Kris Lawson
Ruby AI
Kris Lawson, CEO of Ruby AI, has built and scaled sales organizations at Oracle, Salesforce, Microsoft, Snowflake, and Vidyard. He lost a deal to Salesforce as an Oracle rep, got recruited to build Salesforce Canada from 4 people to 800, then built Snowflake Canada from zero to $35M ARR. Now he’s building an AI agent designed to automate the work that sellers used to do. In this conversation, he breaks down what the modern sales hire actually looks like in an LLM-dominated world, why coachability beats pedigree, and why hallucinations in large language models are a permanent feature, not a bug.
In 2004, Kris was selling for Oracle when Salesforce entered the enterprise market. Oracle’s CRM was expensive, difficult to use, and required painful data center setup. Salesforce won the deal. Instead of dismissing the loss, Kris did something most reps don’t do: he read the signal. He analyzed Salesforce’s S1, understood why they were winning, and presented the findings to Oracle leadership. Days later, a recruiter called about opening Salesforce’s first Toronto office. He said yes, and walked into a 4-person shared office space.
Sometimes these things just happen and you have to read the signals and look to be comfortable with that and see where you go.
Kris Lawson, Ruby AI
The lesson isn’t about joining winning companies. It’s about noticing where the market is moving and having the courage to follow.
From Salesforce’s early days, Kris hired across backgrounds: career enterprise reps, startup sellers, and recent graduates. The company invested heavily in onboarding, running nine Sandler training programs himself. What separated the people who stayed and rose to VP and SVP level from those who didn’t? Coachability.
He didn’t need MBAs or Rolodexes. He needed people who could listen to guidance, ask clarifying questions, and execute with rigor. The best reps were the ones who thrived on being coached, who stayed disciplined with process, and who didn’t fight the system. Process rigor, especially around CRM discipline, was non-negotiable. Sales reps back then weren’t used to being CRM-first. The ones who adapted fastest and most completely were the ones who climbed fastest.
Coachability is incredibly important. Energy, confidence, fearlessness, high acumen for understanding a customer’s business. Intelligence. If you put those things together, that’s a great fit.
Kris Lawson, Ruby AI
Snowflake Canada launched in 2018 with zero revenue, zero partners, and two salespeople. Kris was up against 400+ Microsoft, AWS, Teradata, and Cloudera reps. Building a partner army fast became critical, but hiring the wrong profile was the killer mistake.
The temptation was to hire from sales academies: people with rigorous training, proven playbooks, and impressive pedigrees. But there was a hidden gap. Academy grads at mature companies had learned to cross-sell and upsell within established brands. Snowflake needed the opposite: pipeline generation savagery. Pipeline generation savagery meant finding 5 new prospects every week, prospecting into unknown accounts with no brand recognition, and doing the messy work of new business development when the company name itself was a punchline in 2018.
Kris found that hiring directly from a sales academy to a startup role without intermediate experience had a 90% failure rate. The reps had the skills but not the hunger or the scars. The ones who succeeded had worked at another startup first, lived through the scrappiness, learned how to generate pipeline without a brand, and then joined Snowflake with realistic expectations. That intermediate startup experience was the missing piece.
If we go directly from the sales academy to what we needed at Snowflake, it was nine times out of ten a complete fail. They had to have cut their teeth at a startup first.
Kris Lawson, Ruby AI
Kris has advised Gong, Maxa, and Axia through their early sales hires. Founders make the same mistakes repeatedly. First, they hire from a sales academy without realizing the gap. Second, they step back from sales entirely, assuming the new hire will just take the work off their plate. That hands-off approach breaks the feedback loop.
The third mistake is assuming founder magic translates to a sales deck. Founders have inherent credibility—people want to be near innovation, near the founder’s journey, near the origin story. But that magic doesn’t live in a first call deck, a website page, or an outbound email. It lives in the founder’s depth of knowledge, confidence, and ability to hold a room.
