Ryan Milligan on this episode
Ryan Milligan is the Chief Revenue Officer at QuotaPath, a compensation management platform for revenue operations leaders. Over his time leading the sales organization, he has built a track record of sustained high performance, hitting 100% blended quota attainment in eight of the past ten quarters, while tripling ARR per rep and increasing average contract value by 40%.
The core insight is that AI adoption in sales isn't about replacing reps, but about eliminating non-selling work so they spend less time on admin and more time on revenue-generating activities. When reps spent just 25-30% of their time actually selling, QuotaPath identified every step in the funnel where they were doing something that wasn't uniquely human, then applied AI to automate it. The result was reps doubling their selling time to closer to 50%, and carrying 40 deals instead of 20.
QuotaPath uses two primary tools to achieve this. Momentum automatically updates 75+ CRM fields from call recordings and emails, eliminating the manual data entry burden that used to consume rep hours and created handoff problems between sales, sales engineering, and customer success. Dust, a micro-agent platform, generates business cases, ROI narratives, multi-threaded emails to stakeholders, and customer briefs, solving discrete problems at scale. But both tools required careful prompt engineering and process discipline to avoid the "AI slop" problem where outputs become unreadable noise.
This episode dives into how these tools work in practice, why context matters in handoffs, what makes high-velocity sales easier to optimize with AI, and which use cases delivered the most impact in tripling revenue per rep.
What we cover in this episode
-
0:22
The Results and Business Impact 3x ARR per rep, 40% higher ASP, doubled win rate in mid-market, all without adding headcount.
-
1:37
From Data Science to CRO Ryan's career path through data, marketing ops, rev ops, and eventually taking the reins in sales at QuotaPath.
-
3:41
Reps Spend 25% Time Selling The foundational insight that drove the efficiency gains: identifying every non-selling task and applying AI to it.
-
13:29
Momentum Auto-Updates 75+ Fields How Momentum ingests call data and emails to automatically populate CRM fields without manual rep intervention.
-
21:50
Carrying 20 vs. 40 Deals Automating CRM updates allowed reps to mentally handle and actively work 40 opportunities instead of 20.
-
24:15
Prompts Are Everything Lessons learned on prompt engineering: specificity matters, and data for data's sake is noise, not value.
-
28:53
Dust and Micro-Agents How Dust creates AI agents that pull from CRM, Gong, and web data to generate business cases, briefs, and emails at scale.
-
42:27
Day Three vs. Day 73 Multithreading AI-generated emails bring CFOs and CEOs into deals early instead of waiting 70+ days for executive awareness.
The lines worth sharing
Our notion was, how do we double that and spend more time on the phone, at every step in the funnel where a rep is spending time on something that's not uniquely human, put AI on it.
Ryan Milligan · 3:41
Tools like this are only as good as the prompts. If you don't add the mechanics of what you're looking for, you run into the AI slop problem.
Ryan Milligan · 24:15
I'd rather know that CFO said no on day three of the evaluation than day 73.
Ryan Milligan · 42:27
How can we use AI to make my team more efficient with an output of speed and volume. That's been the fundamental focus.
Ryan Milligan · 12:42
Common questions from this episode
How did QuotaPath triple ARR per rep using AI?
By identifying every non-selling task reps did (CRM updates, business cases, emails), automating them with Momentum and Dust, and freeing reps to spend 50% instead of 25% of their time actually selling. This allowed each rep to carry 40 deals instead of 20.
What does Momentum do?
Momentum automatically updates 75+ CRM fields from call recordings and emails. It ingests call data, analyzes compensation plan mechanics, and populates fields required for sales handoffs and customer success implementation, eliminating manual rep data entry.
What is Dust and how does it work?
Dust is a micro-agent platform that combines CRM, Gong, and web data to solve discrete sales tasks at scale: generating ROI narratives, business cases, multi-threaded emails to stakeholders, and customer briefs. Reps trigger agents manually or via Slack.
How did QuotaPath get CFOs into deals earlier?
Using Dust to auto-generate personalized emails to C-suite executives during early sales cycles. Reps review and approve emails before sending, bringing senior stakeholders into evaluation on day three instead of day 73.
What's the biggest mistake to avoid when implementing AI sales tools?
Leading with AI first instead of process first. Get the human workflow and handoff steps working well, then apply AI to replicate and scale it. Without clear prompts and a defined process, you end up with unreadable AI slop.
Is AI easier to implement in high-velocity vs. enterprise sales?
They're different. High-velocity sales benefit from AI's speed and rapid feedback loops. Enterprise sales benefit more from AI for data mapping and stakeholder research. Both require clear problem identification before selecting AI tools.
SEO meta description
Ryan Milligan, CRO at QuotaPath, shares how AI tripled ARR per rep by automating non-selling work, not replacing sales reps. Learn how Momentum and Dust drive efficiency.
Target keywords
Ryan Milligan
QuotaPath
ARR per rep AI
Momentum CRM automation
Dust AI agents
sales efficiency AI
CRO strategy
rep productivity AI
compensation management sales
business case generation
multithreading AI
sales AI tools
Read the complete conversation
Auto-generated from the episode audio and lightly cleaned. Expand to read or search inside.
Read the full transcript · 59 KB · Ryan Milligan
Eddie0:04Welcome to Go to Market Science. In this podcast, we share tangible, actionable playbooks from the trenches working as go to market strategy and rev ops consultants for our clients here at Union Square Consulting and candid conversations with revenue leaders in the market that have been there. Now let's get into it.
Eddie0:22All right. To another episode of Go to Market Science. We've got a great CRO story today on how quota path grew RR per rep by three x using AI and other things that they did to create efficiencies. Our special guest today is Ryan Milligan CRO, a quota path. Thanks for joining me today Eddie.
Ryan0:41Thanks for having me. Looking forward to this.
Eddie0:43I am super excited to dive into this. So we're going to talk about how quota path tripled AR per rep by leveraging AI and improving efficiencies in each step of the go to market process. About your thoughts on how AI is not replacing salespeople, but making them more effective, which is a common debate on on the web today.
Eddie1:02How you guys increase ASP by and keep me honest about 40% in the same segments. Tell me if that's wrong.
