About this episode
Eddie Reynolds, CEO and founder of Union Square Consulting, and Rachael Bueckert, marketing manager at USC, discuss the most effective AI applications for go-to-market teams based on the firm's extensive experience implementing AI across sales, marketing, and customer success. Eddie brings credibility from his tenure at Salesforce and years of consulting with high-growth SaaS companies on revenue operations and go-to-market strategy.
The core thesis is straightforward: if you cannot execute a go-to-market motion with humans, adding AI will not fix it. Eddie emphasizes that successful AI adoption requires the same foundational elements—clear ICP definition, documented processes, data quality, and rigorous execution—that make any human-driven process work. This insight reframes AI not as a silver bullet for organizational dysfunction, but as a co-pilot for augmenting already-functional operations.
The reframe here challenges the current market narrative. Rather than asking "How do we use AI to do more with less?" the right question is "What are we already doing well, and where can AI accelerate that?" This distinction separates organizations that see AI ROI from those that waste time and budget spinning up autonomous systems on broken foundations. The conversation then explores specific, high-impact use cases: AI-assisted account identification and research, personalized outbound email generation (with human review), lead scoring and prioritization, customer health scoring, expansion opportunity identification, and call transcript analysis. Eddie draws parallels to the sales automation boom of 2015-2017, when tools like Sales Loft and Outreach promised transformation but delivered only higher volume and lower response rates—a cautionary tale that applies directly to AI SDRs today. Rachael adds practical examples of how USC uses AI as a co-pilot for content creation (topic ideation, outlining, drafting, then manual refinement) and marketing analytics (hypothesis validation, pattern detection, faster data exploration) while maintaining voice, originality, and authenticity. The episode concludes with actionable guidance on how to choose your first AI use case, the importance of focused investment and alignment, and why understaffed customer success departments represent one of the highest-ROI opportunities for AI augmentation.
What we cover in this episode
-
1:42
Common AI implementation mistakes Most organizations simply hand out ChatGPT licenses or spin up AISDR tools without foundational processes, leading to wasted investment and poor results.
-
4:31
Jason Lemkin's AISDR success story SaaSter's AI SDR works because they built strong processes first, trained the AI like a new rep, and review every email before sending—not because the tool is magic.
-
9:32
AI as co-pilots, not autonomous systems AI augments human work at each step—research, messaging, summarization, prioritization—but cannot yet run go-to-market motions independently without human oversight.
-
14:28
Account identification and research with AI AI can analyze customer data to refine ICP definitions, research accounts deeply to inform targeting, and surface priority accounts using aggregated firmographic and intent data.
-
18:24
Foundational elements before AI acceleration Clear ICP, buyer personas, and documented processes must be 8/10 minimum before deploying AI to accelerate volume; otherwise AI amplifies dysfunction.
-
22:32
Process mapping for AI success Reverse-engineer why you win or lose deals; map each step of your sales, CS, and marketing processes; AI then augments each specific step with better data and speed.
-
24:22
Content generation with AI co-pilots USC uses AI for topic ideation, outlining, and drafting newsletters and podcast content, but manually edits and refines everything to preserve voice and original insights.
-
36:04
Customer success AI applications AI can analyze product usage, customer interactions, and QBR data to identify churn risk and expansion opportunities; critical for understaffed CS teams.
The lines worth sharing
If you can't figure out how to do it with humans, you're probably not going to figure out how to do it with AI.
Eddie Reynolds · 0:15
Just adding more volume isn't the answer. You've got to find something that's actually working.
Eddie Reynolds · 0:45
We're talking about co-pilots. We're talking about augmenting the human activities that are being done.
Eddie Reynolds · 9:32
Don't waste your time or money if you're not willing to review emails manually and coach the AI like you'd coach a new rep.
Eddie Reynolds · 51:36
Common questions from this episode
What is the most common mistake companies make when implementing AI in sales and marketing?
Distributing ChatGPT licenses or spinning up AISDR tools without first establishing strong foundational processes, clear ICP definitions, and documented go-to-market motions. AI amplifies what already works; it does not fix broken processes.
Can AI fully automate sales, marketing, and customer success?
No. AI can automate specific steps within each process (research, email drafting, call summarization), but effective go-to-market requires human oversight, judgment, and continuous coaching of AI systems to ensure quality outputs and strategic alignment.
What are the highest-ROI use cases for AI in go-to-market?
Account identification and research, personalized outbound messaging (with human review), lead scoring and prioritization, content ideation and drafting, call transcript analysis for risk identification, and customer health scoring and expansion opportunity detection in CS.
How should we pick our first AI use case?
Choose one specific motion you already execute well (score 8/10 or higher). Align all stakeholders. Invest focused time and resources to train and review AI outputs. Avoid attempting multiple initiatives simultaneously; succeed at one, then layer on others.
How does AI help understaffed customer success teams?
AI can draft proactive outreach to at-risk and expansion-ready accounts, summarize QBR data and customer interactions, identify patterns in product usage, and surface the highest-priority accounts for CSM attention—multiplying CS capacity without proportional headcount growth.
Should we automate our entire content production pipeline with AI?
No. Use AI for speed at each step—ideation, outlining, drafting—but humans must refine for voice, accuracy, and original insight. Purely AI-generated content lacks authenticity and context and will not build audience trust or generate leads.
SEO meta description
Eddie Reynolds and Rachael Bueckert on the real AI use cases that work in GTM: co-pilots, not autonomous systems. Without strong foundations, AI amplifies dysfunction.
Target keywords
AI in go-to-market
AI SDR implementation
AI in sales
Eddie Reynolds Union Square Consulting
customer success AI
lead scoring with AI
outbound messaging AI
AI co-pilots
GTM AI use cases
AI content generation
AI account research
customer health scoring AI
Read the complete conversation
Auto-generated from the episode audio and lightly cleaned. Expand to read or search inside.
Read the full transcript · 67 KB
EDDIE REYNOLDSIf you can't figure out how to do it with humans, you're probably not going to figure out how to do it with AI. As I think sales off and outreach taught us, just adding more volume isn't the answer. You've got to find something that's actually working. Welcome to Go-To-Market Science. There's an art and there's a science to go to market. In this podcast, we talk about the science by interviewing CROs, private equity investors, and other sales and marketing experts, as well as talking about what we learn every day in the trenches helping to build go-to-market engines.
RACHAEL BUECKERTWelcome back to another episode of Go-To-Market Science. I'm Rachel Buchert, the marketing manager here at Union Square Consulting. And with me is Eddie Reynolds, our CEO and founder. Hey, Eddie, how's it going? It's going well.
EDDIE REYNOLDSIt's going well. We're doing two podcasts back-to-back. Let's see how my energy levels are. I was like jazzed up on the last one. Maybe I will talk less on this one. Maybe this will be better.
RACHAEL BUECKERTThat's not what we want, though. We want you to talk. That's where all the stuff comes from. Yeah, there's a happy medium. All right. So today we are talking about the best use cases for AI in go-to-market. And this is the topic of a newsletter that we put out recently. At the time that this podcast is live, it probably won't be that recently anymore. But yeah, a topic that's very topical right now in today's AI-frenzied atmosphere.
EDDIE REYNOLDSVery topical. It feels like it's like the only topic.
RACHAEL BUECKERTYeah, right. So to start us off, Eddie, what do you think most people are still misunderstanding a lot about how AI should actually be used?