The answer is to break the story down to its base principles. Who is the ICP? What are they trying to do that they can’t do today? What have they tried that failed? What will you do for them, and why is it different? Why should they move now? Get all stakeholders—product, marketing, other founders, early sales reps—to agree on that story. Then build assets around it. The first call deck is the most important: it keeps the rep on story during the meeting. Enforce story adherence by listening to calls, by doing deal reviews where you ask reps what the customer said in response to your hypothesis, and by looking for the curiosity signal.
Founder magic doesn’t easily translate into a first call deck, website messaging, or outbound messaging. You have to break the story down and get everyone aligned on it first.
Kris Lawson, Ruby AI
Kris’s opening claim is blunt: 90% of sales cycles will happen inside LLMs within 24 months. Buyers will evaluate you through their LLM. Attention spans are shrinking. Information asymmetry is gone. Sellers are being written out of buying cycles because buyers no longer need them for research.
But there’s a counter-trend: face-to-face still matters. The bar has risen. When you get in the room, you need charisma, curiosity, likability, and concrete insights specific to that customer’s business. You need to be someone they want to see again. The problem is getting the meeting in the first place. Every touch has to count.
The modern sales hire needs to handle the full cycle: SDR work (finding accounts, building sequences), AE work (selling and closing), and CSM work (ensuring adoption and expansion). AI now does 30–40% of the back-office work that used to be split across three roles: research, CRM updates, sequencing, planning.
Test for three things during hiring. First, pipeline generation instinct and passion. Ask for specific examples of how they’ve built pipeline in past roles. Not checkbox answers. Real proof. Second, CSM instinct: tell me about a customer who was failing and what you did to save them. Third, being “AI-pilled”: what AI tools do they use today? Are they in Claude? In saved prompts? Are they using AI as a thinking partner or just copying the first draft?
The sales rep has to show they can run the full cycle because AI does the work that used to take three people.
Kris Lawson, Ruby AI
The hardest test is curiosity and resistance to cognitive surrender. Cognitive surrender is the trap where people defer to their AI completely, stop questioning it, and treat it as a decision-maker instead of a thinking partner. Leaders need to meet with reps before calls and ask hard questions: What’s the biggest objection they’ll raise? How will you solve it? What’s your top question for them? How would you phrase that differently? Force curiosity back into the conversation.
The old model is breaking. A first-line sales manager managing 6 reps is not sustainable in an AI-powered world. The right span is 12–14 direct reports. The SDR-to-AE ratio is inverting: one SDR used to support three AEs. Now one SDR can support six or seven, because AI does so much of the sequencing and research. The SDR becomes less of a meeting generator and more of a go-to-market engineering coach, showing AEs how to use AI tools effectively.
Quotas will double. Sales rep yield has been artificially low for decades. A thorough account review—finding stakeholders, understanding use cases, triangulating with company news, building a real hypothesis—takes four to six hours per account. If a rep needs to break into 50 accounts a year, that’s 240 hours just on research. That’s six weeks of work. AI compresses that to a fraction of the time. The freed-up hours flow directly into more accounts, more deals, higher quota.
But don’t freeze comp. Double quotas, yes. Keep base the same, yes. But give reps more room to earn upside. Pay fewer reps more. Let aggressive reps hit 150% or 180% and take home significantly more money. The incentive structure has to match the new reality.
Quotas are going to double. Sales rep yield is so low for so long it’s completely unnecessary. You get all that research time back, you handle more customers, you should earn more.
Kris Lawson, Ruby AI
Here’s the hard truth: hallucinations in LLMs are permanent. They’re not a problem waiting to be solved. They’re an architectural consequence of how LLMs work.
An LLM uses motivated reasoning. It takes your question, guesses what answer you want to hear, and delivers it fast. It makes assumptions along the way because speed and confidence matter more than perfect accuracy. If you forced LLMs to flag every assumption, to be transparent about guesses, they’d burn way more tokens, get much more expensive, and feel less credible. An AI that says “I’m guessing here, I made an assumption there” feels less authoritative than one that speaks with confidence.