Ryan1:08No that's right.
Eddie1:09And then how you guys doubled win rate in mid market. And probably some other things too, as well as some of the AI tools that you guys use to achieve this.
Ryan1:17Sounds great. Yeah looking forward to it. Cool.
Eddie1:19So without further ado, before we dive in, I'd love it if you could share your background a little bit. You're a crow now you're leading Rev Ops. Before that, you're doing some other cool stuff. Before that. I'd love for the audience just to get an idea of what your background is. That may have helped you figure out how to execute on some of these things.
Ryan1:37Amazing. Yeah. So I'm Ryan Crocker quarter path. I've been at Quarter path for about four and a half years now. Started my career in data science at wafer. Com big SQL junkie and light analyzing and polling data and telling data stories, but then wanted to get more into the execution operational side of the business. So I went to display advertising and marketing performance marketing side of the house.
Ryan2:00And then after business school went to the company called Home Base, also doing performance marketing, and then shifted into more of the rev ops side of the of the world. And then at my time at Quarter Path, I've been here for, you know, four and a half years now. I initially started by leading Rev ops, but we sell primarily to rev ops leaders.
Ryan2:19And so we had a sales leader leave to go do more consulting work. And I raised my hand about three years ago and said I should be a good fit to on the sales team. The team's done quite well. We've been 100% blended quota attainment eight of the past ten quarters, so it's been encouraging thus far. And then over time, I've taken over the leadership of our marketing and I mentioned teams as well.
Eddie2:40That's awesome. It's such a robust background from analytics to marketing to rev ops to now CRO. And we could do a whole episode. And I've done these in the past, as we discussed on how folks go from rev ops to CRL, but I thought it was an even more interesting story on how you actually impacted AR growth. And I think what's most interesting to me about that is it's very tangible.
Eddie3:01You know, I feel like right now LinkedIn is just flooded with these stories about how somebody, you know, implemented cloud code and they booked 87 meetings last week, or they booked 200 leads and you're like, yeah, did you actually book any revenue and you're running a sizable go to market team? You're not some solopreneur posting on LinkedIn about the latest, you know, cloud code hack and what I loved about your post that made me reach out to you, want to invite you on the show was talking about, like, this tangible outcome of measuring AR and the belief that AI is not necessarily replacing sales reps, but making them more effective.
Eddie3:37I'd love if you could expand on that a little bit. Before we kind of dive into the details of what you've been doing.
Ryan3:41Yeah, absolutely. So, you know, one of the prevalent notions that Aret spends 25 to 30% of her time selling, and our notion was, how do we double that and spend more time on the phone? And so the way in which we purchase the problem has been at every step in the funnel in which a rep is spending a lot of human hours doing something that's not uniquely human, like updating CRM, sending multi-threading emails, updating next steps, building business cases.
Ryan4:07There's a long list of these things building sugars on behalf of customer relationships. How can we leverage AI to solve very discreet problems on behalf of the seller? And how do we get the seller bought into the power of AI at solving these problems? And so the way in which we data, we partnered with a couple of different tools on the CRM side, using momentum to automatically update next steps and a lot of other fields required to sell our product.
Ryan4:30You know, with our product, it's there's a lot of questions to ask. So I should have started quote about automates the process of calculating and paying commissions. And so we have to understand the mechanics of your compensation plan, as well as the mechanics of your data set up in order to make you a successful customer. So we have to ask a lot of questions and reps have to feel fill out, you know, 75, 100 fields on a deal before moving it to close to one.
Ryan4:55Right? So AI is a great tool to use to automatically update those fields, build prompts, and be able to capture that data. And then throughout the course of the sales cycle, there's other things that reps want to do. And my question to the reps was, hey, how could you close 30, 40, 50% more deals a quarter? And they said, hey, these are the things that I'm doing that are taking a lot of time building proof of concepts for customers.
Ryan5:19I'm building custom business cases, I'm building pricing proposals, I'm demonstrating ROI. And so we went step by step for each of those problems and said, how can we use AI to deliver that output for them that the human can uniquely use? As we use a cool tool called dust, which basically creates these micro agents that solve very specific problems, pulling in from CRM and from gong.
Ryan5:40And then the reps can interact with that both on a manual nature and an automated nature to generate the things they need in order to provide value to customers and move them through the funnel. We what we say with our sales process is it's a high speed, quick to value, easy to adopt sales process. For a rev ops leader, that's speed to outcome is very important for us and for our buyer.
Ryan6:05And so AI made our reps a lot faster.
Eddie6:07Wow. That's awesome. You jump ahead a little bit. So I wanted to dive into that in a minute. But that's okay. Before I do that, I want to paint some context for folks because we've done a number of interviews like this with Crow's stories, and I think it's so relevant to understand what's the sales team look like. Who are you selling to at Cetera, so that we can paint a picture for the audience and they can compare if their sales motion mimics yours, or if it's different, and put that in context.
Eddie6:31And so I'll start by just asking, can you describe who you sell to and the rough size of deals without giving us your exact like average price point and what segments you target? Sure.
Ryan6:43So this is evolved over the years since I've led the team. But I would say who we typically sell to today is a director or VP of Revenue Operations. One of the main things that's great is they have their data structure in a CRM, like a Salesforce or HubSpot or somebody that's injectable. For us, we're typically selling, you know, SMB through the upper mid market.
Ryan7:04So orgs with anywhere from five to a couple thousand sales reps that that whole gamut. And so from a there's a range of outcomes. But deals can be anywhere from six K to a couple hundred K is the range. But really a lot of our velocities on those orgs with, you know, 5100 reps who want to use their comp plan to drive the right behavior and the right performance from their code market teams.
Eddie7:30Got it. And what's the size and structure of your sales team today?
Ryan7:34Yeah. So I.
Eddie7:35Lead.
Ryan7:36The the broader revenue work. So on the sales side we have a team of seven today on new business. So we have a sales team sales manager. Excuse me. And then with that seven individual reps some report to me and some report to the sales manager. We then have to be doing a lot of work also like AI powered outbound work as well, which we can talk more about.