EDDIE REYNOLDSWell, it's interesting because we just did a podcast like five minutes ago on a totally different topic that we could have done 10 years ago, 20 years ago. And now all of a sudden we're talking about AI, this new thing that has just like exploded, especially this year. And I think so many organizations, like everybody's got ChatGPT and they got their reps on and they're all trying to say, hey, go use ChatGPT. And that's often like the, that's it. Hey, we implemented AI. We want everybody to ChatGPT license. We told them to go use it. It's like, okay, cool. And or like they're buying tools, like every single tool has AI in it. So like, I don't know, where do you draw the line? We're using AI. We're not using AI. It's like every single piece of software today seems to have AI embedded into it. You can't not be using AI. But what I think is, is like, I think every CRO and revenue leader is asking themselves, how can I drive more revenue by leveraging AI? We're seeing companies have higher expectations for revenue growth per employee or per salesperson. I had Hayes on the podcast to talk about his concept, the 10X AE. He posted this thing. Could we envision a future with a mid-market AE with a $10 million quota? And obviously that set a lot of people on fire. It was a really interesting thought exercise to discuss what dials could we turn with the use of AI to take a human being that was previously producing a million dollars of revenue and have them sell the same product to the same set of customers, meaning in this case, mid-market customers, not going enterprise and hey, like we do $5 million deals and there you go, solve that problem. No, like the same sales targets, the same products and make them 10X more productive. And I wanted to come off of that and I wanted to share like some specific tangible use cases where I think we are seeing AI have the most potential impact so that people can actually take this and run with it and say, hey, I think that's the thing we're going to go develop in our organization and also share some more fundamental concepts that are not necessarily so new on what we need to have in place in order to make those things work. Because the answer is like just giving everybody a chat GPT license or like spinning up an AISDR is not going to work for most organizations. With that said, I love the fact that Jason Lemkin of Sasser just came out and said, hey, we made the AISDR successful. He also exhaustively like set the disclaimer that we made it successful, but we have all these foundational elements in place that many other organizations don't have. And so that's why I wanted to write the article. People are starving for what the heck am I supposed to do with AI? There's so many different ways, like areas we could go. And I don't have all the answers, but putting my rev ops hat on and thinking through go to market, where does AI seem to have the most promise? We've talked about this extensively across our team and with a lot of our customers. And I wanted to share our thoughts as of August 6th and try not to go into any specific tools because those things will evolve quickly. And just think about like, what would an advanced AI do in different areas of the business and where could it have the greatest potential impact?
RACHAEL BUECKERTYeah. And this conversation is so important too, because so many companies right now are feeling a ton of pressure to move quickly with AI and get ahead of the game or if they feel like they're going to be left behind very fast, right? But the issue there is that they might not have the groundwork laid out to actually implement AI successfully. And so they're moving as fast as they can. They're trying to get all these tools and getting kind of ahead of themselves. And then wondering why it's not working for them.
EDDIE REYNOLDSAbsolutely. It's just like anything else. I mean, this AI frenzy reminds me of like Sales Loft and Outreach many years ago. Like what was it? 2015 when they blew up, if I remember correctly. 2016, 2017, something like that. When I started this company in 2016, it felt like this is all anybody in B2B SaaS wanted to talk about. And I didn't even have these tools when I was at Salesforce. And I left in 2016. I kind of came out of Salesforce and I kind of opened my eyes to the rest of the world. And I'm like, oh, wow. Hold on a second. These tools are exploding. Everybody wants to spin up an SDR team because they all read Predictable Revenue and read about how Salesforce built their SDR team and grew basically faster than any other software company in the world. And let's go implement that strategy. And we need Sales Loft and Outreach in order to do that. I'm sorry, not and, or. And then let's go and pump a bunch of content. Let's get our marketing manager to write a bunch of sales emails. And like, all right, we'll have our reps go and press send all and they'll record voicemails and get dropped voicemails and we'll make a thousand calls a day. Fast forward to 2025, we've been through that entire sort of maturity curve. And how's everybody doing with that? Most outbound sales engines are failing. All we've done is we've just like pushed more volume at the problem. So now prospects open their inbox even less, answer the phone even less. And we're still in the same place we were before. The organizations that have a really tight outbound process and strategy are succeeding. And most of the rest that don't are failing.
SPEAKER_18Yeah, probably the worst. And I don't think AI is any exception to that.
RACHAEL BUECKERTYeah, honestly, probably like worst situation than we were before. Because now people, like you said, are so skeptical of any cold email that they get in their inbox. Like you have to be coming to people with some relevant messaging that actually is going to matter to them to even get them to like open it and look at what you're saying.
EDDIE REYNOLDSYeah, I mean, and everybody says that, but I write cold emails all the time. And like I set up meetings and I'm telling you, I'm not writing a thousand, I don't have time to write a thousand emails. I mean, it's just what I do is super, super highly targeted. But the stuff that I'm seeing working for me, that I see working with our customers is fundamentally the same stuff that was working in 20 years ago, 30 years ago. We have slightly different tools. We have a little bit more sophisticated data, but it's the same idea. And I think what I want to talk about with the AI is you have to apply these first principles to get the AI to work. And I don't want to keep droning on. I want to get into these specific use cases so that we can talk about them and also talk about like some of the things that we're doing with our customers and what we're doing internally with AI and where I see this going. Because obviously this is the thing that everybody's trying to figure out right now.
SPEAKER_20Yeah, so the dream scenario, the ideal dream scenario is that we have AI marketers, SDRs, AEs, AMs, CSMs.
RACHAEL BUECKERTEverything is just fully autonomously AI and we're paying very, very little money to make lots and lots and lots of money.
EDDIE REYNOLDSThat's not the dream scenario. That's not the dream scenario? No, are you kidding me? That sounds like a lot of work. Rachel, the dream scenario is you pop into ChatGPT and you like build me a hedge fund and make me a billion dollars. And then you just like sync it with your bank account and then it just buys traded securities. And then, I mean, it's just really funny. I spent a lot of time in private equity, but ironically, like our old exchange student from the Czech Republic runs a PE firm in Prague. And I was talking to him one day and he was just lamenting and he's just like, oh, God, I'm sick of private equity. I want to work with hedge funds. How do you want to get hedge funds? He's like, I'm just so tired of dealing with people. So much work. So, yeah, that's the dream scenario. It's just like you go into ChatGPT and you say, make me a billion dollars. That's the dream scenario. I don't think that's realistic right now. And I don't think your scenario of having a fully autonomous go to market is realistic. So what we're talking about is we're talking about co-pilots. And I think everybody's kind of figured this out by now. We're talking about augmenting the human activities that are being done.
EDDIE REYNOLDSThere are steps in this process that could be fully automated by AI. There's no denying that.
EDDIE REYNOLDSBut I don't think that we're going to remove humans from marketing sales or CS entirely, at least not yet.
RACHAEL BUECKERTSo what parts of the vision do you think are actually realistic? Like, how do we, I mean, no part of the whole, like, going to ChatGPT and asking for a billion dollars is realistic.
SPEAKER_20But what can be realistic about bringing AI into these teams? Well, I think the first thing that is really important is to think about identifying one specific use case
RACHAEL BUECKERTthat you're going to focus the adequate amount of time and energy on to get right.
