General Episode Description:In Part 2 of this conversation, Bob Kelly, Founder and Chairman of the Sales Management Association, joins Mark Petruzzi and KK Anderson to examine why AI adoption often fails even when organizations invest in the right technology.Building on research across 111 sales organizations, the discussion moves beyond AI tools and into the operational systems required to create measurable impact. Bob explains why automating a broken sales process only accelerates poor execution, how leaders should distinguish between path problems and insight problems, and why CRM adoption continues to struggle when technology creates work for sellers without delivering value back to them.The conversation also explores the three organizational levers shown to drive AI adoption: leadership endorsement, training investment, and AI proficiency in hiring criteria. Bob shares why AI implementation requires clear decision rights, governance, and process ownership, while Mark and KK examine how AI may finally help revenue organizations break down traditional departmental silos.The episode closes with a candid discussion about AI anxiety, the changing role of sales managers, and why experimentation may become one of the most important leadership practices in the future of sales.What You’ll Learn:Process Before AI: Why automating a broken sales process only creates bad outcomes faster and at greater scale.Path Problems vs. Insight Problems: How leaders can distinguish between work that follows a defined decision tree and work that requires judgment.CRM Data and Seller Productivity: Why salespeople rationally avoid systems that consume selling time without helping them sell.The Three AI Adoption Levers: How leadership endorsement, training investment, and AI-focused hiring criteria influence adoption.AI Governance and Decision Rights: Why organizations must define who builds, owns, instructs, and controls AI agents.Key Topics:The 27-point performance gap between frontier AI adopters and other sales organizationsGeneral-purpose AI, agentic tools, and sales-specific applied AIWhy having ChatGPT open does not mean an organization has adopted AIAI as an amplifier of existing sales processesPath problems versus insight problemsProcess automation and AI-powered decision guidanceWhy “bad fast” can be worse than “bad slow”Using AI to capture CRM data from emails, calls, notes, and conversationsSpeech-to-text technology and sales data captureWhy most CRM implementations fail to create value for salespeopleDocumenting new customer, expansion, and cross-sell sales processesCapturing effective selling messages and value propositionsProcess flexibility and salesperson decision-makingLeadership endorsement as an AI adoption driverThe 36-point AI adoption gap linked to leadership endorsementTraining sales teams to use AI effectivelyAdding AI proficiency to sales hiring criteriaAI governance, security, and decision rightsWho owns and controls AI agents inside an organizationAI’s potential to break down departmental silosCross-functional AI workflowsThe future skills of sales managersAI anxiety and uncertainty across middle managementExperimentation as a sales leadership disciplineGuest Spotlight: Bob KellyBob Kelly is the Founder and Chairman of the Sales Management Association, an independent professional organization focused on advancing sales leadership and management effectiveness. Through research, education, and industry events, Bob works with sales leaders to apply evidence-based practices to sales productivity, process design, technology adoption, and organizational performance.He also serves as an adjunct faculty member at multiple business schools and brings an academic, research-driven perspective to the changing role of sales management in the AI era.Resources & Mentions:Sales Management AssociationMicrosoftChatGPTMicrosoft CopilotZapierCRM platformsAgentic AIApplied AIAI governance and securityMiller HeimanTony RobbinsRoman history and leadership researchSales Management Association AI adoption researchKey Takeaway:AI adoption is not primarily a technology problem. It is a leadership, process, and organizational design problem. The sales organizations creating meaningful AI advantage are not simply buying more tools. They are documenting how selling actually works, defining where human judgment belongs, investing in AI skills, and giving leaders clear ownership over how AI is deployed.Technology will continue to change. The organizations that build the discipline to experiment, learn, and redesign their sales systems will be the ones best positioned to adapt.🎧 Listen now and subscribe to Selling Intelligence for more conversations on AI, sales leadership, revenue operations, and the future of go-to-market execution.#SellingIntelligence #AIforSales #SalesLeadership #RevenueOperations #AgenticAI #AppliedAI #SalesManagement #ArtificialIntelligence #SalesProcess #SalesEnablement #GoToMarket #RevenueLeadership #SalesStrategy #AIAdoption #FutureOfSalesMark Petruzzi