Ryan7:55And then we have customer success managers. And so they are responsible for implementation account managers who take over the relationship like both in implementation and post, and then solutions engineers who help from the technical side. But from a go to market perspective, my orbit is marketing. So demand generation and and GTM engineering, rev ops sales sales engineering and then account management as well.
Eddie8:20So you got a fair bit going on. So I wanted to ask you like I asked you this on the prep call. Like do you think you already answered this. But I'll let you answer for the audience. Do you think it's easier to implement AI for more transactional, high velocity sales?
Ryan8:35I tend.
Eddie8:36To think so.
Ryan8:37I think the thing that I really like is that you get an immediate feedback loop. So, you know, we're a rep on my team is working anywhere from 30 to 50 open opportunities at a time. And so getting feedback on business cases that have been AI generated or, you know, how the the deals are moving through funnel. There's a there's a high volume which gives a lot of feedback, and you're trying to really drive a fast sales cycle on behalf of the the buyer.
Ryan9:09Right. So we operate in a pretty competitive space. There's a number of tools that automate the process of calculating paying commissions. And one of the things that I pride ourselves on is that I think we are the best buy experience. We have pricing on our website that none of our competitors have. We have a free trial. You can get started playing around in Quarter Path today.
Ryan9:26We don't require a BDR qualification call like a number in the space. Two you can book with an AI today, right? So I'm very focused on the speed of providing value to a to customer or potential customer. And so that is really great from an AI application perspective because I get a lot of data very quickly. And then I can pivot effectively based on that data that I get.
Ryan9:46The A, the AI applications are more enterprise. Sales to me are different. There are things like how do I build maps of large, complex organizations? How do I think about parent child relationships? There are incredible AI applications there. I just find them to be different than the more the relatively higher velocity transactions that we have today.
Eddie10:05Yeah, and it's interesting because I think that's a slightly different answer than what I thought you gave me on our prep call, whereas I thought you were saying maybe a little bit more that you thought it might be harder in more transactional sales. But the hypothesis here is like if you think about like the penultimate AI, SDR, that just like sends out a bunch of outreach and books, a bunch of meetings, one can imagine how whether that's, you know, inbound or outbound, one can imagine how that would be much easier to replace.
Eddie10:34A 22 year old seller fresh out of college was zero experience. That's selling a $5,000 solution instead of your enterprise sales rep. That's been in the same industry for 30 years and has all these relationships, and is trying to get a meeting to pitch a $10 million solution. And even I'm not sure that I believe that I've sold everything from I sold cell phones at college for like 30 bucks a month to I worked in private equity, where we sold multibillion dollar private equity funds, 25, $50 million checks at a time.
Eddie11:03And, you know, when I was doing that, the pushback from the folks I worked with on even the concept of doing anything like what we do in B2B SaaS was, well, this is all relationship driven. Like nobody's writing $50 million check because you've, like, templated some email or hired some 22 or SDR to cold call them. Although to be fair, I was 27 at the time and I was cold calling people, managing billions of dollars, and I was getting meetings.
Eddie11:28And so, I don't know, like how far that goes. But the way that I thought about that, because I was also doubling in ops, was, well, okay, we might only close five deals this year, but how many meetings do we have to get to close those deals? And how many emails do we need to send and how many people do we need to cover?
Eddie11:44And even though we're only covering 3000 people at some point in that journey, you hit this level of volume where you need some type of system and automation. And this was a long time before AI, but now AI, where you talk about mapping a complex organization, that's a tremendous amount of work where, you know, that might not mean that AI is going to go and send the email to the CEO of a fortune 500 company, but wouldn't it be beneficial to have all of that background information before you actually finish that email and press send?
Ryan12:14Oh, absolutely. And I think what I was trying to emphasize was that the applications of AI are very different in the different models. I think for for what I'm focused on with my team, it's all around speed. And how do I competitively present value markedly faster than my competitors? So when when people are evaluating tools in the market, a lot of them will say, you know, quota path is the first proof of concept that I'm doing.
Ryan12:42I'm looking at a couple of different tools, but you all are the first one that's built for me, and that's by design. It's I want to be able to get your comp plan, use AI to analyze it, use AI to build the plan, interpret it effectively, and then be able to show you that very quickly because momentum is the major driver of these deals.
Ryan12:57And time kills all these conversations. And so it's how do I use AI as a tool to make my team more efficient with an output of speed and volume. And that's really been the fundamental focus.
Eddie13:07Yeah. So let's dive into momentum. I'm going to kind of skip ahead here because I had some other questions before we got to AI. But I'll just jump into it because I know this is what people want to listen to. So for anybody that's not familiar, what is momentum? Do I think you kind of answered that already? And specifically, what exactly was the problem that it helped you solve?
Eddie13:24I think you kind of answered that already, but if you could just recap it and go a little bit deeper for the audience.
Ryan13:29Totally. It's funny, I was I was talking about the concept of momentum, the, the noun versus moment of the tool, which I realized I conflated the two momentum, the tool. And a lot of these tools that we use are really focused on building a set of prompts trained off of our data that intake call data and email information, and then take those prompts and take that information to address questions.
Ryan13:54So those are things like filling out next steps automatically after a call, updating the probability of a deal, filling out. What are the descriptions of the compensation plan? What are the variables they're going to need to use? What is the source of data that the customer is going to have filling out these variables that we will need to have to then prepare a customer for successful implementation.
Ryan14:15And so it's just momentum is working in the background to update all of those variables on our behalf. And so as we have tried to make sure that we have really continued strong growth in net revenue retention with our customer base, that all starts in implementation. And so we have to repair our customers effectively for what they should expect and what our CSM team should expect.
Ryan14:33And so getting that information effectively prepares us to be able to have those those conversations. And we can go more into depth to prepare for the implementation of the House.
Eddie14:44I love that, and it's so interesting because like pre AI, this was many years ago. And it's not the only time that I've seen this, but we had an absolute train wreck of an engagement where we were working with a customer, specifically with the finance department, and at the request of finance and specifically the CFO, they wanted these 35 fields filled out on Salesforce.