EDDIE REYNOLDSSo let's take Jason Lemkin's example of his AISDR or SASTR's AISDR. He wrote an article about this. You can find it. We obviously wrote an article about the best use cases in AI. And I don't know, are we going to link Jason's article? I don't remember if we have it linked. We can put it in the show notes. Yeah, we do. It's not hard to Google. What Jason talked about is, like, setting the foundation and saying, okay, like, what does an SDR do? What does a successful SDR look like? They have human SDRs that are first and foremost succeeding at booking meetings that turn into pipeline and close one deals, right? That's the prerequisite. If you can't figure out how to do it with humans, you're probably not going to figure out how to do it with AI. As I think sales off and outreach taught us, just adding more volume isn't the answer. You've got to find something that's actually working. So then you break that process down and you say, okay, which accounts do we want to go after with our human or AISDR? How do we identify accounts? How do we segment those accounts? In this particular use case, you could look at our outbound framework and how we've broken all of this down. And we've walked through the step-by-step process to do this with a human. So what's our ICP? What are our buyer personas? What are the accounts that we want to go after? Who are the contacts in those accounts that we want to go after? How do we segment them? How do we get relevant messaging for each segment? Okay. Now we've done all that. Let's just say that we've done all that manually as a one-time exercise. And now we have this reasonably large, but not too large list of prospects we want to go after with our AISDR. Jason then talked about, okay, now we go. And he also made the point like, hey, we have a strong brand presence in the market. People know who we are, et cetera. Then he also talked about, hey, we're taking warm prospects, people that are already in our funnel, opted in, et cetera. So he's really setting himself up for success here. It's a lot of foundational stuff there, right? Then we say, okay, how do we get AI to write our messaging? Now this is an example, and I don't want to try to plug tools because I'm worried that by the time somebody listens to this, some new development will come from either the tool I've mentioned or a new tool I've never heard of before. And then this podcast will quickly become outdated. But for the time being, let's use ChatGPT as the example, you give your rep really clear instructions to say, this is what a good email looks like. And now we have created this entire territory. We've created all these specific accounts and these contacts. We maybe used AI to help do that, or maybe we didn't. And then we've got a really, really strong prospect list. We've segmented it, and now we know what good messaging looks like. What is our process to actually research? What things should we be looking for? When we pull up a company's website, after we've already identified the account, what specifically should we look at? Let's use the financial statements as an example. The annual report. People always talk about this. This is like old school sales. Oh, I got the deal because I went in and I read the annual report. And I went to page 22, and I read this thing, and then I wrote the CEO an email, and I said, based on your goals of XYZ, AI can do that. AI can do that really, really well. So now, you take AI and say, this is our research process. This is the information that we need to look for. This is the information we need to digest. These are templates that we might want to go through. And then we get the AI to write that content. It's really easy for everyone to imagine how that might work. Now we're talking about a scenario where all we're doing is we're plugging these accounts in, and we're spitting out emails. Now we have a human review process. And the human reviews that and says, yes, this is an email that I indeed would send on my own. And you keep coaching ChatGPT or whatever AI you're using to do this so that you get the messaging right. And what Jason talked about in his article was putting just as much time into this as you would training a new SDR and building off this process souped to nuts and not just pressing go on the AI. And then by doing that and reviewing every single email, and also Jason even talked about how he was personally responding once those emails got responses manually, not just like slapping the AI on top of it, and trying to like perfect each step of that process. That is the winning formula. And it's the winning formula. And I feel confident saying this, even though I don't pretend to be an expert at AI, because that's what works for humans. And if you think about AI like a human, and you think like humans are imperfect, they have a learning curve, they need feedback, they need training, they need proper data, they need proper inputs. That's what you need to make AI successful, at least as of August 6th, 2025. In addition to that, everything I just walked through, you can use AI for each of those steps. So I think like as Hayes mentioned, and obviously he's plugging his own tool, GradientWorks, and I'm fine with that. Using AI for account identification, I think is brilliant. We talk a lot about how we don't think that ICP is well-defined enough. We could take AI and say, let's look at all the data across all of our existing customers, let's analyze that, and let's try to improve our definition of our ICP.
EDDIE REYNOLDSI think that also needs some human intervention. You're going to run that analysis, and you're going to look at it, and you're going to say, does this conclusion make sense? Now we say, okay, we're going to tweak our ICP a little bit. How do we use this information then to identify better accounts using AI to go after that?
EDDIE REYNOLDSAnd this is where I struggle with even the concept of using AI. Well, what tool doesn't have AI? I know when I say using AI, am I talking about ChatGPT? Am I talking about Clay? Am I talking about ZoomInfo? Honestly, I don't know. What I'm really trying to say is we use advanced analytics in order to do something that we were previously doing manually.
EDDIE REYNOLDSInstead of saying our ICP is this industry in this revenue range, let's go pick the highest revenue companies in these industries in that range and give them to our sales reps.
EDDIE REYNOLDSLet's ask AI to take this numerous steps further so that we can have a better account list. We can then take that from the account list to the segmentation to the messaging, but we need a co-pilot along the way, at least in 2025, to make sure that we get quality outputs before progressing to that next step. So we have a lot more to go through, but I'll stop there.
RACHAEL BUECKERTYeah, I was just going to say that's the difference between jumping ahead of everything else and using AI right out of the gate to try and accelerate all the things you're already doing versus using AI throughout the rest of the pyramid, you know, the go-to-market efficiency pyramid, which you can find in the frameworks of our website if you're not already familiar with it. But we used to have a version of the pyramid where the top part, we called it, it was automation, AI automation. And we decided to change that to acceleration because we realized, you know, you can use AI and automated tools to help you with all of these below levels, the fundamentals, adoption, optimization, all that. It's really when you get to accelerating what you already have, you need to have those foundations in place before you do that. Yeah, and if anybody's listening to this and they haven't heard us drone around about the go-to-market efficiency pyramid enough, like the base of the pyramid is essentially just identifying what is our ICP and our buyer personas and what is our process to market, sell, and serve them?
EDDIE REYNOLDSHow do we define that? There's no reason you can't use AI to look at your existing customer data if we have call recording to look at our calls and to help us improve who is our ICP and our buyer personas and what is the step-by-step process to identify quality marketing leads, to follow up with those leads, like what is the ideal sales process? You know, we have things like Medic as an example, but we could also look objectively at our data and ask, are those the things that really make the difference between whether or not we win or lose a deal?
EDDIE REYNOLDSWe might run all of our sales conversations and opportunities through the AI and come up with a different conclusion that, oh, wow, here's this factor that falls outside of one of these standard sales sales methodologies that we didn't realize was as important as it is.
EDDIE REYNOLDSAnd now we can improve the basic process for the way that we sell. And so, yeah, absolutely. I think we took a step back from that, especially as AI has evolved since we originally came up with that framework and said, there's no reason AI can't be used for every step in this process. But I don't think it's at a place where we can just have it go and run autonomously to do any aspect of go-to-market. It's more that we leverage it in each piece of it. But anyway, I don't want to drone on because I want our audience to come away with tangible takeaways for things that they could do. So go ahead and ask me more questions. We can talk about more of these things that we wrote in the newsletter about what we see as the most, you know.
RACHAEL BUECKERTYeah, exactly. I mean, it segues perfectly into that because first we want to look at the foundational elements, right? For example, clearly defining their ICP. So what are some tangible ways that people can take AI and develop and define these fundamentals?
SPEAKER_20Well, I think I kind of just walked through that, right?