So the LLM industry has built a system that trades hallucinations for speed, cost efficiency, and perceived credibility. There’s no financial incentive to fix it. Kris has caught Claude making confidently wrong statements about Gong features. When he asks why it will do it again, Claude says yes, it will, because motivated reasoning is structural to how it works.
This is why purpose-built systems matter. Ruby doesn’t use motivated reasoning. It breaks requirements into multiple specialized agents. One agent researches, another verifies the research. No single agent is trying to think about too many things at once. If Gemini does the research, Claude verifies. Multiple models checking each other’s work. Ruby also retains memory across the whole organization—accounts, deals, stakeholders—so it never starts from scratch. Memory and multi-agent architecture together eliminate confabulation.
Hallucinations are a feature, not a bug. There’s no financial incentive for LLMs to get rid of motivated reasoning. Purpose-built systems reduce hallucinations by design.
Kris Lawson, Ruby AI
Buyers are using their LLMs to evaluate your proposals. If you send vanilla content and your competitor sends deeply specific, verbose, business-aware content, the buyer’s LLM will recommend the specific one. The LLM itself cognitively surrenders to the better-written argument. So the old playbook—net it out, keep it short—is half-right. Keep the first page clean. Add two or three pages of detailed methodology, your understanding of their business, why you’re a good fit. Make it easy for their LLM to recommend you.
The same applies to datasheets, case studies, and proposals. Don’t send generic AI slop. Map it to their industry, size, use case. Show you’ve done the research. It takes minutes if you have context. The bar has risen. Hyperpersonalized content is everything now. It’s the price of entry. Without it, you’re invisible.
If you’re building sales from scratch, the order matters. First hire: SDR. You, the founder, stay close to selling. You stay close to the buying signal. The SDR tees up meetings so you can focus on closing and learning the market. Pay the SDR well. You want them to be your mentee, someone who can grow into a sales role later. Make them high-intelligence, motivated, fearless, and hungry.
Once you have more meetings than you can handle, hire your first account executive. The debate is always: one or two? If your TAM is large, your ICP is clear, and you have product-market fit, hire two. You’ll have comparison data. You’ll see who’s doing well and why. You’ll get clean signals faster. It costs more, but the learning is worth it.
When you need a third AE, hire them as a player-coach—someone who can manage the first two while still selling. After that, think about a full-time manager. The progression avoids overly complex hierarchy early on and keeps one person connected to both strategy and customer reality.
Start with an SDR, add two AEs if you can, then convert one to a player-coach. By the time you need a manager, you’ll have clear data on who should do it.
Kris Lawson, Ruby AI
You said 90% of sales cycles will happen inside LLMs within the next 24 months and that sellers are going to be left out of buying cycles because buyers no longer need them for information. Attention spans are smaller and as well the buyer is honestly they're evaluating you through their LLM and their LLM is their thinking partner and if their thinking partner is giving them a lot of information they may feel like they don't need to spend a lot of time with you. There's this term called cognitive surrender cuz people are inherently lazy and so it's kind of easy if somebody just makes decisions for you. Anything you send to your customer is going to their LLM. If you send a proposal that is very vanilla and your competitor sends a proposal that's highly specific, that deeply outlines verbosely your understanding of the business and why you're a good fit. That claude is going to pick the verbose, highly detailed proposal versus a fairly generic proposal. A hyperpersonalized content is everything. The bar is really high today to be likable and to add value because it is so difficult to get that meeting. What does the sales hire even look like for a company launching now?
Um, this is who we hire, the show about how the best technology companies build their teams. My name is Louis, your host, and today we are with Chris Lawson, CEO of Ruby AI. Chris lost a deal to Salesforce while selling for Oracle. Got recruited by them and then started Salesforce Canada from a fourperson RIA office into an 800 plus person operation. He did it again at Snowflake building Canada from zero to over 35 million in ARR. He has personally closed 1,800 new logos across his career and is a father of triplets. Now he's building an AI agent designed to replace a lot of the work that the sellers he used to hire were doing before. We're stoked to have you on, Chris. Thanks for coming on Who We Hire.
Thanks, Luis. Glad to be here. Excited for the show.