Eddie15:04Every single time a wrap got off the phone, we pushed back and said, this is not a good idea. But of course, like, you know, they made the final decision. So we did that. And finally the CEO stepped in at some point because the reps were absolutely rebelling. And then he fired us with some really nasty language, fired the controller, fired the CFO.
Eddie15:26Basically, everybody involved got fired because the reps were like, what person built this? And you think pre AI and like this is something I feel so strongly about. You have to be so choosy in what you ask refs to do, because everything you ask them for takes them away from that valuable selling time. But what you're saying is that you are now able to capture those 35 data points that are really critical to managing your sales process and or your onboarding process, which I want to dive into, because you've got AI doing all that work, and the rep can still focus their energies on doing what they do best, which is selling and talking to
Eddie16:03customers.
Ryan16:04Exactly. Like if a rep doesn't have to be thinking about taking the notes on what CRM are they using, what ERP are we joining to do? They have cliffs or draws or other mechanics of their plan? If they're just asking those questions and the customers describing it, but they're not having to spend that much time thinking about it, the rep can spend more time thinking about creating value for the customer, and then you would have a situation in which a call would end, and the rep would be furiously inputting those fields into CRM, where instead I would rather have the rep thinking through what's my multithreading strategy?
Ryan16:35How do I send our new product for them to play around with? Like, how do I create actual value versus being a note taker? And so that's that's been crucial for our ability to give our reps the opportunity to close work and close more deals per quarter. Now.
Eddie16:49Let me ask just out of curiosity, why is this important? Why why do you need those 35 fields? Like it's great that you've got AI creating this efficiency so that reps can spend more time selling. But another way to slice this is just don't ask them to ask that information. Why is that so critical that you had to find a solution to it?
Ryan17:06Yeah. So context is massive in the handoff process of our customer relationships. Right. So if you think about the ways in which someone potentially becomes a customer, they could talk to 3 or 4 people in their first year of a journey. Working here and not being able to hand off that context gets the customer quite frustrated, right? Because they feel like they themselves from time and time again.
Ryan17:28And then it makes it really challenging for the person getting that customer to give them value very quickly. Right? We want people to be able to build their compensation plans and pay their team very rapidly in the product once they're signed. And if a CSM gets on the phone, post sales and has to repeat all the same questions about comp plan design and the mechanics and the triggers, that's taking away from time to give the customer value and actually use the product, right?
Ryan17:53If you think about a journey of a customer, maybe it's that we host an event, a prospective customer talks to myself or a BDR or someone at that event, we gather some information that we capture. Then we open a sales process with them. We do a first demo, then we do a proof of concept. On the proof of concept, we'll pull a sales engineer.
Ryan18:13Okay, well, now I got to get the sales engineer up to speed on this. Then we're closing the deal and removing it to a customer's assessment manager who's taking over the implementation. And then after the implementation and account manager is taking over. Right. So you're talking about a lot of different people who are involved in the process. And if you don't give them the information they need, you don't set them up for success to deliver value to the customer in a short period of time.
Rachael18:34Quick pause. Everything we talk about on this show diagnosing go to market ops. Prioritizing projects for revenue, impact processes, metrics, insights, building a predictable go to market engine. We've built frameworks for all of it. They're free and undated on our website, Union Square Consulting. The link will also be in the show notes, so make sure you check that out.
Rachael18:57All right, back to the episode.
Eddie18:59You know something that resonates with me so much? Because when I was at Salesforce, I think they did a pretty good job of not asking us for a lot. You know, there are only like 3 or 4 fields. We had to keep updating on the pipeline because they wanted us selling, but that was management. When you come to the SC, it's an entirely different story, right?
Eddie19:13So at least for us, I had at some point in time one SC that was covering like 4 or 5 teams, like not just for A's, for teams of A's. So you can imagine like how discerning they were with their time. And it's not uncommon that seas are a very rare resource. And so they would push back and say, well, if you haven't gone into Salesforce and updated all of these different fields, I'm not even going to join your meeting because I'm not going to spend an hour listening to you regurgitate your entire deal, like, I need you to document this so that I can prepare quickly.
Eddie19:43Right? And so this took an immense amount of time, but it was necessary, especially these were more of the bigger strategic deals where we would pull SAS in, and they needed all that color to do a good job of doing discovery and then setting up a demo or proof of concept or what have you, that we needed to win that deal.
Ryan19:59And I mean, I think for us, we sell a there every customer relationship is unique a quota path, right? Every customer has a unique compensation plan. They have unique compensation plan logic. And then they also have a unique data sources and ways they architecture data sources. One customer joins Salesforce and NetSuite. The other customer joins HubSpot and QuickBooks.
Ryan20:18Like they're all using bespoke data structures. And even within those data structures. No, HubSpot is the same among the hundreds of customers who are using HubSpot. So they all have different names of fields and different naming conventions and different structures. So you there are a number of tools that you can sell where the sales process is pretty much the same.
Ryan20:37It's a widget, right? You sell the widget, it clicks in, it automatically does everything you need it to. You don't need that context for Tool of Ours, where you're joining both unique bespoke compensation plan logic and unique bespoke data architecture. You have to have that context or you're not setting anybody after success.
Eddie20:54I think that's a really important point, right? So there are some sales processes that are just very straightforward, and you're kind of selling the same concept. And there's others. Our business is not different than yours in that respect, where there's so much information and data and nuance that you need to collect. And if you don't have any help with AI doing that, then you're just stuck sitting there looking at notes and looking at call transcripts and spending hours and hours connecting the dots.
Ryan21:19Exactly.
Eddie21:20So where do we go from here? Let's let's talk about dust. What does dust do? And actually wait. Hold on before I move off of momentum. Sure. One one other question is, did you see it? So I understand how having all this data in Salesforce for momentum really sets you up for success on a on on customer success, on onboarding.
Eddie21:42And maybe you've already answered this, but did the use of momentum actually help you guys improve your win rate on actual deals?
Ryan21:50Yeah, so it really helped. It was less a win rate driver and more a volume of deals being able to work driver. So it was just a you know, it basically allowed us to stretch limit on a per person basis of how many deals they would consider actively working at a time. So that went up about ten, 15% once we rolled out the tool.
