The companies getting financial returns from AI redesigned their workflows around it.
McKinsey surveyed 1,719 executives across 97 countries in August. 6% of their companies credit AI with 5% or more of EBIT and call the impact significant. KeyBank’s August report found 54% of middle market companies rank AI and technology expansion as their top capital investment priority for 2026.
Nearly three-quarters of that 6% fundamentally redesigned their workflows around AI. Among everyone else, one quarter did.
Company size shows up in deployment. At organizations below $1 billion in revenue, the share that are scaling AI agents held flat at 22% this year. At larger firms it climbed from 27% to 40%. Your larger competitors will keep widening that gap on tooling. Redesigning a workflow is the move available to you this quarter.
The redesign is what my podcast conversations keep circling back to. Five strategies came up again and again.
1. Scaling “Digital Workers” to Automate Back-Office Complexity
Executives defend margins by deploying specialized “digital workers” to absorb administrative labor. They build automated loops around the high-volume, low-margin workflows that choke growth.
“We’ve rolled out over 25,000 digital workers to this industry… Customers using our platform are tripling, literally tripling their free cash flow or their profit margins within four months.”
Fred Voccola, Chairman and CEO of Simpro Group
The method repeats across functions. Leaders break a workflow down to its leading indicators, then automate the coordination between the steps. Custom platforms now handle scheduling, invoicing and call routing. One firm walked up the stack and cut investment memo drafting from thirty hours to four. Margins expand without a matching rise in headcount, which is the whole point.
2. Building Sandboxed Internal Agents and Co-Pilots
Leaders bypass public AI tools and build secure internal agents trained on proprietary data.
“We created our own AI agent called Jarvis. It’s become a one-stop shop for our salespeople… All of this has exponentially increased productivity.”
Ankur Ahlowalia, CEO of Chronus
These co-pilots compress software release cycles from weeks to days. The sandbox keeps proprietary databases and customer records out of public models. Ask yourself where your customer data goes when your sales team pastes a client email into a chatbot. Build the sandbox before you scale the tool, or you will spend next year auditing what already left the building.
3. Cultivating a “Human-in-the-Loop” Empowerment Strategy
Scaling operators reject pure automation. They position AI as a sidecar that handles the manual labor so people focus on trust, relationships and advice.
“When you’re approaching creating a product with the angle of empowerment versus replacement, it fundamentally changes how the product is being designed.”
Vic Ye, CEO of Cara
Frontline workers who believe AI will make their job easier will bring you automation ideas. Give them a channel to submit those ideas and a name attached to each one. In relationship-driven sectors, the route to a human stays visible in every workflow these leaders build, because clients who fail to find it leave. Automating the paperwork moves professionals from task processing into high-value advice.
4. Accelerating High-Volume Execution and Strategy
In creative operations, executives use generative AI to run high-volume campaigns and deploy tactical assets at a pace no human team will match.
“Volume really is winning right now… we’ve built out AI to automatically launch creatives… Now media buyers can work on five times the number of accounts. They’re much happier.”
Jonathan Snow, Chief Innovation Officer at Avenue Z
The same approach reaches the executive suite. CEOs now use AI as a strategic thought partner to red-team major decisions. They feed the plan to a model, ask it to argue the opposing case, then rewrite before the board sees it. One global subscription company planned a 9,000-person international convention in Belgrade this way and skipped a mid-six-figure meeting planner entirely.
5. Transitioning Hiring Frameworks Toward “AI Fluency”
Talent strategy is shifting toward screening for AI fluency across technical and non-technical roles.
“None of our customers are hiring an engineer unless they are AI fluent or trying to get there… And we’re seeing this trend of not just engineering teams hiring engineers, but marketing and operations teams hiring engineers.”
Dylan Serota, CEO of Terminal
Software development costs keep falling. Non-technical departments now hire their own engineers to build custom workflows and bypass the IT queue. Nearly a third of the companies McKinsey surveyed walked away from a software purchase this year because they could build the feature in-house with coding agents. Ask your next marketing hire how they use AI to do the job.
Small business CEOs report a cost on the other side of this. AI-generated applications flood their inboxes, and vetting authentic candidates takes more time than it did a year ago.
Conclusion
These executives rebuilt how work moves through the building before they bought anything. They started with the workflow, then picked the tool.
Your larger competitors nearly doubled their agent deployment this year. Smaller companies did not move.
Pick one workflow this quarter. Map it down to its leading indicators. Redesign it around AI, then measure the margin before and after. That number will tell you whether the redesign is working, and it will tell you before you commit next year’s budget.
I interview senior leaders on the latest in leadership and AI on my podcast, The Scaling Executive Podcast. If you are a senior operator ready to share what you are discovering, come on as a guest. See if we’re a fit.