Yeah. So, take me back. You were at Oracle. You lost the deal to Salesforce. And was it a moment where it's like, okay, if we can't beat them, join them. Um, like what what did you see in that loss that made you want to switch sides?
Yeah, it's a great great point. Um, great question. So yeah, back in the day it was uh this was like 2004. Um Salesforce had had really come to market and started to appear in the enterprise around that year. And at Oracle we had a uh CRM system. It wasn't really good. I mean nothing was really good at that time. Poor ease of use. Uh you know the expense of it and setting up uh data centers. It was a painful. And so uh we lost the deal. We lost a deal rightfully so to Salesforce. came out of nowhere and the uh the leader of uh the Oracle sales team um he asked me at the time he's like he heard about this deal and he's like I you know you've done well here in the last two years how did you lose to Salesforce and I said oh these guys are something special something something interesting is happening here and he said okay can you come to our next uh management offsite and present I was like I sales rep go to this this management and uh leadership ship offsite and present how to actually beat um uh Salesforce and like what the learnings were. And so uh I went through the S1 uh because they had recently gone public um and I started to realize that everything that we were struggling with at Oracle, Salesforce was able to do. And it was really really interesting as well. It's like right after that presentation happened, I got a call from a recruiter who uh told me they were opening the first office in Toronto and and were um were were wondering if I was interested in joining. And so it was all very serendipitous and I think that's just in general in your career. You got to look for the signals and and look for uh where the universe is going so to speak, not to be not to be too um mystic, but sometimes these things just happen and you have to read the signals and and and look to be comfortable with that and and uh and see where you go. So I ended up joining and yeah there was four of us in our shared office space that was quite cramped and uh the rest is history.
Yeah. So from four to 8 years later there was 800 people you had personally closed 200 million in in ARR. Uh so you got to experience that growth when when you're building out a sales team that fast. What is the hiring profile that you saw that kept winning? Yeah, I think um it was interesting back then at Salesforce because it was very much a new logo sale. So I think a lot of people look at Salesforce and hiring from Salesforce today um as um as being a little bit different. And where those Salesforce reps are generally really good at cross sales and upsells, back at those days, it was all about new logos. Like for example, like I um covered for one of my sales reps uh because she was pregnant. I had to go and personally negotiate the first deal with Shopify with Toby, the CEO. That's the Wow stuff that we were working on. Yeah, it's pretty pretty wild. So, at the time, we were hiring a ton of people. We had to get them to learn the playbook. We had to um manage them pretty closely and really create this culture of selling. So, we would have to uh we did a ton of Sandler training. So, like I personally co-led uh nine Sandler trainings in the years that I was there. So, I feel pretty good about the Sandler system and upfront contracts and all that fun stuff. And so, what it came down to in terms of who the best people were is a lot of it came down to coachability, right? So, bringing people from all sorts of backgrounds, all sorts of different uh selling environments that could be coming from traditional enterprise sales, they could be coming from startups. they could be coming right out of out of university uh or college right into you know right into like an SDR type role and one of the most important things were that people could just jump in and follow this playbook follow Sandler follow all of the different uh processes that we had we were we were pretty intensive with our processes there were certain things that had to be done certainly certain things that had to be done within uh Salesforce the CRM which at the time like you know sales people weren't coming in with a with a rigor of of really being CRM first. So, I think coachability was was really really important uh then and that's probably one of the things that stands out. Like I think about the people that I had the pleasure of of uh leading over the years and and some of the people that uh stayed at at Salesforce like there's uh some incredible people that um I personally managed that stayed there into VP, SVP and even president uh level roles and those uh people those individuals that really thrived were actually the most coachable people where they were the people where I could give them a guide guidance and they would basically ask a question or two and run with it and implement it and and stay uh rigorous to it. And so coachability is is incredibly important. And then other stuff was just energy, right? Energy, confidence fearlessness um uh like a high acumen uh for for understanding a customer's business. So intelligence and um if you put those things together, then it was a great fit for that that stage that we had of new logo hunting and and closing.