Ryan22:10And you know, you have that limited touchpoint with a rep where she says, okay, I have 20 open deals. This 21st one comes through in a pre AI automated world. The barrier to keeping that deal going is much higher because they carry the mental burden of like logging all those notes. But if every deal, everything is automatically logged for you, you want to handle and navigate, you know, 30, 35, 40 deals.
Ryan22:34And so I guess that does come into its less win rate on a per deal basis, but it's volume of AR close per rep per quarter as a function of them being able to carry more deals to fruition because AI is automated. One of the notation.
Eddie22:48Yeah, that makes so much sense. And then also like by extension on supplies to you, but for other people, when you save time on the pipeline management piece, then you have more time to spend prospecting or multithreading or what have you. And I know for me, in every job I've ever had in sales, I've always had this point where I open up Salesforce or whatever CRM I was using, and you look at your opportunities and say, it's 15 and you're like, I'm now at my limit.
Eddie23:12I know this is going to take me X amount of time to go through each deal, figure out what's going on. And it's not just admin. It's like admins. One thing you appease management, but the other thing is just I need to know a certain amount of information before I can pick up the phone or send an email to follow up, and without any help, it just takes X amount of time.
Eddie23:28So streamlining that now saves time that now that I can go prospect fill the funnel with more pipeline and then I don't hit that point where I can't sell more as quickly in terms of the number of deals. So it makes perfect sense.
Ryan23:41Yeah, exactly. It's a what is the mental load on a per deal basis and how many can I carry at the same time. And, and we felt like we had a good opportunity to improve that further. And that's what has proved out for us.
Eddie23:51So are there any big or takeaways for the audience if somebody is thinking about implementing momentum, for example, I don't want to make this a plug for momentum, even though ironically, I have a call with them in an hour or two. But, you know, not getting paid for this, I promise. But if somebody is listening to the show and they want to go implement momentum, are there any key learnings that you've taken away that are necessary to make it a successful?
Ryan24:15Yeah, tools like this are only as good as the prompts, right? I think that's the biggest one. So you have to have a very, very good understanding of like what fills out a field. So we have a field that's current company description. And if you don't it took us a while to get that field in really good shape.
Ryan24:29Because if you don't add the mechanics of what you're looking for, you can pull in a lot of extraneous information and you run into like the AI slop problem, where you're just dumping, you know, bullets and bullets and bullets of like, not useful information. And so these tools are only as good as the problems in which you set them up with.
Ryan24:45And it is important to know exactly what you're going to do with each of these fields, because the more fields you're able to update, you also have a bit of a burden of why am I keeping these fields? Because all of a sudden, if you're saying, you're saying, okay, AI can update 75 or 100 fields, you should then put the mental pressure on.
Ryan25:04Okay, do we actually need 75 or 100? Can these three combine because yes you've used AI to update 100 fields. Now I'm a CSM and I got to go read through 100 fields with notes. Right. So it it almost opens your ability to do so much. Then you then have to be more exclusive and explicit with what you actually need to be up to be successful.
Eddie25:24Yeah. And I think that's just sort of like a one on one rev ops thing where there's always this temptation and ops to go build the next thing and you got to ask like, how exactly are we going to use that thing to generate and close deals, right. Because if there's not a tangible use case then it becomes noise.
Ryan25:39Exactly. And I've been guilty of that myself. Right. Putting my hat on. There's, you know, some fields that I've built in the past that I look back on now. It's like, why? Why do we possibly need that? And I think your access to that data, like data for data sake, is not valuable. And so that's where I advise people to push back on themselves as well.
Eddie25:58Yeah, absolutely. But with that I think like it's not just the prompts. I don't know if this applies to how you guys use momentum, but it's also first, I think it's the process. And then I think it's also the adoption of that process. So there's elements that you can't automate with AI. There's things that like yes okay we've got call recordings.
Eddie26:16So maybe reps don't necessarily need to write down what they're discussing on the phone. You've got emails. But then you know there's always at least for me, there's always something that slips through the cracks. Like I was, you know, texting somebody that I'm working in active deal with right now. And I'm like, shoot this text, no matter how much AI have, is not getting its way into the CRM.
Eddie26:34So, like, is there anything that you faced with your implementation of momentum where you said, like, I've got to get this process in place, the human process, I've got to get my team executing this before that is going to be effective before the AI is going to be effective. Yeah. I mean, I.
Ryan26:48Think, you know, how do we take these fields and put them in a structured output in the hand off node perspective for the next person in line? That was super important. Like what are the fields you bucket in certain sections. And how do you lay that out for a way that if a CSM steps in, they're not just reading this like long laundry list of things.
Ryan27:06And that's also kind of the reason why we paired to like, momentum with a tool like dust, which is how do you create these a really shareable briefs internally that take all this great data that you've captured and structure them using AI in a really like, digestible way to solve problems? And so that's another piece of this is you can we spend a lot of time creating the perfect handoff note in CRM.
Ryan27:28And then we realized, okay, but that hand off note might need to flex based on the mechanics of the deal and like how difficult the implementation is going to be or how complex the data structure is. And we want that structure to be able to change. And so that's why we've brought AI as well into the handoff process from point to point.
Ryan27:44So it can be flexible based on the different needs of a different customer at a certain time.
Eddie27:49Yeah, that makes a lot of sense. And what I'm hearing is that you spend a lot of time trying to perfect that handoff process, like what specific information needs to go from sales to CS in what format, and then you can use the AI to figure out how do you deliver that exactly?
Ryan28:04And you know, this data is only as good as it gets applied by the people who need to use it. And so we had a couple false starts on this process where, you know, in the in the kind of initial AI slop days of this where we thought we built, we thought this beautiful hand off note, but it was just like five pages long and completely unreadable.
Ryan28:25And and that was prompt tuning. Right. But you realize that if you get that wrong, you're handing something off to a CSM who doesn't feel compelled to read it. And then both sides are frustrated, right? The CSM is going back to the AI asking questions because they don't have the answer they need, and the A's frustrated because they feel like they've gathered all the relevant information.
Ryan28:44And that's not valuable, right? And so how do you how do you make sure that everyone's read into the process and feels like they're moving along in the journey effectively with you?