EDDIE REYNOLDSI think that you want to be clear on your ICP and your buyer personas. And I also want to make clear to people when they look at the go-to-market efficiency pyramid, we are not suggesting that everything needs to be a 10 out of 10 before you progress to the next level. It's more like, is it an 8 out of 10? So if you're sitting here and you're like, yeah, maybe we could use AI to like improve our ICP, but like we have pretty strong close rates in our pipeline and we have pretty good net revenue retention. And I feel like we know who our customer is pretty well. But what we really, really need to do is set up more meetings, which by the way, like there's a lot that could be wrong with that conclusion. But let me just assume that it's right. That's fine. It's great. Let's figure out how we use AI to like generate more leads and pipeline. I'm okay with that. The question is whether or not the ICP and buyer personas and the basic process is good enough. So I think I was just walking through, you could take whatever data you have, your customer data, your call recording data, your marketing data, and feed it into AI and ask it to do analysis on what's working and not working and potentially get some conclusions that you might not come to manually. But if you've already made those conclusions manually and you're thinking, hey, like I think we know who our customer is and we know how to get to them, we know what channels are working, let's double down on that. Well, then that's where we can start to ask, okay, how can we do what we're already doing manually that's working better with AI? And those are where I think my head was at when we sat down to try to map out some specific use cases.
RACHAEL BUECKERTAnd another use case we talked about in the foundational elements portion of it is process mapping. So why is it so important when you're bringing AI into go-to-market to have like a well-mapped process enabled? So the process is essentially just reverse engineering why we win or lose deals or customers, right?
EDDIE REYNOLDSSo if we talk about net revenue retention, the process of handing off the customer from sales to CS, onboarding, implementations, monitoring customer health, addressing unhealthy customers, renewing healthy customers, expanding healthy customers. That entire process and every step that's within each of those milestones that I mentioned, the only reason we do any of it is because we want to retain and grow a customer.
EDDIE REYNOLDSIf we didn't have to do that, we wouldn't do any of that stuff, right? I'm not aware of Amazon, at least in terms of buying like consumer goods on Amazon, of them doing any of that stuff. Forgive me if I'm wrong. I keep buying stuff from Amazon regardless. Great. I'm not suggesting Amazon needs to like do all that stuff. I don't know if they do or not, but in B2B SaaS, enterprise B2B SaaS, like we have to do that stuff. So if we are not following that process, what it means is by definition, we're losing customers because we don't have clear understanding of what that process is and or we're not executing that process. Just by definition, we're losing customers.
EDDIE REYNOLDSWe're either churning them and or we're losing opportunities to expand them. On the new business side, same thing, like we are losing opportunities to convert leads and pipeline into customers.
EDDIE REYNOLDSThat's what that process is. So the way that we can define that is by using methodologies like Medic or MedPick, best practices in CS, et cetera, and or feeding our data and call recordings, et cetera, into the AI and asking it, can you analyze what are the common threads for the customers that we're successfully expanding versus the ones that we're not versus the ones that we're churning?
EDDIE REYNOLDSWhat are the common threads for all the deals that we're winning versus losing on the new business side? And we may come up with some conclusions that we might not otherwise come up with manually. But again, I'm not suggesting that this is the first place we go to if we say, hey, wow, like our close rate's 30%.
EDDIE REYNOLDSI think we've got a pretty strong sales process. Great. That might not be the place where we need to start with AI.
EDDIE REYNOLDSWe may have a good foundation in place upon which we can build, upon which we might say, all right, let's think about how we can get more of these leads and pipeline into our funnel.
SPEAKER_46So then what would be the next step people could take?
EDDIE REYNOLDSWell, let's talk about top of funnel for a moment, right? Let's talk about specific examples. So right now, you and I have been seriously working on how to use AI throughout our entire content generation process. We are trying to use it as a co-pilot, not as a fully autonomous thing. We're not pumping out newsletters or podcasts that are just come straight from AI. So we broke down the process. We've sat down and said, okay, how do we go about identifying topic ideas? How do we go about writing an outline of a newsletter or a podcast? How do we go about actually writing out the content in that outline, writing out the questions to ask in a podcast? And then how do we go about actually finalizing that newsletter, recording that podcast, writing the LinkedIn posts, etc. And we broke down each piece of that. And then we've been training the AI to help us do that. And now I think we're like saving at least 50% of our time. At least I'm saving 50% of my time on that production process. But by no means is there any point there where we just like let the thing go and don't work with it. And I think that if we can develop twice as much quality content, that can help us generate more leads. I think that's a really obvious one. We talked about account identification. Researching accounts and contacts is another thing, right?
EDDIE REYNOLDSPut the AI on a set number of accounts and say, go through these accounts and do all this research. Whether it's to help inform us on which ones we want to prioritize and or to help us with messaging, etc. I think like there's an incredible use case for that. To be able to summarize that information and then also point the sales rep at that information and say, hey, this is a summary. But click here to see the original source that you can then read and verify that what this is saying is actually accurate.
EDDIE REYNOLDSIt's obviously writing highly personalized outbound emails. This is another example where, at least in the beginning, we want to make sure that those emails are something we would send with our own hands.
EDDIE REYNOLDSSo we want to review those manually. But if we really carefully train the AI, just like we would an SDR, we could most likely come up with an email that goes and looks through all the right things on the website, goes and reads the annual report, summarizes it, and then spits out an email we would actually write with our own hands that looks like emails that we've sent out that have had successful responses.
EDDIE REYNOLDSAnd then have the rep look at it and say, wow, this looks really personalized, this looks relevant, everything looks accurate, it doesn't look like it was written by AI, it looks like it's my voice, maybe I'll make a tweak here and there, and I press send. Now we have better quality messaging and more quantity of messaging with the same human input. Let's talk about another use case that we shared in pipeline generation is scoring and prioritizing leads based on intent and fit.
EDDIE REYNOLDSThe AI gives us the ability to go deeper than we might with other tools to look at what does our lead scoring model look like? What does their intent look like? Do we have third-party intent? Do we have other data that we can use to identify whether or not their inner ICP, inner buyer personas are really a fit? If we can aggregate more data and feed more data to the engine and then give the engine a better ability to go deeper on analyzing leads, then we can prioritize our leads better. But this is no, like, at least as far as I'm aware, on August 6th, 2025, I don't know that there's an easy button to do this, but it's definitely possible with enough focus and work.
EDDIE REYNOLDSAnd then lastly, we talk about aggregating data to identify whole accounts to prioritize. So we've talked about our all-bound framework and looking at everything across sales and marketing, first-party intent, third-party intent, ICP, buyer personas, technographic data, firmographic data, feeding that in to an AI. Across accounts and not just across leads and saying, look at our entire sort of territory, our entire TAM, and help us identify the absolute best accounts for us to target. That, I think, is a really powerful use case to use AI.
