By John Barrows | September 2026
The AI sales skills that actually matter are not prompt engineering or tool selection. They are the judgment skills that have always separated great salespeople from average ones, and that AI is now making impossible to hide. The last 10% of any piece of work, the part where you decide if the output is accurate, if it sounds right, and if it is going to land the way you need it to, is still entirely yours. And it is the only part that matters.
Why outsourcing your thinking to AI is dangerous
Researchers at the MIT Media Lab put EEG caps on 54 students and had them write essays over four months: one group used ChatGPT, one used a search engine, and one wrote without any AI assistance. The ChatGPT group showed the weakest brain connectivity of the three groups, and 83% of them could not quote from the essay they had just written. A separate Microsoft Research and Carnegie Mellon study of 319 knowledge workers found the same pattern: the more people trusted AI to handle a task, the less critical thinking they applied to it.
The problem is not AI. The problem is using AI as an answer engine rather than a thinking partner. Ask a question, get an answer, move on. That is a habit that works until the moment it does not. In sales, it fails in front of a prospect. And by the time you find out it failed, you have already lost the credibility you needed to get back in.
The M&A email trap
Here is a specific example of where this goes wrong. Take AI-generated outreach using a merger and acquisition as a trigger event. AI will write a genuinely good-looking email to a VP of Sales whose company just got acquired. The problem is that “good-looking” is not the same as “accurate.”
If that email lands in the inbox of someone who has actually gone through an acquisition, someone whose job security, team structure, and entire operating reality was suddenly in question, it reads immediately as written by someone who has never been inside that situation. The email might get a meeting. But the meeting will expose the gap. You walk in assuming you understand their world, and they figure out in the first few minutes that you do not. You lose the credibility faster than you built it. The worst outcome is not the email being ignored. It is the email working well enough to get you on the phone, where the lack of context becomes obvious.
The same tool that wrote that email could have built the context that would have made it land. Ask the AI what happens to a VP of Sales at a small company getting acquired by a large one. Ask it what the phases of an acquisition look like and what a leader prioritizes at each stage. Ask it what fears and pressures are real during each of those stages. That conversation gives you enough context to write something that proves you understand the situation before you ever ask for a meeting. That is the difference between AI as a shortcut and AI as a tool for learning faster than experience alone would allow.
Two rooms, the same foundation
Picture two groups of salespeople: a room of students just entering the profession who have grown up with AI as a default tool, and a room of senior reps with years of territory relationships and no particular reason to start using AI. Both groups need the same sales fundamentals. Both need to understand how AI fits into those fundamentals. But they need it explained from opposite directions.
The students are fluent in the tools. What many of them are missing is the context that makes the output trustworthy. They can produce a polished-looking first draft instantly. What they cannot do yet is evaluate whether that draft is any good, because they have not lived through enough of the situations the content is describing. The veterans have the opposite problem. They have the context and judgment that took decades to develop. What some of them have not done yet is figured out how to put AI in front of that judgment so it handles the work they are tired of doing. Once they do, their experience becomes a multiplier rather than a credential. Their context is exactly what makes AI a superpower rather than a liability. When AI makes something up, they catch it. When it gives a generic take, they know enough to push back and ask for a different angle.
What AI does to the 10-60-30 rule
There is a framework worth understanding in sales: 10-60-30. Ten percent of any team will take what they learn and execute at the highest level, because that is who they are. Sixty percent will do something different because it makes sense and is within reach. Thirty percent will not change anything.
AI has shifted what those numbers mean in practice. The bottom 30% are already losing ground, not because AI replaced them, but because they have refused to adapt to a world that has moved. The top 10% will always be the top 10%, with or without any new technology. The group worth paying close attention to is the 60%, and specifically the bottom half of it. The reps who have been getting by on average execution are now competing with AI tools that can produce average work at volume. That is a problem most of them have not fully understood yet. AI does not replace great salespeople. It exposes the gap between average and great, and it makes that gap visible to everyone around them.
The 30-60-10 workflow for using AI on your work
The version of that ratio that matters most for daily work is inverted: 30-60-10. The first 30% is you. Your context, your understanding of the situation, your knowledge of the audience. The more you put in here, the better everything that follows. The next 60% is AI taking that input and producing an organized, coherent draft or plan. The last 10% is you again, making sure the output sounds like you, says what you actually believe, and is accurate to the specific situation in front of you.
Gary Vaynerchuk describes a similar pattern: he wants to ideate and bring his intuition at the front end, let his team own the middle, and come back in for the final layer. He said AI works the same way for him – strong at the upfront prompt, doing the work in the middle, then needing someone to edit before it goes anywhere. The sequence is the same as 30-60-10. If you skip the 30% of context at the front, the 60% in the middle produces something generic. If you skip the 10% at the end, you are putting work into the world that has not been filtered through the judgment that only you have.