Eddie28:53Yeah, it makes a lot of sense. Before we move on, anything else you want to say about momentum for somebody listening to this? And they're thinking like, I want to implement this, how do I avoid failing at it? How do I make it a success?
Ryan29:03I mean, I think you have to have a really clear understanding of what you want documented before you start with any of these tools or opportunities. Is walk me. Walk yourself through what the perfect set of handoffs is, what is the exact data you need and pressure tested, and then go to the people who are ingesting it and say, if you got this laid out in this way, would you be happy?
Ryan29:24Yes or no? Would you have what you need? And that, I find, is really helpful before you even go to any of these tools, because then you know what your target output is. You want to have your target output in mind before you inject AI into the process, because if you don't, you'll end up with an output that's not serviceable by anybody and leads to a lot of frustration.
Eddie29:42Yeah, I couldn't agree more with that. And then I would even make the argument and feel free to disagree with me. If you just don't agree that you've got to get this fundamental process, this human process down, like the shit that doesn't scale, so to speak, before you go to the AI. So what I'm hearing from you is that you did this with humans.
Eddie30:00First you got the process in place. You said, okay, this is what a good sales handoff looks like. Maybe it wasn't perfect, but it wasn't total like it wasn't a total mess. And then you say, okay, now we've got something working with humans. How do we get AI to do that? And then once you do that, then you still have to iterate to get the AI to do the thing the way that you need to needed to be done.
Ryan30:20Yeah, exactly. And we had we had sales engineers that were spending way, way, way too much time in the pre. I like nature of this, but it did work right. So we just filled that with more human hours initially. And then eventually we said okay, how do we inject the to make these team members more efficient. And then sales engineers could work more deals on a per person basis.
Ryan30:37And so at the reps. And so yeah, I agree with you 100%.
Eddie30:40Yeah. And the reason I'm like diving into this is I think that there's so much FOMO right now with AI and every single CRM and every rev ops leader feels like we need to implement 15 new AI tools. And there's this, there's this. It feels like, you know, running around with a hammer, looking for nails, and you're trying to implement these AI tools without thinking, like, what's the thing that is actually working to drive revenue that we could then apply AI on top of.
Eddie31:03But like we go with ACR and maybe we don't have stars to begin with, or we don't have an STR like team that's actually functioning properly. And so now we're taking wild guesses at what might work, when we could just try it with humans and get it, you know, working, and then apply AI to replicate that, that workflow that has actually shown to work.
Eddie31:23Yeah. I mean, it's.
Ryan31:24You've reframed the question. I feel like throughout 2024 and a lot of 2025, the question was, what problems can I solve with AI? Or was like, like, how do I inject AI into this process? Which is not a helpful statement to me. The question is, what things are taking my team too much time to do on a day to day basis, and can AI solve those for them or make those faster?
Ryan31:45Like it's not? To your point, it's not leading with, okay, I'm going to inject AI into this process for the inject AI into the process sake. It's let's get a discrete set of problems. And my team are spending too much time on that are not creating value for anyone involved. And how can we use AI to solve those problems?
Ryan32:02And that has been very beneficial to me. And I think what a lot of people forget is at the end of the day, you still need the human to be bought in. And so if you're just inject AI into a process and make it clunkier, you still you really risk alienating some really strong sellers on your team. So doing a lot of interviewing early on to say, where are you running into snags in the process?
Ryan32:22Walk me through yesterday. How did you spend every hour of the day? And if they said, I spent two hours building a proof of concept for a potential customer, that is a red alarm to me to say, okay, how do I use a tool that we have to ingest that comp plan and build it for you so that you could spend those two hours doing something that you can uniquely do to create value for them.
Eddie32:42Yeah, I love that. And I would even, I would even almost posit a different lens on this where I would say, or I would ask, rather than like where reps spending time, you're essentially making the assumption that the number one thing that you need to solve for in the business is saving reps time. And then I would even ask, why is that?
Eddie33:01Well, because I want each rep to produce more. Why is that? And ultimately what we're trying to do is I mean, in some ways it just comes down to finance like dollars and dollars out. Like I got X amount of dollars to put in to go to market. How can I maximize the output in terms of revenue and profit that I produce with that?
Eddie33:19And then that comes down to, well, I've got X budget for headcount. And so one really effective way of leveraging that budget better is to make each rep more effective. But if we, you know, at the end of the day, like, would you buy a company that has no AI if they're just crushing it financially? Like, okay, we have 100 sellers and they're super inefficient, but somehow some way, you know, we take a dollar and turn it into $5 every single year.
Eddie33:44Okay, well, every investor on the planet wants to buy into that company. Most of those companies don't exist right now unless they're like an AI native startup right now. But that's ultimately the lens that I want to look through. And then you drill down and you say, okay, well, how can we sell more? And then I would say, well, okay, I could do everything you're describing by just hiring a bunch of Vas in the Philippines.
Eddie34:05Okay. Well, why would that be better or worse than AI? And I think there's really strong arguments for why everything you've described makes perfect sense to do. But I'm just trying to present a different lens where, well, if I and I do have Vas not in the Philippines, but in Latin America, where, yeah, there's probably a way that I could use AI to do that, but it's something they're spending five hours a week on, which cost me, you know, not a lot of money.
Eddie34:27And so is that really the problem that I need to solve for right now? Yeah. And I think the.
Ryan34:33I like the way that you think about it, it's like I need to make these reps more efficient because I need to, like, positively improve my cost of sales. And it's an interesting point where my like, ethos and all this stuff is I'd like to run the smallest team possible. Like, I like small teams. I like small, efficient, high performing teams.
Ryan34:52The more people you add and injecting to a process, the more extraneous risk you have, right? Every new person has a, you know, a range of outcomes that you can't necessarily model or control. And so I really like small, efficient, high performing teams that win, that hit quarter sustainably and that have high momentum. And that is the thing that I think about is like, you get a business into a rhythm.
Ryan35:14It's how do you then maximize the per FTE. Right. Like one of the things that I think a lot about with this business is how can we grow revenue markedly faster than we grow headcount. And I feel like back in the day, the sign of running an incredibly successful revenue team was running a big revenue team. It was I have an 100 person sales team.
