Something weird happened on a client call last week.
I was sitting there, listening to a very smart agency owner explain his new AI lead generation service. He was brilliant. He knew every LLM model, every scraping tool, and every API integration under the sun.
The problem was, the prospect on the other end of the zoom looked like he’d been hit with a tranquilizer dart.
His eyes had glazed over about four minutes into the talk about "vector databases" and "autonomous agents." He didn't want a "transformative AI solution." He wanted his sales team to stop wasting four hours a day manually formatting spreadsheets.
If you’re out there selling AI services right now, you’ve probably noticed that the honeymoon phase is over.
People aren't buying the "magic" anymore. They’re buying results. If your sales process feels like a struggle, it might be because you're still selling the rocket ship instead of the destination.
The problem with selling "AI implementation"
Setting out to sell "AI" is like trying to sell "electricity." It's too big, too vague, and honestly, a bit exhausting for the person on the receiving end.
I've been noticing a pattern with the agencies that seem happiest and most profitable. They almost never lead with the tech. In fact, they barely mention which models they’re using until the contract is signed.
When you sell "AI," you invite a thousand questions about security, ethics, and job replacement. When you sell "a way to respond to every lead in 30 seconds," you invite a conversation about growth.
If you want to improve your sales approach today, try this. Look at your marketing materials and delete every instance of the word "harness." Then, replace your technical feature list with a list of the annoying, manual tasks your client hates doing.
How to find high-quality AI leads without the hype
Most lead generation for AI services is currently a race to the bottom. I see so many consultants sending cold emails that look like they were written by, well, a very bad AI.
"I can help you build a custom GPT!"
Cool. So can a teenager with a YouTube account.
The bit most people miss is that high-quality leads aren't looking for a builder. They are looking for a translator. They have a business problem, and they've heard AI might fix it, but they don't know how.
Instead of broad outreach, I've found that specific, "boring" niches work best.
Instead of targeting "small businesses," target "boutique law firms that struggle with document intake." The pitch isn't about AI. The pitch is about the three hours of billable time they’re losing every single day.
If you want to see how we handle the heavy lifting of finding these opportunities, you should check out SalesM8. It’s built to actually help you find people with real problems, rather than just blasting the internet.
Building an AI sales process that actually converts
I keep coming back to the idea that the best sales process is just a series of small, honest realisations.
In my experience, the "Discovery Call" is where most AI agencies die. They treat it like a technical requirements gathering session.
"How many tokens do you think you'll use?" "Do you want a chatbot or an internal tool?"
Wrong. Those are "how" questions. You need to stay in the "why" for much longer than you think is comfortable.
Here is a simpler way to think about your sales flow:
- The Ghost Pain: Ask them about the work that gets done after 6 PM. That’s usually where the automation opportunities are hiding.
- The "What If": Paint a picture of the world where that 6 PM work is done by 2 PM. Don't mention the tech yet. Just let the idea of free time sit there.
- The Prototype: Don't send a 20-page proposal. Send a 60-second Loom video of a "ugly" prototype that solves one tiny part of their problem.
- The Pilot: Sell a small, fixed-price project before you try to sell a five-figure retainer. It builds trust faster than any slide deck ever could.
This approach works because it feels human. In a world where everyone is trying to automate the relationship, being the person who actually listens is a massive competitive advantage.
Common questions about selling AI services
I get asked a lot of questions about how to price and position these services. Here are a few that keep popping up:
Do I need to be a developer to sell AI consulting? No, but you need to be a great plumber. You need to understand how data flows from point A to point B. The "selling" part is understanding the business logic, not the Python code.
How do I handle the "AI will take our jobs" objection? I usually tell clients that AI won't take their jobs, but a competitor using AI might. It’s not about replacement; it’s about giving their best people their brains back.
Should I charge per hour or per project? Avoid charging by the hour at all costs. If you build a solution in two hours that saves a company $50,000, charging for two hours of work is a crime against your own bank account. Charge based on the value of the problem you're solving.
If you are looking for more articles on AI or want to dive deeper into these frameworks, I've written extensively about moving away from the "implementation" trap.
Stop overcomplicating the pitch
Let me share what worked for me when I stopped trying to sound like a genius.
I started telling prospects: "I don't know if AI can fix everything in your business, but I bet we can fix this one specific headache by Tuesday."
That's it. No "digital transformations." No "paradigm shifts." Just a promise to fix one specific headache.
The crazy part is that once you fix the first headache, they’ll practically beg you to look at the rest of their business. That’s how you move from being a "vendor" to being a partner.
If you're feeling stuck on how to position your specific service, feel free to book a consultation. I’m always happy to look at a pitch and help you find the "hidden" value you're probably ignoring.
At the end of the day (sorry, I couldn't help myself), selling AI is just selling. It’s about people, their frustrations, and the hope that tomorrow might be slightly less annoying than today.
Keep it simple. Be helpful. And for the love of everything, stop talking about tokens on the first date.
If you're looking for a smarter way to manage your sales pipeline, check out SalesM8 — it's the tool we built to make AI-powered selling actually practical.
About the Author
Steven Tann is an AI consultant, author of "You're Selling AI Wrong", and founder of SalesM8. He writes about AI, sales, and running a business from a narrowboat on the English canals. Connect with him at steventann.com.