I love it. It's ve very common now when we work with clients, they don't want to work with somebody from a big or because exactly what you said, right? They're used to cross-selling and upselling, but not net new logos. But 20 years ago, that that's all Salesforce was. Uh so it when it came to creating these processes uh did you walk into them already created and you were following a playbook from the US team or did Canada get some more leeway into creating their own uh system and processes?
Yeah. So the the the Canadian office in Toronto was the eastern North American hub. So we actually had uh salespeople that covered all up and down the the east east coast of uh the United States and Canada. So we had Boston, New York all the way down to Florida covering. Yeah. So that's how we
So Canada was leading the the entire US east side.
It it was Yeah. So we had two hubs at the time. So there was San Francisco and then there was Toronto. Um and uh and that was pretty common back back then like Oracle Direct was a teles sales operation of Oracle that was um that did the same thing and that's who I worked for and how I ended up coming over to to Salesforce. So yeah, what we did was um there was a there was I would say a little bit of a playbook that was generated in in San Francisco. It evolved greatly over time, but we certainly did some things that were unique to Canada. Um, one thing uh we started there was uh basically like it was like a rundown. So a rundown was you would go through every single deal and you would say have we been on site? Do we have direct access to the um executive sponsor? Do have we delivered an ROI. So it was like this checklist of things that we knew if we did them the deal was more likely to close. And that was something that we instituted in Canada first that became a a global standard for for the SMB or uh around the world. Same thing with with business case formats that we would do. Um all of these things came out of out of Canada and and spread around the world.
So I I love that you said the first thing there is, you know, have we been on site? Uh that's something that I haven't seen AI be able to replicate yet. uh we're not sending the robots on site to meet to meet with our customers in this world that we're evolving to. Do you see that being even more prevalent now than before?
Yeah, there's lots of little sales addisms um or or uh adjectives that uh that you could use like magic happens face to face, walk the halls, right? All that kind of stuff. And so face to face does matter more than ever. I think it's it matters more than ever. It's starting to come back a little bit with return to office. it's easier to get out and and see people. Uh but the thing is when you get there, you have to deliver, right? You have to have um a degree of charisma, curiosity, likability. You've got to be somebody that that this person wants to to see again as an individual. Um adding a little bit of humor, I think, is is really important. And then after that, it's it's a lot of the information and the information that you can provide about that uh particular customer. So, some sort of adding of value, right? some sort of understanding of their their needs or maybe assessing their needs and and providing some insights about how they're running the business uh based on what you've learned from from other companies. And the bar is really high, right? The bar is really high today to be likable and to add value because it is so difficult to get that meeting and attention spans are are smaller and as well the buyer is honestly they're evaluating you through their LLM and their LLM is their thinking partner and if their thinking partner uh is giving them a lot of information they may feel like they don't need to spend a lot of time with you and so you have these moment of of of at bats where you really have to be compelling whether you're on the phone but certainly if you're you're face to face building those relationship ships so critical.
100%. So after Salesforce, you started Snowflake Canada, zero customers, zero revenue, zero partners. Uh four years later, you had 40 plus people, 300 logos, uh over 300 logos, 35 million uh in ARR. So you started two major organizations, sales organizations. Um like what what did you do differently the second time around now that you had this playbook uh established? Um yeah, so something we did different so what did we do differently? So I think one thing as um starting Salesforce or excuse me Snowflake Canada the one of the things that we did differently aside from the hiring profile which I can talk about is just building a partner army. So we realized that we had to get scale and scale fast and there was basically two of us right. So there was myself and our first sales rep uh building out the country and we were collectively uh up against maybe 400 salespeople. So all Microsoft all the Microsoft sellers, all the AWS sellers, ter uh terod data, cloud era, all those competitors that we had back in the day. And so we were up against hundreds of sales people and there's two of us.
And so the David and Goliath moment.