Welcome to Selling Intelligence. I'm Mark Petruzzi.KK Anderson

And I'm KK Anderson. Every week we bring you conversations with operators who have been in the room and come out with something worth using the very next day.Mark Petruzzi

Our guests today just completed a benchmarking study of 111 sales organizations. And what on? AI adoption. The findings are not what most leaders expect. Only 4% of firms have no one using AI at all. But only 9% qualify as what the research calls frontier adopters. And those 10 firms are outperforming the rest of the field.By on average twenty-seven percent is points on sales objective achievement, nineteen points on individual quota attainment, and nearly double the rate of sufficient selling capacity. The question this research forces is not whether your team is using AI, it is whether they are using it in a way that actually moves the numbers.KK Anderson

Bob Kelly is founder and chairman of the Sales Management Association, a global, independent, professional organization serving more than ten thousand sales leaders. He has no product to sell and no platform to push, which is exactly why this data matters. Welcome to the show, Bob. We're happy to have you.Bob Kelly

Thank you so much for having me. It's so nice to be with you.Mark Petruzzi

So Bob, your research covers over one hundred companies. But before we get into what leaders are doing, give us the honest snapshot. What is the typical sales organization actually doing with AI right now? And how does it compare to what people say they're doing and what the media outlets are saying people are doing as well?Bob Kelly

the word I would use is dabbling. the typical sales organization is dabbling with AI as opposed to integrating it in a pervasive and sort of formal way in the work streams in the firm. this is a reflection of the uneven nature of how AI affects work in general and including sales work.But the interesting thing is to understand just how few firms are ignoring AI or are untouched by AI. When we look at the adoption curve and history of other technologies, CRM, for example, or other platform technologies, they've taken much, much longer to gain a foothold in most organizations. We're seeing very fast pickup inThe use of AI, even if that use is uneven.KK Anderson

So Bob, when we had our our preparation conversation for this podcast, you used a phrase that I hadn't heard anywhere else and I thought it was fabulous. You said you said there's a lot of machine stench. Right. So walk us through like what is that and why is it showing up so broadly in this research?Bob Kelly

Machine Yes.Well how isGreat term, isn't it? I was just about to think of that term when I heard someone else use it. So I've naturally adopted it as if it were my own, but it is plenty descriptive, right? And I think we all have encountered AI generated content. I mean, the more common term is slop. We see it in social media, but we also see it in our inboxes, we see it in work product withinKK Anderson

It is.Bob Kelly

the companies that we work for. and its problem is that it represents an inauthentic product. And so this when used inappropriately or i in the wrong context actually I believe does more harm than good. it communicates sort of in auth authenticity rather thanclarity and equality.KK Anderson

Everyone AI it's like AI gave everyone a on the road and no one can move. And everyone sounds the same. So it's figuring out how do you take that superpower and that super intelligence and leverage it in a way to create differentiation rather thanBob Kelly

We wantYeah, justYou know, IKK Anderson

than the stench of sounding the same,Bob Kelly

I think the mistake so many make is that they're using AI to substitute for their own voice or work product. And when that happens, ⁓ they're pointing out their own superfluity. do we need a human if we're just gonna have machines generate communication? so I think when used appropriately, we use it to augment our abilities, our voice, ourproductivity and when we use it poorly we've we make the mistake of y substituting a machine generated product for something that we should have more say in.Mark Petruzzi

So, Bob, you mentioned in the study that only four to six percent of companies are avoiding AI entirely. That actually concerns me that, there is everybody's diving in. And, only a very small percentage are not. but when it really comes down to it, there's still a huge disparity between those frontier adopters, as you described, and the laggers, the rest of the theBob Kelly

Yeah.Mark Petruzzi

hack is that the gap between the two of them growing and and how fast? How fast is that growing and changing?Bob Kelly

