Measuring AI Skill Gaps in GTM Teams: Comparing Self-Reported vs Actual Proficiency

·Aug 11, 2026 01:52 AM(edited)·

curious what others are seeing with AI skill gaps in GTM teams. been looking at assessment data and the self-reported vs actual skill gap is around 18 points on average. most sales reps say they're proficient with AI tools but when you watch them actually use them for prospecting or outreach, big disconnect. anyone measuring this on their teams or just vibes-based?

💡1

7 comments

· Sorted by Oldest

Patrick A.
·
(edited)·
I think there's a huge difference around of how someone uses genAI. I would argue prompt work is just one piece of the puzzle that gets you in the door. So I look for more than just "I can prompt" when talking to others. I will be opening a role soon and have been thinking about this question alot.
Lara D.
·
·
That's a big problem facing today. Companies are using AI to find employees and potential employees are using AI to (hopefully) land a job. Lot of resumes end up with hallucinations on them, whether the hopeful employee intended the padding or not. Something needs to change in the hiring flows, but what?
Patrick A.
·
(edited)·
I think diligence matters; Lara D. I put the onus on the user for that. If someone doesn't want to check their work thats an individuals issue. As it pertains to ATS leveraging Ai - that comes down to understanding what the context actually means when implmenting- you know, Plato. I kind of see throwing on AI for the sake of throwing AI on as the same as putting a spark plug in with out learning to gap it correctly.
Lara D.
·
·
I 100% agree that the applicant owns the accuracy of what they submit. I’m thinking one layer beyond that, though. If self-reported AI proficiency is consistently overstated, whether intentionally or not, relying on resumes and self-assessment won't work. Hiring teams needs a way to validate the applicant can use the specific tools effectively.
Patrick A.
·
(edited)·
What I am saying its on both - hiring teams and applicant. Hiring teams cant honestly expect great applications if the job postings lack context, not check their work either and aren't involved in learning how this AI use case actually works when implemented. Hiring and HR have a plethora of tools and assessments as it is plus the majority of applicants have degrees (that are required). It's reasonable to argue that its not possible to get a good data set if both ends of the equation are flawed. I mean the best way to validate would be to just pick up the phone, no? Another option would be to just have the candidate show you on a call (or in person) how they use it.
Lara D.
·
·
Patrick A. - If you were opening that role tomorrow, what would really give you confidence that someone could do the work without forcing you to become an expert in evaluating every tool they claim to know?
Patrick A.
·
(edited)·
I wouldn't take the time to learn every tool a person claims to know - especially if I don't use those tools. Its pretty common for JD's to list the tools being used. I can confidently say that when I ask "Which SLM/LLM are you using?" and then someone goes "Cursor!" that they will not make it through. I am abnormal in the fact that I will really go out of my way to meet people in the middle by working hard to talk to them, try to understand them so they are heard and make that value judgment myself. Those lil extras still matter to me. #mainer
❤️1

Patrick A.
·
(edited)·
I think there's a huge difference around of how someone uses genAI. I would argue prompt work is just one piece of the puzzle that gets you in the door. So I look for more than just "I can prompt" when talking to others. I will be opening a role soon and have been thinking about this question alot.
Lara D.
·
·
That's a big problem facing today. Companies are using AI to find employees and potential employees are using AI to (hopefully) land a job. Lot of resumes end up with hallucinations on them, whether the hopeful employee intended the padding or not. Something needs to change in the hiring flows, but what?
Patrick A.
·
(edited)·
I think diligence matters; Lara D. I put the onus on the user for that. If someone doesn't want to check their work thats an individuals issue. As it pertains to ATS leveraging Ai - that comes down to understanding what the context actually means when implmenting- you know, Plato. I kind of see throwing on AI for the sake of throwing AI on as the same as putting a spark plug in with out learning to gap it correctly.
Lara D.
·
·
I 100% agree that the applicant owns the accuracy of what they submit. I’m thinking one layer beyond that, though. If self-reported AI proficiency is consistently overstated, whether intentionally or not, relying on resumes and self-assessment won't work. Hiring teams needs a way to validate the applicant can use the specific tools effectively.
Patrick A.
·
(edited)·
What I am saying its on both - hiring teams and applicant. Hiring teams cant honestly expect great applications if the job postings lack context, not check their work either and aren't involved in learning how this AI use case actually works when implemented. Hiring and HR have a plethora of tools and assessments as it is plus the majority of applicants have degrees (that are required). It's reasonable to argue that its not possible to get a good data set if both ends of the equation are flawed. I mean the best way to validate would be to just pick up the phone, no? Another option would be to just have the candidate show you on a call (or in person) how they use it.
Lara D.
·
·
Patrick A. - If you were opening that role tomorrow, what would really give you confidence that someone could do the work without forcing you to become an expert in evaluating every tool they claim to know?
Patrick A.
·
(edited)·
I wouldn't take the time to learn every tool a person claims to know - especially if I don't use those tools. Its pretty common for JD's to list the tools being used. I can confidently say that when I ask "Which SLM/LLM are you using?" and then someone goes "Cursor!" that they will not make it through. I am abnormal in the fact that I will really go out of my way to meet people in the middle by working hard to talk to them, try to understand them so they are heard and make that value judgment myself. Those lil extras still matter to me. #mainer
❤️1