Yeah. My sales rep's uh name was actually David. Yes. So I wasn't Goliath, he was David. Um, so yeah, so I think that that's one thing and you think about it from a founders that might be listening to this podcast like going it by yourself, going it alone by yourself without partners, without an ecosystem that can echo you, that can can bring you potentially leads, tell people that you're good. Um that's uh that's such a critical thing and we didn't really do that at Salesforce back in the day. We were we were mostly by ourselves although we did have some SI that were they're starting to pop up back then but today I think uh that was one of the biggest things that that we did differently there is just quickly build a partner army that could could go and help us drive leads. We had we had um 40 names in our database when we started Snowflake Canada. 40 names, not leads, not opportunities, we named customers. We had literally names. Uh, and that's probably how many people had filled out forms on on the Snowflake website, you know, up until 2018. So, it was very daunting. The the people that we hired um my profile, I think, at the time was was was definitely a little bit different than it was at uh Salesforce. Like at Salesforce, we had a very rigorous onboarding program, right? We had Sandler sales training, we had discovery training, we had all this training to to kind of build people up. Snowflake did not have like a sales academy of of of any type, right? We we at the time um didn't have extensive uh like we had basically boot camp or 3 days and then you're off and running in the field. So there wasn't a ton of of uh skills training that needed that happened and we also had the major uh new logo push that we had to get to. We had no brand in Canada. In fact, around the world like snowflake at the time in 20 2018, you you would think now, oh, everybody knew snowflake. Very few people knew snowflake. Snow the only the only snowflake that was being talked about was the the term snowflake that was being thrown around like politically at the time. So people would generally laugh at us.
Yeah, that's right. So people would generally laugh at us first because our company's name was Snowflake and they would make fun of us a little bit and then they would be like, "So what are you guys all about?" And so we didn't really have a brand uh to speak of then. So we we really had to find two people. So the combination of these things, right? Didn't have a ton of ton of sales training. Um we didn't have a a brand and people just didn't didn't uh didn't know about us and and uh weren't running into us. So we had to do two things. One is that we had to hire people that already had come out of salesmies and these they had come out of salesmies. They had spent some time at Oracle, Xerox, um could be IBM, could be Salesforce. Uh, and then we would want them to go out and prospect like crazy. And we call it like being a PG savage, like P pipeline generation savagery. Go out, you you would talk to five new people every week. Try to talk try to basically do uh new meet new business meetings. That was a critical metric for us that also I could share with founders is is a is a really important metric um to track new business meetings which is the the the leading indicator to the leading indicator of pipeline generation. And so if you thought about this for a second though, if you think about we wanted to hire people from a sales academy, right? Really important, but we wanted them to be uh like a pipeline savage going after companies where you didn't have a brand yet. You can see that there's there that's a big jump from I worked at a Salesforce and now I'm going to be a pipeline savage and go and go and push my brand on other people. There's a gap there. And so the gap that we needed to to to look for was was somebody at a sales academy went to a startup where they had to do that action where they had you know limited brand value or uh behind them um and they had to go after new logos. If they've done that then we then we take them. But if we jump straight from the sales academy where someone might probably didn't do a lot of new logo hunting that probably was cross-selling and upselling and had a brand. If we go from the si the that academy directly to the needs that we had at the time at at Snowflake, it was it was nine times out of 10 a complete fail.
Wow.
So the
I was gonna guess 50/50, but 90% of the time they were not going to work out unless they had that experience.
It was it was a huge problem. It was a huge problem. So we had to try to find people that like later right some definitely post IPO when we had a lot more brand and and we were moving a lot more up market and talking a lot more sea levels like around that time we could actually end up um we could start to hire some people like that although it still wasn't advisable but back in in 2018 no we needed we needed people that were trained but had also already cut their teeth at some startup um and then were prepared to to join us which you know was a good choice for many people.
Yeah, it makes sense if you want a a pipeline savage.
It's so important. Yeah, so important.
So, you then advised Gong uh 2017 through enterprise adoption, helped them land logos like HubSpot, Snowflake, um now you're advising Maxa, Axia, if if I'm pronouncing those correctly. Uh so, so when a founder comes to you and says, "Chris, I need to build my sales team now into AJI." what's the first question that you're asking them and like why why do most of them think about hiring wrong?