Well, it the answer is very, very fast. The study that we did was concluded I guess about three months ago. It's likely out of date. so we see these adoption and the development of the products that we're even talking about moving so quickly. the the the thing to keep in mind is that or a relevantPoint here is that so many AI products are freely available and they are in the general population. In fact, the most widely adopted tools are those that anyone can pick up, chat GPT or perplexity or any of these sort of language-based chat uh-based tools. And so we see lots of people picking them up, fooling around with them. And this is unlike other technology.platforms that we've seen impacting the sales organization like CRM or SPM or sales enablement or any many many other tools. So this does beg the question, will this be a phenomenon driven by consumers? personal adoption? will it be democratized in that way or willcompanies be required to adopt sort of institution wide, highly controlled platforms to provision AI.KK Anderson

really interesting and one of the incredible benefits of leveraging AI in your sales organization is that you can take that tribal knowledge of your best your your top sales performer and figure out what it what is it that makes them the best and then automate some of those research tactics or techniques or processes in your in your AI agentic workflow.Bob Kelly

You know, theKK Anderson

And and that's something you could never do inside of, the confines of a sales force or hub spot.Bob Kelly

Yeah, th this is the promise. This that idea gets sales leaders excited. in practice, it's I don't know that we're quite there yet, but in general we see companies adopting the easy tools to pick up. These are chiefly language-based models, and they're used for language-oriented tasks, things like communication andcrafting emails, sometimes quite derivative, machine stench gen sort of emails, but also doing things like understanding researching prospects and markets. and where we see less adoption so far is in the things that require more advanced tools, including agenticand sales specific tools. and so what we found in these so-called frontier firms, which represent maybe the top ninetieth per the 90th percentile of adoption and usage, ⁓ is that they're much more likely to use tools that are customized to sales work and include agentic elements. So they're actually automating work and inincorporating AI capabilities within those automation streams.KK Anderson

talk us through what these, the frontier adopters and the research that you talked about. So what was the what was tell the audience, what was the criteria you used to define them and like what does that what does a frontier organization actually look like in practice?Bob Kelly

Well, w ⁓we had have you as you pointed out, roughly a hundred firms in the study. So we looked and we in the study measured their rates of adoption, the frequency with which the typical salesperson uses tools, and the kind of tools that they use. So we took those firms that had the highest rates of adoption and in which salespeople are using tools most frequently and in whichThey are using tools, they're using the general purpose tools, but they're also using agentic tools and custom tools that are geared toward sales specific tasks. And so those those are the basic criteria. We identified roughly 11 to 12 firms in that group and branded them frontier sales organizations. We did this in part.Because the sponsor of this research who underwrote the research was Microsoft, and they're using this term in the way they are marketing solutions to frontier firms. So in part we did this to curry favor with our sponsor, but also because this term frontier has been widely adopted to describe the latest models that are used in these solutions, but also the sort of uneven andperhaps dangerous nature of adoption itself. Ethan Moloch's great term, the jagged frontier, is widely used to describe how it's an uneven place being on the edge of adoption with these tools. The tools do many things really well and other things, easy things, weirdly, quite poorly. So the it it's a it's a shifting, sort of an uneven place.to be and that that's why we like the term in addition to describing those leaders in organiza in our economy in sales organizations that are adopting AI.Mark Petruzzi

So Bob, I I kind of echo that that thought process and and what what I'm starting to see, like everyone, right? I jumped in early to just the whole opportunity of AI and what you can do and the research capabilities and and then move very quickly into the agentic side and really creating agents and and building them for our clients.And that moved to a whole nother level of productivity. But we learned something in that phase two, and that is how valuable all the systems and tribal knowledge that a company has and how to for AI really to be playing its best game, it's gotta have that that information as well. So that brings us to applied AI, right? That's kind of wherewhere the terminology is coming together and and when you can do that as an organization we at AGS do that for our clients we do that with with a couple of our partner organizations that we work very closely with like Humagentic. when you can go to that next level and really integrate within the systems that are already in place takes you to a whole nother level. did you see that at all in this study and what's what's your overallPoint of view on that.Bob Kelly

