in-house AI talent, they will lean toward tools with strong defaults, clear workflows, tight onboarding, and built-in guardrails. Basically, the less a customer has to architect from scratch, the better. A flexible platform sounds powerful, but if the customer lacks technical depth, flexibility can just mean more ways to fail. ### Why is governance part of this story? Because once AI moves from experimentation into live marketing workflows, mistakes get expensive fast. Bad outputs can hit customers, brand voice can drift, compliance issues can surface, and teams can lose track of what was machine-generated versus human-reviewed. The capability gap is not just about productivity. It’s about control. If adoption rises faster than governance, the organization gets speed without reliability. ### Is this just a marketer problem? Not really. Marketing is just a clean place to see it. The same pattern is showing up across a lot of white-collar work — software spreads faster than institutions learn how to absorb it. But marketing feels it early because the function sits close to content, media spend, customer interaction, and brand risk all at once. ### So what’s the bottom line? AI adoption is becoming ordinary. AI competence is not. The companies that win from here probably won’t be the ones using the most tools. They’ll be the ones that close the gap between access and ability — and make AI usable by normal teams, not just power users.