🔧 Back to Basics: Operations in the Age of AI

Each year, companies spend billions on consultants searching for the next big thing — reorganize, automate, outsource, innovate. They crave frameworks, benchmarks, and transformation playbooks.

I used to be one of those consultants. We delivered exactly what clients wanted — the glossy reports, the success stories, the new acronyms. But often, what they wanted and what they needed weren’t the same.

📉 The Irony of Innovation

Most of my clients were service companies — banking, insurance, healthcare, retail. They dreamt of cutting-edge solutions but ignored operational fundamentals. They were chasing shiny objects while stepping over gold.

And now, with AI? We’re seeing the same playbook all over again.

🛠️ Old Tools, Still Sharp

I started my career at GE, where “Ops” ran the show. To move up, you had to know how the business made money — and how to make more of it.

Today, operations is sometimes looked down on in the strategy world. But execution still wins games. You can’t out-innovate broken processes.

One principle that never gets old: bottlenecks and variability matter more than most realize.

🚧 Bottlenecks: Fix the Closest Alligator to the Boat

Eliyahu Goldratt’s The Goal (yes, the one every MBA has read) gave us the Theory of Constraints:

1. Identify the bottleneck.
2. Exploit it.
3. Align everything else.
4. Elevate it.
5. Repeat.

In manufacturing, this is second nature. But in services? Not so much.

Case in point: I recently worked with two VC firms looking to apply AI. Both wanted to automate pitch deck creation — a task for Associates.

But the real bottleneck wasn’t Associates — it was the MDs. Only so many deals can move forward if the MD can’t review them fast enough. Automating non-bottlenecks just increases inventory.

📊 Variability: The Hidden Killer in Service

Ever had a terrible customer service experience? Odds are, it wasn’t the average performance that let you down — it was the variability.

At GE, we waged war on variability using Six Sigma. Today, most companies don’t even track standard deviations on their dashboards.

AI has the potential to change that. Early studies (like BCG’s) show AI can dramatically reduce performance gaps between low and high performers — making your workforce more consistent, more reliable.

💡 Final Thought: Use AI to Solve Real Problems

AI is cheap, smart labor. But if you don’t know what work needs to be done, or what’s truly slowing your business down, AI won't save you.