We we didn't go super deep to understand the degree to which firms are integrating agentic solutions within their own environments. we did sort of investigate, we asked, are you employing agentic solutions? and we ask, are you using custom solutions that are sales focused? We also ask if they are making use of platform solutions, existing solutions like CRM.or SPM sales performance management platforms in which there are embedded AI capabilities. This is likely the most this is the likeliest way that many firms will begin adopting these agentic solutions kind of pervasively they will exist within the comfortable and ofKK Anderson

Tools they have.Bob Kelly

a platform in which they're already using which solves the chief concern about distributing this capability across the enterprise, namely security governance, the keeping data secure and understanding just how people are using the solutions.KK Anderson

I thought it was incredible that the data showed that ninety percent of frontier organizations met or exceeded their sales goals in the preceding twelve months, versus sixty three percent of the non frontier firms. Like that's compelling.Bob Kelly

Yeah, yeah, it's a big number. so it suggests that frontier firms, so called frontier firms, are much more likely to be productive than more laggard firms. I mean, it's a clear delivery of the promise of AI i as it relates to productivity.KK Anderson

Right. Really incredible. Okay. So we your your research study also talked about selling capacity. and it's let's see, the data says I believe it was 60% of frontier firms say they have sufficient capacity to cover available opportunity compared to 32% of non-frontier firms. And I'm assuming, Bob, that's going more towards their productivity as well.Bob Kelly

So this is a a really interesting finding. again, this data is self-reported, so it is subject to the caveats that you'd have in such a study, but it's almost the only way to conduct quick benchmarking research. But the fact that these firms are so much more likely to indicate that they have sufficient capacity does suggest.An impact of using AI that we all are hoping for, which is it allows us to do our work faster. Now, if you follow the logic through, ⁓ it's reasonable to suggest that this surplus capacity will be short-lived. So where we are perhaps in the curve of adoption of AI and sales is that we're using AI to great effect to make salespeople more productive, but we haven't redeployed that capacity.In ways that are likely to follow on from AI adoption. We're likely to see different s kinds of organization structures, perhaps different spans of control. We're almost certainly likely to see different jobs in the Salesforce. And as we absorb an understanding of how AI will impact all of the work of those new people, we'll probably see them.become every bit as overburdened and overworked as people used used to be before AI. they'll just be doing more. so that that's something to keep an eye out for. those excess capacity is something that you you don't want too much of it. but it's better to have a little bit than to be understaffed.Mark Petruzzi

Mm.KK Anderson

You heard it here first. Bob's prediction. The pendulum's gonna swing back the other way, isn't it? That's that's pretty funny. And I believe it. Hey, I believe you.Bob Kelly

Right now.Yeah.Right.there's an inter another interesting difference in these frontier firms compared to other firms. We asked about the sentiment about AI. what is the promise of AI in various aspects of managing the sales function, making decisions for the sales function? And we found that those firms that I'll put it this way adoption is positively correlated with enthusiasm inMark Petruzzi

All right.Bob Kelly

Estimating AI's impact. Those firms that are using AI the most are the most sanguine about how AI will help them in the future. and believe it, its impacts are far more broad based than firms with just a limited, sort of naive understanding of AI.Mark Petruzzi

Stuff, Bob. let's move over to topic three, the management question. How AI changes the sales manager's role. And your research in this area shows that 100% of frontier organizations expect AI to have a high impact on first-line sales manager responsibilities over the next two to three years. That is not a heads number.What does it actually mean for what a frontline manager does every day? And what does it mean for how CROs should be developing those managers over time?Bob Kelly