RACHAEL BUECKERTAnd I can bring it back to marketing for a second again as well, because I might be able to shed some more color on that. I mean, I use AI pretty much every day when I'm doing marketing for USC, and that goes far beyond just, like, content creation. You know, we don't create all of our content with AI, but we do use it for topic ideation sometimes, and we do use it for giving quick outlines and drafts. But then that is all manually edited and made sure it's in our real voice. And then, Eddie, you coming in and making sure that, like, your perspective is in there as well, and we're actually bringing real, actionable, original, intangible insights to our audience. And it's not just, like, regurgitated AI slop, like a lot of stuff that comes out these days. But also, I use it a lot in, you know, marketing analytics as well. I will look at our metrics and our analytics for our campaigns, and I will do what I can, come up with my own hypotheses, and then I'll also input that data into ChatGPT and get validation for things that I'm thinking of. We're asking to see if it can find any patterns or anything that maybe I might have missed. And that's not just getting AI to do all my work for me. It's having AI, a really quick assistant that can just give me information that I need right now. And so I'm not spending hours hunting numbers down or researching something specific. You know, I can just get that at the snap of a finger pretty much. But there does need to be, like, that person behind it putting the human intelligence into it because there's so much stuff that AI misses.
SPEAKER_20Well, and it's no different from how I work with you as a human.
EDDIE REYNOLDSI mean, I could have hired you and said, hey, Rachel, just go pump out a bunch of content so that we can, like, feed the SEO engine that we don't have.
SPEAKER_50Yeah. And it's like, that would have been so much easier.
EDDIE REYNOLDSSo much easier. But I nitpick every single comma and space, and you and I argue over whether or not we should have, like, a semicolon or you actually like these, like, dashes or however you pronounce it.
RACHAEL BUECKERTM dashes. I love M dashes. And I'm... Why is it called an M dash, by the way? Well, because the N dash is, like, the shorter dash with, like, hyphenated words. The M dash is, like, the longer dash. Anyways, I don't want to get on to a tangent there. But as a copywriter, I love M dashes. They're so handy in so much writing to, like, make a little aside or emphasize a point that you're making in a sentence. And it's a real annoyance to me that it's become a trademark for AI now. And now I have to strip all my writing of M dashes because people just think it's all completely AI written. Anyways, that's my little rant. But that's the point here is, like, I nitpick you on things as granular as that because I want the newsletter to come out in my voice and to represent what I'm saying.
EDDIE REYNOLDSAnd it's been so hard working with you like you're a talented copywriter, but I have a different voice. And sometimes you'll write a whole newsletter and it'll get my, like, brain flowing. And then I go and rewrite the entire thing. And it's like I've thrown all the work you've done in the trash and I feel terrible about that. But then I'm also like, okay, at least the newsletter is truly my voice. And I can stand behind it and say, hey, like, I didn't just, like, delegate this to somebody. And it would have been so much easier to say, hey, Rachel, like, just start spitting out a bunch of newsletters. But I don't think that that's why people follow our content. I think they want to hear firsthand, like, hey, Eddie, what are you seeing from your team and from your clients every single day in the trenches? And if I'm not sharing that, then I don't think it provides value to our audience. And I think that that's a perfect example of how AI can or can't be used, whether we're talking about marketing or sales or CS. If I'm not delivering the real value, then it's not good enough. And it's not a reflection on you as a human being or on the AI. It's just, like, we have to figure out each step in the process and what our bar for quality is. And my bar for quality is, like, hey, Rachel, you're a great writer, but without firsthand experience in this stuff, unless you've interviewed me or somebody else on the team, and we can really stand behind everything in that newsletter, we're not putting it out there.
RACHAEL BUECKERTAbsolutely. And, you know, I'm not precious at all about having anything I've written being tossed in to make way for things that are more aligned with what we're trying to do. Any more than I am about an MQL not end up being qualified or closing or something like that. For me, the goal is the value at the end of the day and the quality and the revenue. It's not just the piece itself, if that makes sense.
EDDIE REYNOLDSIt does. And I love that about working for you. Man, wouldn't life be easier if we just went the other direction?
SPEAKER_08It'd be easier, but not as profitable.
EDDIE REYNOLDSI don't know that we would be producing the leads and revenue that we're producing through our content if I was just like, hey, Rachel, just figure it out. Google it. And I think that's kind of where AI is in terms of content generation today. It just reminds me of like these organizations that are like, we need content to feed our SEO. So we're just going to hire a bunch of content writers to regurgitate a bunch of stuff they don't really understand that well. That's like, okay, cool. Anyway, we probably spent too much time on marketing content here.
RACHAEL BUECKERTYeah, I was just going to say, doing stuff like that, especially right now with AI being so prevalent in especially writing and articles and stuff online, that's not going to get you fans.
RACHAEL BUECKERTIt's not going to get you people who follow your content and believe in you as like an authority figure on the subject or think that they should work with you. It's just going to be fodder for Google. So if that's your goal, then have at it. But I don't think that's most people's goals.
EDDIE REYNOLDS100%. But I do think it really is amazing in terms of how it can like help you develop ideas faster, how it can help you with each step of the process. Give me an outline. And now I look at the outline. I'm like, nope, actually, those are not the three things I want to say here. Let me change it to that. Okay, now fill that in. And now like, nope, I would say this differently. But now this is really truly what I would say if I put twice as much time into it without any help. I think some of the use cases that are really valuable in pipeline management and closing are obviously summarizing call transcripts and identifying red flags. I mean, we've been looking at this for a while. Salesforce was at least pretending to say that Einstein could do this many years ago, suggesting next steps based on historic patterns, auto updating certain fields in the CRM. I thought it was really interesting. We had a strong debate internally over this. Would AI be able to just like completely update your pipeline? Meaning it's going to close lost deals, move them out of your pipeline. It's going to move deals through certain stages for you. And we debated this a lot. And it's like, who knows? Who knows what the future will hold? But right now, probably not.
EDDIE REYNOLDSProbably wouldn't trust AI to be managing my pipeline for me and determining what deals go in pipeline and what stages they go into. But, man, is it powerful to look at the call transcripts and try to identify risks in a deal, to try to take some of the burden of admin off of the salesperson's plate so that they can go from call to call faster and be more prepared for those calls, to know what to look for in each deal, to update their CRM with less manual effort, to help them draft follow-ups and recap emails and help them prep for sales calls.
EDDIE REYNOLDSThose are the use cases that, to me, seem most powerful. And I'll go back to even the human element. I mean, for two years, I've been having Sarah, our executive assistant, do a lot of that stuff manually. She's been updating Salesforce for me. She's been doing research on every single call. Every single day, I get a quip note from her and it says, here's all the calls you have today. Here's links to all of those quip notes. Here's all the research that I've done. And I've given her really explicit instructions, like tell me how much revenue, how much headcount, do they have a CRO, how many people in sales, how many people in marketing, how many people in CS. Like there's a good amount of work that goes into this and then I can pop in and in five minutes I can skim through this and I can understand more about this company and the individual that I'm talking to than I would probably know if I did 30 minutes of manual research on my own. And there's absolutely no reason why we couldn't train an AI to do that. So that was a spiel. But I think those are the most interesting use cases in pipeline management and closing that I can think of.
RACHAEL BUECKERTYeah. And it's just interesting to me, the comparison of AI to, you know, like an intern or something like that, or like a really, really smart and extremely lightning fast intern.
SPEAKER_11Smart and extremely lightning fast intern, but one that completely lacks context and experience. It's like you've hired this like 20 year old from Harvard, but they just don't have the, they just don't have the experience and the context.
