Read Time: 2 minutes. |
| | | | Last Call to Level Up Your Leadership Game | Our top-rated MGMT Accelerator starts February 18 at 11:00 AM ET. | In eight 90-minute sessions over four weeks plus 3 group coaching sessions, you'll unlock your biggest management challenge while building the skills and supporting AI-powered systems. | Perfect for leaders with 3-10 years of experience who want to deliver more impact and level up their leader capabilities. | Don't wait! We won't offer this program again until May. | |
| |
| | |
|
The Question Every CEO Is Asking |
I spent three days this week with 100 CEOs last week. |
|
Topics ranged from developing leaders to finding A-player talent, but no topic was more common than AI. |
Specifically: Is anyone actually getting an ROI from it? |
These are pragmatic business owners who don't want to throw money away chasing the latest fad. But they also don't want their business eroded overnight while competitors figure it out first. |
Here's what I learned. |
|
Can AI Replace Entire Functions? |
The short answer: Not really. Not yet. |
Some CEOs have attempted to replace entire functions, but their results were mixed at best. The few who tried have walked it back for now. |
Why it's not working: AI can handle tasks, not judgment. It can answer questions but can't navigate ambiguity. The error rate was still high enough on what were perceived to be critical tasks that they couldn't afford to keep running the experiment. |
The trap: Most of the CEOs said they had gone too big upfront. IF they had to do it over again, the would have started smaller. Ideally, augmenting people vs trying to replace them. |
|
Where AI Is Actually Delivering ROI |
The real wins are individual, not organizational. |
Engineers coding 6x faster using Cursor and Claude. Managers automating admin work that used to consume hours. Sales teams using AI to prep for calls and draft follow-ups. |
The pattern: No one mentioned layoffs, but several said they had slowed hiring. Teams staying the same size while output grows. |
One CEO's example: "My development team writes 6x the code they did a year ago. Same headcount. Massively more output." |
The highest ROI right now: |
Your own workflow (start here) Employee experiments (let them find the wins) Specific, repeatable tasks (not entire jobs)
|
Hint: The main LLM companies - OpenAI, Anthropic, xAI, etc - are all leaning into their coding capabilities with the hope it'll be able to accelerate the coding of their own models. Despite the doomsday scenarios a recursive AI might suggest, chances are this function will advance fastest, which means early adopters will benefit. |
|
Are Customers Open to AI? |
The answer: It depends. |
Some businesses are seeing great results. Customers prefer not talking to humans if AI has the answers or can solve their problems faster. Inbound is shifting from phone to chat. |
But here's the problem: Many leaders set the standard for AI at near-perfect. |
This is like holding autonomous cars to zero fatalities while humans cause 40,000+ deaths annually. The comparison to perfection slows adoption. |
The better standard: Compare AI to what you're getting now, not to what's theoretically possible. In many cases, your team isn't getting to the work or delivering it at the quality you'd hope. |
|
The Two Standards That Actually Matter |
Standard 1: Speed |
AI types faster than you. It analyzes massive datasets faster than your best analysts. It responds to customers faster than your support team. |
If speed matters in your business, AI wins. |
Standard 2: Unlocking What You're Not Getting To |
One CEO implemented AI call answering at his pizza place. Result: 30% more orders taken (with fewer errors) than human counterparts. |
The win wasn't "better than humans." It was "capturing orders we were missing because humans couldn't answer fast enough." |
|
What the Coders Figured Out First |
Developers are out in front on AI adoption. Why? |
They changed their measure of success. |
They're not asking "Can AI write perfect code?" They're asking "Can AI help me write more high quality code faster that I can then review and refine?" |
The answer is yes. Massively yes. |
The lesson for everyone else: AI isn't replacing your judgment. It's multiplying your output. |
|
The Pragmatic AI Strategy |
Based on what's actually working: |
Phase 1: Personal Productivity (Start Here) |
Use AI for your own workflow first Find 3-5 tasks AI can do faster or better Measure the time savings Share what works with your team
|
Phase 2: Team Experiments |
Let employees experiment with AI tools Create a channel for sharing wins Celebrate speed and output gains Consider letting adoption unfold organically Be mindful of secure data
|
Phase 3: Customer-Facing Applications |
Start with low-stakes interactions Set realistic standards (better than nothing, not better than perfect) Measure what you're unlocking, not just what you're replacing Keep humans in the loop for complex situations
|
|
Your Next Move |
This week, pick one task that takes you 2+ hours and see if AI can cut that in half. |
Don't aim for perfect. Aim for faster.
Don't try to replace yourself. Try to multiply yourself. |
Because the CEOs winning with AI aren't the ones with the biggest budgets or the fanciest tools. They're the ones who changed their standard from "perfect" to "an improvement over what we're getting now." |
And that standard? It's a lot easier to hit. |
Dave |
PS - Our February MGMT Accelerator is 3 days away almost full. Will you join us? |
| | | | Ways To Work With Us | MGMT Accelerator: A live cohort-based leadership development program. MGMT Fundamentals: A two-week training program for new managers. Custom Programs: Workshops built and delivered for your company. 1:1 Executive Coaching: C-suite leaders looking to scale. Keynote speaking: Leadership lessons for your event or offsite.\
| Learn about them all at: davekline.com |
| |
| | |
|
. |
No comments:
Post a Comment