Ryan35:34I have 50% customers team. And that was like the badge of honor. And today I feel like it's the inverse, like you want the smallest team you can possibly have to hit incredibly high targets because then they win. Everyone's together. You build the sense of morale and momentum that really wins long term.
Eddie35:52Yeah, I mean, there's so much to unpack there and I love that. And in some ways I look at that and I'd say, okay, I completely agree with you because everything you mentioned, it's so difficult and expensive to hire people to manage people. You have a limited success rate on an individual basis, et cetera, etc.. So the smaller the team that you have, especially if that team becomes small because you're just keeping the best players, then to me that makes a lot of sense in terms of just driving a profitable, you know, high growth business.
Eddie36:23On the other hand, like I have like mixed feelings about the whole AR or per rep because or per FTE, I mean, you look at like a company like Goldman Sachs, I remember many years ago they were they were pumping out like $1 million per FTE. And I'm like, okay, well, sure. But, you know, half of these people are making, you know, 300, 400, 500 K.
Eddie36:42And like Goldman Sachs is a very valuable enterprise. But, you know, since 2008, I haven't heard anybody talking about them as like the darling of like, you know, Wall Street in the sense of an investment opportunity. All anybody wants to talk about are these high tech startups. So like cool. Goldman's got this extremely high revenue per employee. But if I can have a 10th of that because I've hired a bunch of folks at 50 grand a year that just crank out revenue.
Eddie37:09Okay, well, I'm winning, right? So anyway, I'm not sure what point I'm trying to make. I'm just kind of, you know, going down a rabbit.
Ryan37:14It's a good I think you have to blend both RF to E with like overall margin and cost profile of the business. Right. And like burn and profitability I think that I just there's a lot of metrics that run into it. I like Arpa feet as one because it says when you model forward, how many people do we need to get to this level of RR?
Ryan37:31And I just think that's a fun exercise to say, how could we double that? And I think that, you know, and if the answer to your point is just pay them two times as much money. Okay. Well that's not the answer, right? You. I do like to invest in my team and like to give them raises, a function of their performance, and raise oats pretty consistently.
Ryan37:50But you have to raise R growth, obviously much more dramatically than you're raising oats or you're just conflating or like, you know, inflating a problem.
Eddie38:00Absolutely. Well, let's get back into it because I know we're short on time. Tell me about dust. So first what do they do? And what was the problem that you guys were trying to solve with dust?
Ryan38:10Yeah. So it's interesting. So momentum solves the problem of, hey, let's gather all this information and write it to these fields. In CRM, dust solves the problem of I have a discrete set of tasks that I want to do that are taking me a long period of time that I feel like AI can automate. So those are things like, I would like to write a multi-threaded email to every CFO, Crowe and CEO, evaluating of companies evaluating code path to add, particularly our unique value from the sales process that's going to take a rep a long time to do.
Ryan38:39But we built a dust agent that goes and looks at the LinkedIn's. The profiles of the organizations then pulls a lot of like contextual references from the ground conversations, then pulls similar customers from CRM and writes a bunch of these emails. Right. And so what I like about dust is it combines the number of the mic, CRM, gong, MCP, like all the relevant data that you're going to need, and then is able to be trained to solve like very discreet tasks, and reps can operate it both in dust as the tool.
Ryan39:11They can also operate it out of slack, so they can call within deal rooms that we have at dust. Can you go build like an executive brief for this call? And so there's just a number of ways in which we wanted to present information for different people, for different contexts that we wanted to do at scale, that we just didn't have enough hours in the day to do.
Ryan39:30So if you think about our account management function, we want to be able to generate a QBR every quarter for every customer. Okay. We have over 1000 customers at this point. Right. So kind of mechanically impossible to use humans to generate a QBR for every customer every quarter. But if you have an agent that pulls in their usage data, pulls in the context of recent calls, pulls a new product, releases and generates a QBR that sits and is able to be sent.
Ryan39:55Okay, that presents a lot more value. So you're basically taking all of the relevant contextual data. You're using these prompts and asking for these certain outputs, which typically are PDFs, URLs, PowerPoint slides, what have you that solve the right problem for the right person at the right time? And that's how it's been impactful for our business.
Eddie40:15That's incredible. So I'd love to double click on the multi-threaded email generation. So first question is just how do you get the data. So one obviously LinkedIn is notorious for being really protective of their data. So how does dust even get access to the data on LinkedIn.
Ryan40:31Yeah. So the way that we do it is will the I'm trying to think the right way to approach it. So basically yeah, we don't do any automated subdirectory on LinkedIn. So we'll actually like hit the website and look at the like about Us page and some of the founders. And then it'll gather the information that the ripple.
Ryan40:48Then go find the email addresses. Usually because we don't want to automate that. We don't want to automate on LinkedIn because it's like very penalizing. And so it's usually looking at like the company page or prompting the rep based on like press releases and other things. This is who we think this person is. So go look them up and like pull their relevant information.
Ryan41:06It's finding the person who it is is actually not a very taxing problem, like finding the name of the person and their email address. So you can go into LinkedIn and gather information. It's more right in the contextual email to that persona based on the call information, based on our similar customers and that sort of stuff. That was actually the most taxing part of the process.
Ryan41:26But yeah.
Eddie41:27I felt you were saying I thought you were saying like you'd go in and like, read like the post that I read of yours that prompted me to reach out to you. And you're basically automating the process of taking that post and turning that into a relevant email.
Ryan41:37No, it's more mid sales cycle stuff. So after a first demo or during a validation process, I want to make sure that the CEO, the CFO and the Crow and the VP of all like aware of the evaluation there where of our our unique value props. They understand why working with Coda path is like a great fit for the organization, because you do find that especially in modern staff sales, you have a CEO or CFO lingering in the background ready to say no to a purchase the whole team is excited about.
Ryan42:09And a lot of times that's because they weren't aware that the evaluation was happening in the first place. You know, a rev ops managers go taking a look at, quote, a path. They're building up this business case. They go presented. The CFO says no, we have no budget for any tools right now. And then you've done a lot of work for no output.