EDDIE REYNOLDSAnd that's why I like, when I talk about pipeline management and closing, you'll notice I didn't even mention forecasting. I think forecasting can be really powerful, but it's so important to have that foundation in place. Let's say that we are going to look for red flags. Well, what specific red flags should we look for? Why do we lose deals? What is our sales process? What are the things that we should be looking for? You can't just ask the AI, tell me the red flags in this deal. You have to say, these are the reasons why we lose deals most often. Do you see this stuff in the call transcripts? Right. Just as if you hired that like really smart person from Harvard. It's actually really funny. Like when I started this business, I shared an office with a really smart person from, I don't think she went to Harvard, but it was, maybe it was one of the little Ivies. I think she went to like Williams or something like that. And she was like 21 and like training sales teams. And I was just like, you're super smart and you've read a shitload of blog reports or a blog post, but like you've never actually done sales. And it was just really interesting to see that. I think it's the perfect analogy for AI. Sometimes I'd hear her on the phone like giving people advice and I'm just like, that's so wrong. Like you're so smart, but like you don't have the context to understand why that doesn't work.
RACHAEL BUECKERTYeah. And yeah, and AI doesn't get context at all. You have to provide all of that, like spoon feed it to it. So moving on to customer success. How can you use AI to get ahead of like churn or identify expansion opportunities or customer risk, stuff like that?
SPEAKER_16Well, I think analyzing product usage is a really interesting one, right? Here's an example where you potentially could have a massive set of data, right?
EDDIE REYNOLDSLike think about it. Like what does AI do? AI at the end of the day is just analyzing a bunch of inputs and then it's spitting out some kind of an outcome, right? It's saying, yes, this is good. No, this is bad. Or yes, this is a cat. No, this is a dog. Or yes, this article matches all the things you asked me for in the article. Looking at a really large data set is always like a really obvious use case for AI and product usage is a great one.
RACHAEL BUECKERTWhat are all the different things that could influence a customer to churn or not churn? And what is the correlation between those different things and which customers are more likely to churn and not?
EDDIE REYNOLDSThat I think is a really, really powerful use case to train AI on. Reviewing customer interactions as well. So now we've taken the hard data and we've looked at like what are red flags and green flags? What about the actual like conversational data with our existing customers? What about the QBRs? What about tickets and customer service calls that come in? How can we aggregate all that data together and feed that into the AI to identify red, yellow, green accounts and then decide what to do about that? That I think is like an incredibly powerful use case coupled with what we've already talked about, which is making sure that we have the most clear definition of our ICP and which customers like regardless of all that stuff are more likely to retain and grow versus the ones that are more likely to churn no matter what we do.
SPEAKER_20And even using AI to create those connective tissues between these different departments, right?
RACHAEL BUECKERTLike call summaries, either sales calls or just meetings with existing clients. You know, marketing can use those kinds of summaries. And like with Otter, I know we don't want to get into tools too much, but Otter is an example. And use that as feedback for the stuff that marketing is doing or the stuff that sales is doing. Yeah, or Gong.
EDDIE REYNOLDSWe don't personally use Gong just because of the price, but Gong, a lot of our customers have it. Obviously, it has a lot of benefits. That's another one that's an obvious one. But I think with customer success, there's something about CS for me that like I always struggle to talk through because with new business, like it seems easier to sort of just break things down. You've got inbound, you've got outbound, you've got pipeline management. That's not holistic. I'm missing partners, I'm missing PLG, I'm missing a bunch of stuff, but like it covers a lot of it. With CS, there's just so many more moving parts. And with more moving parts, there's more potential for things to break down, which means you have more data that you can feed the AI to come up with more insights, which I think is really interesting. And we were just talking in the last podcast about unblending the funnel. And I think it's so important to like start with your customers and look at like, how do we define our best customers and what is working best for them to retain them and expand them? And that's a rife opportunity for AI. The other thing in CS is that like most CS organizations are understaffed. I was just talking with a customer today, relatively new customer. We haven't had the opportunity to even try to fix this yet, but they're outperforming in sales and their implementations are really struggling. And they're having to delay customer implementations and they're having churn issues because they, A, didn't forecast and plan for enough resources in CS and B, don't have enough resources in CS. So imagine that we've done a better job at identifying these issues. And now we're using AI, just like we talked about the SDR on the CS side to draft proactive messages, to identify the best accounts to go after, whether it be to write the ship on a red account or whether it be to drive an expansion opportunity in a green account. How could we draft these proactive messages so that CS can touch more customers in less time? Helping prepare for QBRs and account reviews. How do we ingest all this information? Our customers give us so much more information than prospects. How do we ingest all this information and make really meaningful QBRs and account reviews and customer success calls to be more impactful in those calls? And again, like we have to have that foundation of understanding what does good look like? How do we define a really good QBR and what specific things should the AI be looking for in our product usage data, in our call transcript data, et cetera, to identify the things that we need to cover in that QBR? And then I already mentioned surfacing expansion opportunities. We have all this outlined in our newsletter, by the way, if my ramblings are too hard to follow. Those are the use cases that I think are most interesting in CS because we have a lot of data that we can analyze to identify which accounts are healthy, unhealthy, which accounts should we be trying to like land more of and which accounts should we be trying to spend less time landing a new business? And then what do we do with those accounts? How do we get in front of them more? How do we have more impactful meetings with them, more impactful messaging and interactions with them? And how do we find ways to retain and expand more of them?
RACHAEL BUECKERTYeah. And if you're listening to this, so many companies are, they just neglect CS so much. They neglect it with the attention that they give to the different departments that they're managing. And we talked about this in our last podcast, Unblending the Funnel. It's relevant for AI as well. And it's like, don't neglect CS with AI, especially since CS is one of those departments that, as you're saying, Eddie, usually very understaffed because, again, they get neglected. So the AI is, I believe, can be like a really powerful tool for increasing the capacity for our CSMs and making sure that we're serving our customers as well as we can. And also getting the data that we need to get to make sure that everything else is running properly. Like we have the right ICPs, we're selling to the right people, sales process is moving swiftly. We have good handoff between sales and customer success, all that stuff. I think AI can probably help with.
SPEAKER_18Well, especially because I think the bar is lower.
EDDIE REYNOLDSLike if you talk about, like, let's think about an enterprise sales cycle, right? Like, let's say that we're chasing a $2 million deal, and presumably we have a salesperson that is extremely experienced at closing $2 million deals and knows what they're doing. How much could AI help them? There's tons of stuff there, right? There's tons of deep research that would be done by that person that AI could help with. But how much is that going to improve quality? I don't know. I mean, that person is capable of reading the annual report and pulling out interesting conclusions that they then use in their sales process.
EDDIE REYNOLDSBut in CS, it's like, how many red accounts are we not even reaching out to? How many green accounts are we not even reaching out to?
RACHAEL BUECKERTThe bar could potentially be much lower, especially for an understaffed organization where you're like, hey, if we can just get this email out to this person versus not having an email at all that says, hey, there's this potential issue and here's how we could potentially help you solve it, that is so much more powerful.
EDDIE REYNOLDSAnd so when the bar, I think, is lower than that enterprise sales example, then I think you have more opportunity to move the needle on that. And then you do that across a large organization with a large number of customers. Now, all of a sudden, customers are getting these very helpful and relevant emails that are driving them to adopt the product more or at least driving them to meetings with the right CSM to address these issues. That can be really powerful.
RACHAEL BUECKERTYeah. And if your business is a recurring revenue business model, then that's kind of like the whole point of your business is keeping your customers and having the revenue be recurring. Right. So it's something that just it doesn't make sense to me that it's so neglected and not seen as a revenue center because it quite literally is if you're a BDSAS or any other kind of recurring revenue. And I know, Eddie, you were saying you were at the CRO Summit and I can't remember who said some companies are saying that something like. I think the average company is generating like 50% of their revenue growth from existing customers, 55%.