Yeah, this is a very interesting question. It's it's very interesting to our organization, given where we sit and who our ⁓ our members are. We have a lot of sales managers in our membership. it's important to understand that the sales management role, especially the first line sales management role, has been changing significantly before the introduction of AI. The role has hadAnd this is also a reflection of technology use and better data, but the role has had a diminishing amount of a sort of supervisory responsibility and an increasing ⁓ amount of focus on developing people, but also in a set of kind of integrative disciplines that I'll describe in a moment. soOf course, we all, think of that first line sales manager is maybe a a really good salesperson that was promoted into management. and it's it's of course a much different job than being an individual sales contributor. The the places where we see sales managers having the most impact are in actually improving the proficiency and effectiveness of the people that they are managing. that⁓ is a very much a ⁓ role that requires human to human communication and intervention and relationship. it's it's hard to automate that. the supervisory stuff is relatively easy to automate. nowanother way to think about this question, Mark, is to say what are the competencies and sales managers that will become in the forefront in the future? And increasingly we see this happening already. And there are a small set of competencies that I think will characterize high performing sales managers in the future. As it relates to AI, I do believe that proficiency or at least fluency inapplying and understanding and applying AI tools to sales work will be required for sales managers. Now this involves a lot of kind of com subordinate skills. It it it requires someone to be able to take a typical sales job or task and deconstruct it into its component parts to understand how we might automate that, whichportions are appropriate to automate, which portions should retain a human in the loop. And also it requires a process discipline. Sales, the sales function has always been process focused, or at least it has been in the past 20 or 30 years. But in the future, I think this sort of process rigor and the ability of a manager to apply process discipline to the selling function.And to diagnose the issues as would a process engineer. these are critical skills that w we'll have to have in sales management roles in the future.KK Anderson

So interesting.Bob Kelly

what else?Yeah. ⁓ the to return for a moment to this idea of developing people, despite what I've said, despite claiming that it's so important for people to be involved in this process, I believe that the currently today, the biggest impact we're seeing in AI and sales is in the area of learning and development. I think for the for most firms.the place where AI can have the biggest impact is in training, coaching and developing salespeople simply because the amount of quality activity in those areas is quite low or non existent in ⁓ many sales forces today.KK Anderson

So as sales organizations get smarter and sharper and and the applied AI really takes hold, and then the sales managers are not spending as much time on things like pipeline inspection or call reviews because the AI is listening to the calls, right? And the AI is pulling out insights. the AI is predicting the forecast, right? That well, this is like the vision, right? So if AI takes those off the plate.Bob Kelly

No right.Maybe. Yeah, yeah, yeah. ⁓KK Anderson

and I know you just said, hilariously so that, Something else will fill that time on the plate. but what what should the sales manager be focusing on with that time instead? Like they're gonna be getting back so much time. Where should their priority be? What's the w how will they have the greatest impact?Bob Kelly

Well, let's also point out that the typical environment that sales organizations operate in is very change intensive. in part but for the same reasons that we're that we're the same things we're talking about today. their customers are also being impacted by AI and b and by technology. The way they prefer to buy is changing all the time. they'reInterest in meeting with salespeople and the value they expect to get from such investments in meeting time changes all the time. They're all these new competitors, etc. So often even if only implicitly, we expect that sales manager to be the change leader to enact those adaptations that we have to have in the marketplace. soI expect that that will continue to be the case. Sales managers will have to lead change in organizations and there will continue to be plenty, plenty of it. it will be coming faster than it has in the past and and there'll be greater change impacts than we've had in the past. ⁓ so ⁓ I expect managers to be highly focused in that. They also are going to have to ⁓be skilled at diagnosing issues, diagnosing what's going on with pipeline issues. Will AI help our pipeline measurement and tracking? It it may well make it easier to do. Will it result in better quality? I'm on a little skeptical, but only because we've done plenty of research about pipeline management, and we find that the sort oftracking constructs that many firms use to track deals or opportunities are often wildly inaccurate at forecasting results. so much so that I've come to the conclusion that we often I believe use these tools for more than trying to simply predict outcomes. We use them to foster conversations.So another way to think about this question is will AI result in better quality interactions between management and ⁓ individual contributors to foster quality and figure out how best to close deals? I'll I'll see that. I'll believe the pipeline accuracy and efficiency impacts when I see them.But the good news is we blow a lot of time. Managers blow a lot of time fooling around with forecasting and pipeline tracking, especially in complex selling environments. And if we could have that go away, that would be a a good thing.Mark Petruzzi

Excellent, Bob. Okay, this is where we will leave part one. In part two, Bob gets into what the research says about the management role specifically and why process has to come before AI, or you are just accelerating the wrong things. And also we're going to cover the three organizational levers that the data shows actually drives adoption. We look forward to seeing you there.KK Anderson