#15: How AI Changes Sales

Hey friends, and a special welcome to all new subscribers!

I spent this week in Barcelona for a big film and cinema convention, meeting with studios and various industry stakeholders on behalf of the new agency we launched earlier this year (Mojo Works).

I’ll be sharing more about how we’re building this agency from the ground up with an AI-first mindset over the next few months.

For now, let’s move on to this week’s topics: How AI changes service sales and a framework to bring AI into your workday gradually.

PS: feel free to connect with me on LinkedIn and say hi! It’s always fun to connect with new readers.

PS2: Thanks to everyone that took the time to answer the poll last week. I’ll be releasing the first step-by-step playbook (on competitor research) soon.

– Martin

The Idea

How AI Changes Sales

For decades, every service business had the same rule: you can't afford to do actual work before getting paid.

So everyone followed the same playbook—qualify, pitch, negotiate, then deliver.

That constraint just died in many industries.

The New Math

AI is making many types of work basically free. What used to cost $500-2000 in labor now costs $2-5 in processing.

Smart operators figured this out first. Instead of "Here's what I could do for you," they're saying "Here's what I already made for you. Want more?"

Examples everywhere:

• Personal branding consultants deliver full LinkedIn audits before prospects even respond

• Real estate photographers create listing videos from Zillow photos, then cold email agents

• Marketing agencies build competitive analyses from public data

• Interior designers generate three room concepts from one photo

Approaching new business sales this way creates an interesting dynamic, and a competitive advantage:

Proof beats promises. Prospects aren't imagining your work—they're holding it.

Reciprocity becomes your closer. When someone delivers unexpected value, they feel obligated to return the favor.

Speed creates moats. While competitors schedule discovery calls, you're already delivering results.

This doesn’t mean you should be giving everything away.

Instead, consider how you can build two offers:

Entry offer: Partial but valuable work that costs pennies to produce. This is your marketing.

Core upsell: Full engagement worth 5-10x your entry offer. This is your revenue.

Interior design example:

• Entry: Three AI room concepts (cost: $3, time: 15 minutes)

• Upsell: Full design package + implementation ($15,000)

The Tactic

The Autonomy Slider

Andrej Karpathy, one of the co-founders of OpenAI, recently did a talk at the tech incubator Y Combinator. It’s long and technical, but very good. (You can watch it here).

Anyways, there was one key takeaway I wanted to share with you.

He referred to it as “The autonomy slider”. A useful framework to use when thinking about how to bring AI into various parts of your workflow and workday:

Take any task. Give AI three levels:

Level 1: AI suggests stuff, you pick what to do
Level 2: AI makes a plan, asks permission, then does it
Level 3: AI just handles it, you check the results

Start everything at Level 1. Graduate tasks as you build trust.

An example for content creation could look something like this:

• Month 1: "Give me 5 headline ideas"

• Month 2: "Write this section, I'll approve it first"

• Month 4: "Write the whole post, I'll edit"

• Month 6: "Handle everything, ping me when it's live"

Most people either micromanage AI (pointless) or let it run wild (disaster).

The autonomy slider lets you find the sweet spot for each task, and adjust it over time.

Start paranoid. Get lazy gradually.