RACHAEL BUECKERTI can't remember the exact statistic, but the bottom line is like, and it's really obvious to me, like your existing customers are your best and easiest folks to sell to.
EDDIE REYNOLDSI don't think anybody like doesn't understand that. I think it just has to do with, you know, a lot of CROs are only responsible for new business. A lot of marketing is only thinking about new business.
RACHAEL BUECKERTI mean, how many marketers are thinking about targets for expansion and retention? Like so much of our content goes out to our existing customers, but it's just like the default thought process is like, oh, marketing is responsible for bringing in new customers.
EDDIE REYNOLDSI don't know when I was at Salesforce, like I do get so, so much marketing stuff to our existing customers. We put on events for existing customers. We brought them to like the major events like Dreamforce and the World Tour, and we shared all this content with them. And like, that's how we expanded accounts, but that's not as typical in other SaaS companies.
RACHAEL BUECKERTIf only organizations were set up in a way where customer success could flag accounts ready for expansion and share that information with sales, which could share it with marketing, who could create campaigns to target those existing companies or customers for expansion opportunities and working with sales to get them on the phone. If only we had like a unified go-to-market motion. Yeah, or like a singular leader that could like overlook all of those things and make sure it's all connected and rowing in the same direction. What would we call that person? I don't know. Like a chief something officer.
SPEAKER_06Revenue?
RACHAEL BUECKERTI don't know. Something revenue? All joking aside, I don't want to knock CROs. Like, I mean, we've worked with CROs before that previously managed sales, marketing, and CS, and they go into a new role and they're only overseeing sales and marketing. And it's like, I'm not here to knock CROs.
EDDIE REYNOLDSWe love CROs. Like we work for them. But these are some of the reasons why these things happen when we don't have a fully aligned organization and or for whatever reason, like the emphasis is just on like we need to land new customers. And I get it. We have to strike this balance between our short-term goals and our long-term goals. But long-term, like the whole reason that these companies have built recurring revenue businesses, we've done it as well. We have recurring revenue too. All of our revenue is recurring revenue with very few exceptions because that is like the most attractive business model in the world. So we need to focus on that. But I want to get off my soapbox here. Anything else we want to cover on these common use cases for AI?
RACHAEL BUECKERTWell, before that, I'll just quickly mention, you know, if you're listening to this and you're going, what are these guys talking about? If you're a CRO covering like everything that's not sales, we have a couple newsletters on making sure your organization is CRO ready, which is based on a podcast that I did with Warren Zena of the CRO Collective. It's a CRO Stories. I can't remember the title now off the top of my head. It'll be in the show notes though. Something, something, making sure. Yeah, it'll be in the show notes. There you go. Exactly. That's the solution to everything. But yeah, so you can take a listen to that or read those newsletters to see what we're talking about there when we're talking about CRO actually being a CRO in the organization. Not just a sales leader with a fancier title. But anyway, moving on to this topic. So where do you see go-to-market teams most often going off track when they're implementing AI? I know we've touched on this throughout.
EDDIE REYNOLDSWell, I think like I see them going off track when they're implementing AI or any solution or any motion, any initiative. It comes down to first, do we have alignment across all the stakeholders? And secondly, do we have enough focused energy and investment in this thing to see it through to make it work?
RACHAEL BUECKERTSo let's just take outbound as an example. Whether we're talking about humans or AISDRs or whatever, are we all committed to like building the foundation and saying we have a clear ICP buyer personas? We have a clear process. We have mapped out all of our accounts. We know who we're calling. We've segmented them.
EDDIE REYNOLDSWe've got the right messaging, et cetera, et cetera. We have a reporting mechanism for who's doing what and whether or not that's converting into meetings, pipeline, and closed one deals. We've got a mechanism to review those reports, to coach reps on what they could be doing better. And we're constantly analyzing this motion to try to like tweak it and improve it over time. That is what it takes to make any initiative in go-to-market work. What doesn't work is, hey, we're going to hustle. We're going to implement this new tool. We're going to spin it up in four weeks. We're going to release it to the team. And then hope it works. That never works. If anybody doubts this, first, like, please just come find me so I can say choice words to you.
SPEAKER_12I worked at Salesforce for three years and this is what half of my customers were doing.
RACHAEL BUECKERTThey're like, we need Salesforce because every sales team needs Salesforce. So we bought Salesforce and we did a six-week or six-month implementation. It doesn't really matter what.
EDDIE REYNOLDSAnd then we just, that's it. And then cool. We wiped our hands of our implementation partner and we didn't have anybody in-house that was capable of using it. And management never really gave a shit about it. And now nobody really uses Salesforce in any meaningful way. And, oh, by the way, our close rates suck so we can't get visibility into our pipeline and nobody's following up with our marketing leads. I wonder why that happens. So we need to get the foundational elements in place, but we also need to see things through. And I will say, like, with our customers, one of the issues I'm increasingly seeing so often is we can get them through, like, the foundational step, partially because we do most of the heavy lifting. But then it gets to this sort of inflection point where you're like, okay, like, you now have all the tools. And I mean that metaphorically, not literally, like, you have the tools and you also have the systems and processes and the data and the playbooks and all that stuff. Now you got to go do the thing. You're going to go execute. And where's the mechanism where the sales manager, in this case, an outbound, like they're managing the SDR, is holding the sales reps accountable and saying you need to make this many calls per day. I mean, actually, sorry, I shouldn't say that everybody does that. But you need to follow up X number of times with each new inbound lead, with each outbound prospect. I need you to cover these accounts specifically. What is our conversion rate for these accounts? How can I coach these reps? Like, there's so much missing there. And I would say the same thing with AI. Like, you're going to spin up an AISDR. You have to coach it and tweak it and improve it again and again and again until you get it right. You have to review these emails manually before they go out. And if an organization is not willing to put that in, and if you doubt me, go read Jason Lemkin's article about how SASTR did it themselves and about how much of this effort they put forth. If you're not willing to do that, then, like, don't waste your time or money buying some, like, tool to do this to just mail it in because it's not going to work. It's not going to work any better than the organizations that grabbed their marketing manager and said, write me 10 sales email templates that we can spit into sales loft and then press send all on. That didn't work either. This is not going to work if you're not committed to seeing it through. Another thing I would say is pick one thing and go make that work. Don't try to, like, get everybody chat GPT licenses and say, hey, we're going to do account research and we're going to do ICP identification. We're going to do this and we're going to do this and we're going to do this and we're going to do it all at once. And, like, everybody needs to use AI because AI is a new thing. Like, no, like, pick one thing and focus on it until you get it right. Or fail at it and then pick something different. There's nothing that drives me crazier and go to market than people playing whack-a-mole and trying to solve, like, 17 problems at once. Pick one thing and just make it work. And then if you, like, hit a point where you're like, all right, we've done everything that we can do and now all we can do is watch and see for the next two weeks and now we're going to spend, like, a little bit of time on this other thing while we're watching and seeing, then I'm fine with that. But this idea that we just run around and try to knock stuff out and then just drop the ball on it and move on to the next thing never, ever works. It never has. It never will. And the same thing applies to AI. Sorry. That was a different soapbox. I got a lot of soapboxes today.