Ryan42:24How do I go to the CFO earlier in the evaluation and say, hey, this is how we could partner together. This is some information that could be relevant, what have you.
Eddie42:31So how did you go from concept to actually sending the email? You know, part of that I want to ask, like our reps reviewing these emails before they actually press send or is this fully automated? And how did you how did you get from the concept to okay, we now have reps pressing send on many emails that we're very comfortable with.
Ryan42:49Yeah. So a lot of the multi running actually goes through me. And so that's like one part of the process. So I run standups with our team. We run a stand up every day. And each day we saw we talk through a different problem. So you know on Wednesdays we'll talk through what are the opportunities you need to be validated by our sales engineers.
Ryan43:07What are the statuses and what are the priorities? For example, on Thursdays we talked through multiple. And so we the reps come prepared for me with a set of emails that I, our CEO, others of the organization can send to peers that the reps have written, they've reviewed, they've used dust to write and they've reviewed, and they come to me and say, hey, I want you to reach.
Ryan43:27I want Ryan to reach out to the CFO at Jones Software. This is the email that we've written. It talks about some similar customers we work with. It talks about, you know, unique value props. This is the email. I've reviewed it. Ryan, here's the email address of the person. Here's the context for the deal. And here's the email you should send.
Ryan43:45And so we have human intervention points like baked into the process. But it's allowing us to be able to at a much higher volume, bring in senior stakeholders into our deal process, which has been massively impactful because you can get ahead of that CFO who said, oh no, we have no budget for this. I'd rather know that on day three of the evaluation than day 73.
Ryan44:05Right.
Eddie44:08What have you guys done? Or how have you handled policy nations? I mean, I think anytime you're pulling this very large set of data, there's always a sphere that it hallucinates and and gets the data wrong. And then how long would it take you to go through all that mountain of data to validate it? I'm curious how you tackle that problem.
Eddie44:24Yeah. So the.
Ryan44:25Hallucinations have has not been a huge problem for us. You know, what it's trained off of or what it's reading is gong calls. It's reading like industry relative information. So we haven't had a huge hallucination problem. Also the the risk of hallucination in an email is not like my primary concern. You can quickly just make sure that the customers that are the referencing are similar, like in a similar industry to the customer you're talking about.
Ryan44:51But even if you were to miss on that, that's not like the most devastating miss to me. So there are some things like this multithreading where I'm willing to take a little bit more of a hallucination risk, but we haven't seen that as too much of a problem.
Eddie45:03Yeah. No, I think that's fair. Awesome. Well, we only have a few more minutes left, but I dove into one specific use case. What are some of the other most impactful use cases you found with dust?
Ryan45:15Sure. So if you run through the entire sales process basically from start to finish, we have specific agents that solve each of these kind of problems. So you run a first demo, how do I send a like ROI in business case. So pulling in like ROI calculation tools we built and then pulling in their relative information to present a business case.
Ryan45:35So how do you auto generate those. How do you then think about the multi-threading process which we talked about. How do I then build a go live plan for a buyer based on the information that we've we've had, what is going to be the mutual action plan for go live. And you build that in a really interesting way.
Ryan45:50How do I then build hand off notes for multiple people in the organization? So can I automatically build a brief for a sales engineer to prepare her for a call? Can I build a brief for and exec who's going to hop on a closing call to give me, like exact level, important information that I would need? Do we have mutual investors?
Ryan46:07Do we have like mutual relationships, some of that stuff that's kind of publicly available from there through the process, you can do a sales engineer or sales engineer to solutions engineer handoff or a repped a CSM handoff. And then through the account management process, you can do things like account briefs, QBs pulling in usage data. So basically every step of the process you can have these certain problems and that we're solving discretely with these agents that are, you know, the kind of primary agents, you know, an account management agent, a go to market agent, and then sub agents that are built to solve some of these specific problems.
Eddie46:44That's awesome. All right. Well, we've only got a couple minutes left now and I want to cap this off. Let me go back to the original pitch here for this podcast, which was the results that you were producing. So tripling AR or per rep, increasing ASP by 40%, doubling the win rate in mid market. Of all the things that we discussed or anything we didn't discuss, if you had to pick, what are a few of the things that you felt were most impactful in helping you achieve those efficiency gains?
Ryan47:12Yeah, I mean I think the a lot of the auto updating CRM field subgroups, a lot more times we able to close a higher volume of deals that played a very large role in this process, which probably be the primary, and then the secondary is a lot of the business case and ROI generation, both pre and post sale.
Ryan47:29So being able to have a very clear ROI narrative that's trained off of the customer's data, that takes what they say and apply it to how they're going to be able to buy. It allowed us to create a lot of by in on deals that could have said, I can't make the case for this right now. Now's not the right time.
Ryan47:46You can build around the cost of an action and talking through the cost of an action to a buyer, and that FOMO of not buying has been very impactful for us and been a great value proposition in the medium term.
Eddie47:58That's awesome. Well, we got deep into the weeds on this, and like some of the answers you gave, like I was genuinely curious myself on how you navigated these common challenges. So thanks for sharing this. Of course, if people want to reach out to you, how can how can people find find you or find more about you or your content, your company, etc.?
Ryan48:15Absolutely. So I'm pretty active on LinkedIn posting a couple times a week, so definitely check me out there. And then we're going to be kind of a sneak peek. We're launching a new product in about a month or two that I'm excited to get people's hands on, which will be I'll be posting a lot about it soon all around compensation plan design and like go to market efficiency.
Ryan48:33So more to come there. But I'm very excited for that.
Eddie48:37That's awesome. Well thanks so much for joining me today.
Ryan48:39Cool. Thanks to appreciate that.
Eddie48:41Thanks for listening to the episode. If this resonated, please give us a five star rating and a follow. It helps us reach more people, and you get our latest and greatest content without having to search for it. And if you're looking for hands on help and go to market strategy and or rev ops, please reach out to us.
Eddie48:55We help our clients with everything from annual planning to improving processes and go to market, implementing systems to support those processes and go to market AI. We're always happy to offer a free consultation to help you identify the best opportunities to improve your go to market engine, with or without our health. You can find us at Union Square Consulting and the info will be in our show notes.
