A tough topic but an honest one: fear is beginning to set in with a lot of CMOs and CROs I know.
That the new AI world is one that … won’t have a place for them.
And it may not. The tough, honest start-up truth:
- 1% of your team is epic
- 5% of your team is great
- 15% of your team gets sh*t done
- 50% of your team does OK-to-solid work
- 30% of your team adds little value at all
This has honestly always been true. It’s just in 2021, there was so much organic demand for everything in business software, it masked things. And the 50% and midpack could, e.g., still be decent closers.
Not today.
AI is coming for the last two groups. Faster than most want to be honest about.
We’re also creating entirely new jobs. But are they the ones you can, or are willing, to do?
AI needs new leaders:
- Someone needs to orchestrate AI workflows.
- Someone needs to QA AI outputs.
- Someone needs to train AI models on your specific use cases.
And as Balaji pointed out recently: AI prompting scales, because prompting is just typing. But AI verifying doesn’t scale, because verifying AI outputs requires deep human judgment.
Think about it:
- Proof of human – Is this content actually written by a person?
- Proof of authenticity – Is this information real or hallucinated?
- Proof the output is not AI – Does this pass as genuinely human?
- Proof the output is actually right – Is the AI’s analysis accurate for our specific business context?
Every piece of AI-generated content, every automated decision, every AI recommendation will need human verification. That’s not a task you can outsource to more AI – it requires domain expertise, business judgment, and contextual understanding.
The question isn’t whether jobs will disappear – it’s whether your current team can evolve into these new roles or if you’ll need to hire fresh talent who already thinks this way.
Here’s What I’m Seeing in the Field
I was talking to a CMO at a $50M ARR company last week. Built great teams. But he’s losing sleep because he just realized that 60% of his marketing ops, content creation, and campaign execution can now be done by AI tools. And he doesn’t really understand them himself.
His exact words: “Jason, I have 12 people doing work that Claude and a few automation tools could handle. What the hell do I do with that?”
The Math Gets Brutal Fast
Let’s say you’re paying that middle 80% an average of $75K each (loaded cost ~$100K). On a 50-person go-to-market team, that’s $4M annually for work that AI can now do for maybe $50K in software costs.
Your board is going to figure this out. Your investors already have. Your competitors definitely will.
The Thing Everyone’s Missing
The companies that win won’t just fire their way to efficiency. They’ll transform their way to dominance.
That 50% doing “OK-to-solid work”? Some of them can level up. Give them the AI tools, eliminate the grunt work, and some will level up. They’ll improve the depth of customer relationships, they’ll build more pipe and close more deals.
The rest? Replace them. You have no choice in the end.
The real opportunity: While your competitors are cutting headcount, you’re creating super-powered teams.
What They Say They Are Already Doing
Atlassian says they’re not laying off their content team – they’re retraining them to be AI prompt engineers and strategic storytellers. But can everyone really be a top tier prompt engineer?
HubSpot’s sales team uses AI for all prospect research and initial outreach. But instead of cutting SDRs, they moved them into consultative selling roles. But can most SDRs really become deep product experts? And do they even want to be?
Salesforce is working to reskill 1000s of customer support reps into sales positions. But … will this really work? Can an entry level support rep learn to sell complex enterprise analytics software?
Can Folks Really Be Reskilled? Benioff Says “Yes” But It’s Not Clear Folks Can or Want to Learn to Be AI Natives
Marc Benioff keeps talking about reskilling at scale. Salesforce is spending $500M on it. Sounds great in theory.
But here’s what I’m seeing in practice: It’s way harder than anyone wants to admit.
I invested in a Series B company that tried to retrain their entire demand gen team on AI tools. Six months later? Only 2 out of 8 people were still there. Because only 2 out of 8 adopted them consistently. The rest found workarounds to keep doing things the old way.
The brutal truth: Some people just can’t make the mental shift.
The Three Types of People You’ll Discover
The AI Natives (20%): They pick up AI tools instantly. They’re already experimenting with ChatGPT, Midjourney, and automation workflows on weekends. These folks will 10x their output.
The Reluctant Adopters (50%): They’ll use AI tools if forced to, but they’re not innovating with them. They see AI as something that threatens their job security rather than amplifies their capabilities.
The Resisters (30%): They actively avoid AI tools. They have philosophical objections, or they’re just overwhelmed by the pace of change. These are your biggest retention risk.
What Benioff Gets Right (and Where He’s Being Too Gentle)
Benioff is right that reskilling is possible.
But he’s wrong about the timeline and success rate. Most reskilling programs fail. Not because the technology isn’t there, but because people fundamentally resist changing how they work. And reskilling for AI has to happen faster, with bigger results, than prior transitions.
I’ve seen marketing managers who spent 15 years perfecting Excel workflows refuse to learn no-code automation tools that would save them 10 hours a week. The tools aren’t the problem. The mindset is.
The Uncomfortable Question
Are you prepared to make tough decisions about people who can’t or won’t adapt?
Because that’s what this comes down to. You can provide all the training in the world, but if someone fundamentally resists becoming AI-native, they’ll become a drag on your entire team’s productivity.