This newsletter started as a long, messy brain dump. Two training sessions in one day, a research study, a Gen X tangent, and a framework from a Gary Vaynerchuk reel. AI found the thread and organized it into something coherent. The last 10% was making sure it sounds like me and says what I actually believe. That is the only part that required something AI does not have.
Curiosity as the superpower
The students who will be dangerous in five years are not the ones who are most fluent with the current set of tools. The tools will change. The ones who will matter are the ones who are curious enough to use AI to build context they do not yet have, rather than to skip the step of building it. Curiosity is the AI sales skill that compounds. The faster you get curious about a buyer’s world, an industry, or a situation you have not lived through, the better your 30% gets. And the better your 30% gets, the better everything that follows it.
If you are early in your career, get curious enough that you earn the right to own the last 10%. If you have been doing this for decades, start your 30% this week, because your last 10% is worth more than you think.
Frequently asked questions
What are AI sales skills and why do they matter now?
AI sales skills are the combination of prompt fluency, contextual judgment, and critical evaluation that allow a salesperson to use AI effectively without replacing their own thinking with it. They matter now because AI can handle the research, the structure, and the first draft of almost anything a rep needs to produce. What it cannot do is evaluate whether the output is accurate, relevant to the specific buyer, or actually likely to land the way it needs to. The reps who develop AI sales skills build better input at the front end and better judgment at the back end, which is what separates work that stands out from work that looks like everything else.
How do you use AI in sales without losing critical thinking?
The key is treating AI as a thinking partner rather than an answer engine. Instead of asking AI to write an email and sending it, ask AI to help you understand the buyer’s situation first: what pressures they are under, what their priorities are at this stage, what a person in their role is actually dealing with. That conversation builds the context that makes the output useful. Then use AI to draft. Then edit the draft yourself, specifically looking for places where it is generic, inaccurate, or missing something only you would know. That last step is where your critical thinking goes, and it is not optional.
What is the 30-60-10 framework for using AI?
30-60-10 is a way to think about the human-AI split on any piece of work. The first 30% is the human: context, intent, knowledge of the situation, understanding of the audience. The next 60% is AI: organizing, drafting, structuring, expanding on the input. The last 10% is the human again: final judgment, accuracy check, voice and tone, and making sure the output says what you actually mean and will land the way you need it to. The 10% at the end is the most important and the most commonly skipped. It is also the part AI cannot do for you.
Why does sales experience make AI more valuable, not less?
Experience is what makes AI output trustworthy. A rep with deep experience in a specific industry or buyer type can evaluate AI output against what they know is actually true. They can catch hallucinations, identify generic takes, push back when the framing is wrong, and add the specific context that makes the output relevant rather than plausible. Without that experience, AI produces something that sounds good but may not be. With that experience, AI accelerates the work that the rep already knows how to evaluate. The experienced rep is not competing with AI. They are the quality control that makes AI worth using.
How do you evaluate an AI-generated email before sending it?
Ask yourself three questions before sending any AI-generated outreach. First: does this accurately reflect what this specific person is actually dealing with right now? Generic trigger-event language that could apply to any buyer in that role is a sign you skipped the context step. Second: if this person got on a call with me, could I back up every claim in this email with real knowledge? If the email implies you understand their world but you do not, the meeting will expose that gap. Third: does this sound like something a person wrote, or does it sound like something a tool generated? If the answer is the latter, it is going to read that way to the recipient too.
Keep going
The AI sales skills that hold up – discovery, objection handling, building real trust with buyers – are at the core of Filling the Funnel and Driving to Close. See what teams are building at learn.jbarrows.com/pages/results.
John Barrows helps sales leaders decide whether to replace, rebuild, or retrain their teams for the AI era. For 25+ years he has worked with the world’s most demanding sales organizations, including Salesforce, LinkedIn, Google, Amazon, and Okta, building the frameworks that became Filling the Funnel and Driving to Close. Today he advises CROs and VPs of Sales on AI readiness, team restructuring, and go-to-market strategy, drawing on exposure to every type of B2B sales organization over that span. John believes sales is a science, not a personality contest. His training focuses on the fundamentals that hold up regardless of what the market or the technology does next. He has been featured in Forbes, Harvard Business Review, Inc., Fortune, Entrepreneur, and the Boston Business Journal. He is the host of Make It Happen Mondays, author of I Want to Be in Sales When I Grow Up, and an LP at GTMfund.