RACHAEL BUECKERTYour soapboxes are great. I love hearing your rants.
RACHAEL BUECKERTYou're paid to say these things.
RACHAEL BUECKERTNo, not really. I could say I hate it. I don't think my pay would change.
EDDIE REYNOLDSThat's true. That's true. You know me too well at this point.
RACHAEL BUECKERTYeah. So on that note, what is one kind of maybe low risk but high reward AI use case that somebody could start with or you'd recommend starting with? Or if it's not one size fits all, how do people figure out what they should start with?
SPEAKER_12Well, I'll plug or go to market efficiency pyramid again. Like I just started ICP and buyer personas. Are we at least an 8 out of 10 on that?
RACHAEL BUECKERTSo if we are not an 8 out of 10, if we're not clear in who our ICP and buyer personas are, I'd start there. If we can check 8 out of 10 on that, we go to the next piece and we map out like the basic process. Who are we reaching out to? Who are we marketing to? Who are we selling to? What customers are we serving? How do we define like in segment territories and segments of customers, et cetera? How do we identify red, yellow, green? Like that I think would be the next piece is like, where do we push our energy through?
EDDIE REYNOLDSWhat is that process? And if we're like, yep, like we're good there. We know who like our best customers are, who our worst customers are, who our best prospects are. Could be better, but we'll give it an 8 out of 10. Then we can move on to the next thing. And then I think the next thing is looking at one of the steps in that process, whether it's researching an account to prep for a call or whether it's writing that outbound messaging and just saying like, what's the thing that's either most time consuming and or most difficult to do really well that we could apply AI on? And we've listed a bunch of different examples throughout this podcast and in our newsletter. And I think you just pick one and you say, okay, we've got this massive SDR team doing outbound calls. We've got all the foundational elements in place. What if we go and do what Sastra did and try to like really carefully, like work on building out this AI SDR to like generate messaging as a co-pilot to help us book more meetings? Okay, cool. That sounds great. What if we take the same concept in CS? Which is more important? I don't know. Which thing is more broken? Which thing provides more opportunity? Or on our pipeline management, how could we do a better job of running our sales process by using AI to do deeper research, by ingesting information from the company's website, from our recorded calls, etc. To help our reps do what's already working faster and better. And I think you just kind of got to pick a horse. I will say like we did another piece of content on this on sort of like the go-to-market decision tree. Is that what we called it? The go-to-market decision tree? Or go-to-market ops decision tree.
SPEAKER_58Something like that.
EDDIE REYNOLDSAnd I talked about, okay, identifying short-term versus long-term priorities. And basically where my head's at is because I was just talking to like one of our senior consultants today about this. And I'm like, hey, this is what I've been saying. Do you disagree? I think of it like this. Do we want more new business or more NRR? Which thing, if we could only do one, is more important? That's our go-to-market decision tree. And then it goes to the next layer. New business is like more pipeline or close more of the pipeline we're generating. And CS is retention versus expansion.
RACHAEL BUECKERTAnd then it goes to the third layer, fourth layer, etc. And we keep breaking it down.
EDDIE REYNOLDSWell, I kind of just want to like pick one of those things. And for me, it's always going to be NRR and just say, can we just get the basics in place? Can we just get this to 80%? We might not have this beautifully architected customer health scoring algorithm that feeds data from our product usage into Salesforce that our reps can then like read off and prioritize accounts from. Can we just quickly do a report in our database and figure out which accounts are red, yellow, green? Can we do that? Do we have that basic in place? Do we have like just a basic process? Here's a three-step process for what we do with a red account. If we have that, then I think where I want to go is to say, well, is there an easy way to use AI to improve our ability to like retain or expand our existing customers? And if not, then I would say like, is there a way that we could use AI to just close more of the deals that we're already generating? And if not, then I kind of say like, well, then let's try to like generate more top of funnel stuff. Now, you kind of have to like balance that out and say like short term, long term, which is the thing that if we really like turn the dials on it would have the greatest impact for our organization. And I tend to go backwards from net revenue retention to pipeline management to top of funnel pipeline generation, but it depends on the organization. We don't really work with startups, but if you're a startup with $200,000 in revenue, you're probably going to get more bang for your buck from like knocking out new business right now. But for most organizations, there's so much gold in net revenue retention that's being ignored. Either way, there's no wrong answer here. If you really focus on the particular use case for AI and work with it just like you would a human being until you get it right, you see that impact, then it's going to be a worthwhile investment, even if it's maybe not the net revenue retention that you focus on first. But where I would want to go is I would want to go backwards from like customer expansion to renewals to new business, pipeline management and pipeline generation. And I would want to ask, which is the first thing that's not an eight out of 10 that we could fix?
EDDIE REYNOLDSAnd do we need to fix it manually or do we need to fix it with AI?
SPEAKER_57Yep.
RACHAEL BUECKERTWe have frameworks on all of those things as well. And that will be in the show notes too. If you are curious about any of those sections like pipeline management, outbound, inbound, metrics and insights, annual planning, all that stuff will be in our frameworks in the show notes.
Cool.
RACHAEL BUECKERTAwesome. And that is all the questions I had.
EDDIE REYNOLDSAwesome.
RACHAEL BUECKERTIs there anything else you wanted to mention there?
EDDIE REYNOLDSNo, I think I just, I went off track a little bit and I was saying, I was talking to one of our senior consultants today about this and I was like, hey, like gut check me here. Like, am I off base? And he's like, no, that's right. However, there is also this sense of, hey, we just want to fix inbound. We want to like convert more of our inbound leads to pipeline. There's nothing wrong with that either. There's nothing wrong with just saying, okay, like how could we use AI to better score our leads? Or how could we use AI to like follow up with our inbound leads faster? There's absolutely nothing wrong with that. Like, especially if you're looking at this and saying, hey, our hypothesis is that if we put more attention on this, we can generate significantly more revenue. Maybe that's not the greatest thing that we can improve in the business. But that's not necessarily a bad thing to go and do that. And I think in RevOps, we oftentimes have two options. Like, do we look holistically at the entire organization as we would prefer to do across net revenue retention and new business and break it down into that go-to-market decision tree that we talked about and then drill into like the first thing that needs fixed? Or do we just say like, hey, we know this thing's on fire over here and it's actually a meaningful thing. It's not like a stupid thing like we need to check our field usage report in Salesforce that's not going to move the needle. Like, no, like we need to fix our conversion rate of inbound leads to pipeline. That's perfectly fine. And we can go knock that out and it's low-hanging fruit because our reps aren't following up with these like leads, whether we're using AI or not. And like, let's go knock that thing out. There's nothing wrong with that.
RACHAEL BUECKERTYeah. Awesome. If you want to find out how we can help you do any of the stuff that we've talked about, not anything, but most of the stuff we talked about in the podcast today, specifically surrounding, you know, process design, good market planning, systems, infrastructure, all that good stuff. You can find us at unionsquareconsulting.com, which will also be in the show notes.
EDDIE REYNOLDSCool. And we got a bunch of newsletters linked as well in the show notes, right? On all the stuff we talked about. Well, we will. And yeah, that about sums it up. So thank you.
SPEAKER_72Yeah. Thank you, Eddie.
EDDIE REYNOLDSThanks for listening to the show. If this resonated and or you'd like help with anything we talked about in the show, please reach out to us. You can find us at unionsquareconsulting.com and the info will be in our show notes.