The Warning Signs Are Already There
Think about it: The Sales Rep That Never Learns the Product. You know the one. They’ve been at your company for 18 months and still can’t demo the core features without fumbling. They rely on SEs for everything. They think “product knowledge” means knowing which slide deck to send.
The Customer Success Manager That Just Tries to Upsell You. Every interaction is a ham-fisted attempt to expand the account. They don’t understand the customer’s business. They can’t articulate value beyond “you should buy more seats.”
The Director of Marketing That Just Wants to Hire a Bunch of Agencies to Do Campaigns. They outsource everything – creative, strategy, execution. They’re basically an expensive procurement person with a marketing title.
Here’s the thing: These people already don’t add much value. They’re coasting on process, not insight. They’re order-takers, not value-creators.
AI is going to expose them completely.
When ChatGPT can research prospects better than your lazy sales rep, when AI can analyze customer health scores more accurately than your CSM, when automated campaigns outperform your agency-dependent marketing director… what exactly are you paying them for?
The uncomfortable truth: If you wouldn’t trust them to learn AI tools, you probably shouldn’t trust them with their current responsibilities either.
The $240K Question: What Are We Actually Paying For?
Let’s get brutally specific about the math here.
Do we really need $160K OTE sales reps that only call back 30% of the leads? Do we need $80K SDRs that just send spray-and-pray emails?
I know companies spending $2.4M annually on a 10-person sales team where half the reps can’t be bothered to follow up on warm inbound leads within 24 hours. Meanwhile, AI SDRs are responding to prospects in 60 seconds with personalized research that took humans 2 hours to compile.
The wake-up call: Many of your “he’s OK, we need the body” mid-pack performers are about to be exposed as expensive underperformers.
When AI can qualify leads better, research prospects faster, and follow up more consistently than your current team… what exactly are you paying them to do?
This isn’t about being heartless. It’s about being honest. If someone’s primary value is executing tasks that AI now does better, cheaper, and faster, then we need to have a different conversation about their role.
The Weekly AI Test: Your Secret Performance Review
Here’s what I tell every CEO I advise: Bring in a new AI tool each week if you can, each month at a minimum. And see who rises to the occasion and leverages it. And who stares blankly at the Zoom.
AI is evolving at breakneck speed. What took 6 months to build last year now takes 6 days. The tools that seem impossible today will be table stakes next quarter.
I know a VP of Marketing who introduced a different AI tool every Friday for 3 months. Video editing AI. Social media schedulers. Content generators. Email optimization tools.
The results were brutal and revealing:
- 3 people dove in immediately, started experimenting over weekends, came back Monday with creative use cases
- 5 people tried them halfheartedly, used basic features, didn’t see the point
- 4 people literally ignored every new tool introduction
Guess which group got promoted? Guess which group got “reorganized” out?
This isn’t about being tech-savvy. It’s about intellectual curiosity and adaptability. The people who light up when you show them a new capability versus the people who immediately list reasons why it won’t work – that tells you everything about their future value to your company.
The Speed Trap: Why AI Makes Humans Work Harder AND Smarter
Here’s a perfect example of what I mean.
I know a company that deployed AI SDRs that research prospects and send personalized outreach in under 60 seconds. Response rates are 3x higher than their human team ever achieved.
But here’s the catch: When the AI responds to an interested prospect in 60 seconds, the human sales rep can’t wait 4 hours to follow up.
Suddenly, your “solid” sales rep who used to batch all their calendar scheduling at the end of the day? They’re a bottleneck. The AI creates the opportunity, but the human kills it with slow response times.
One rep adapted – set up instant notifications, started responding within 10 minutes. His close rate went through the roof.
Another rep kept doing things the old way. “I’ll get back to them tomorrow.” His pipeline dried up completely.
Same AI tools. Completely different outcomes.
This is the future: AI amplifies both great performance and mediocre performance. The gap between your best and worst performers is about to become a canyon.
The Hard Conversation You Need to Have Today
Sit down with your team. Show them this math. Be honest about what’s coming.
Then ask: “How do we get every person on this team into the top 20%?”
Because here’s the reality – AI isn’t going to replace great salespeople who build relationships. It won’t replace marketers who understand customer psychology. It won’t replace leaders who can navigate complex deals.
But it will absolutely demolish anyone whose primary value is executing tasks a machine can learn.
The 90-Day Action Plan
Month 1: Audit every role. What percentage of their time is spent on tasks AI could handle?
Month 2: Start the transformation. Give your team AI tools. Measure output changes. Identify who adapts quickly.
Month 3: Restructure around human+AI hybrid roles. Double down on your adapters. Have tough conversations with those who can’t make the leap.
Be honest. The bottom 30% of each functional area in your company may have no role in the Age of AI. And in fact, you will probably move faster without them. That’s perhaps the most important point.
Bottom Line
The AI revolution isn’t some distant future thing. It’s happening right now.
Your choice: Lead the transformation or get transformed by it.
The companies that figure this out now and not in 12 months will have an insurmountable advantage. The ones that don’t… well, their talent will be working for the companies that did.
What are you going to do about it?