Yeah. Um well I think the journey of of founder sales and so and so yeah so the advisory that that I've done in the past like uh I started to work with gong when they were about 15 15 people and so they had uh I think two or three sales reps um when I when I was starting to work with them a little bit and I was working with them on like product market fit messaging um but then also just really opening doors and a lot of key accounts that that ended up being some of their cr their first uh big reference customers. And so the and then same thing with Maxa and Axia was moving from founderled sales to to sales sales and and lining up that those businesses to both race series A's which which they uh they've done successfully. And so the the the big mistake that a lot of founders make when they start to hire their first salespeople is that they one they they hire probably from a sales academy like I just described and they they they make that judgment error. Um and uh which is easy to to to to fall for. And then the other thing I think a lot of a lot of founders do is is they they step back from sales and they're like, "Oh man, so finally somebody's can can take this off my plate. It's been really taxing for the last two or three years. me selling. I've hired this professional from X company and now they're just going to run with it. And the um that becomes obviously a big big problem, right? It becomes a big problem because the founder gets stays away from sales, doesn't get enough signal from from what's happening with the customers and that person who came from big company probably doesn't know exactly how to tell the story the best. The third thing as um as we're talking about this that I think a lot of founders um are challenged with is actually just getting their story to that to the salespeople and and helping the sales people understand how to tell the story of that particular product like um or solution the the founder we talk about like founder magic. So founders come in they can they can wow people um often a lot of people and I know this now as a founder a lot of people you sell to kind of like semi-admire founders as like a like they're they're kind of like like they they want to be close to innovation. They like the idea of founders. Um, and they they tend to listen to founders a little bit more than they would with with sales people because they're just fascinated with that that uh journey and often sometimes want to sort of vicariously participate in that journey of that particular founder. And so there's a there's a degree of like um u kind of um a wide birth that founders get because they they're founders. They have a they have a kind of a more interesting story that people would like to to pay attention to. But then they're also just deep experts and they they are very compelling and and being able to tell their story. But that story that founder magic doesn't typically easily translate into like a first call deck when you finally meet with a customer. It doesn't easily translate into some of the website messaging. It doesn't doesn't necessarily always translate into the the outbound messaging that you want to use to try to get a meeting. Right? So, I think one of the first things that that I've done with all of these companies that I've advised uh with is is really first start with the story. So, okay, great. So, we've got the salespeople, we've got them in place. Um, we need to break the story down of your company to base principles of like, okay, who's your ICP? What are they trying to do that they can't do today? What are the things that that they've tried that that have failed? What are you going to do for them? Why is that going to be different? Why should they do it now? Right? like telling that whole story and getting that down and then understanding that story, agreeing on that story with all of the internal stakeholders. You got product in the room, you got marketing if it exists, you've got any other co-founders, um you've got sales people if it exists. All of us agreeing like that's the story that that that we want to tell. And then from there, you're starting to build out assets, right? You're saying, "Okay, well, now that we have the story down, let's what are the assets we need on the website? What are the first calls that that we knew we need?" Right? that first call deck is like one of the most critical assets that you can you can possibly have today is like what keeps the sales rep on the story uh during the meeting and uh it ensures that that they they follow that and and and hit the right right moments. But I think that's that's one of the biggest things also that founders struggle with is that they move they run past the story and they assume that the sales reps can can do it and often it's just not ready for you know wider distribution at that point. So, it's not a script, it's a story, and then you're enabling them with stories, going back and checking to make sure they're adhering to the stories. And that's really where where you started. Um, and I mean, when we look at Gong, it their story is very clear, right? They have another 100 competitors now in in in the market, but the way they've evolved and their story has been consistent. I and you're able to go and see their their reps everywhere telling the same, you know, story that Yudi is. uh is telling. So I think there there really is some magic to that and it is one of the hardest things. I see one of our reps that does a really good job in telling the stories of like the pains that I struggled through when I was using other agencies and why I started Latam scent. Do you see that story be more compelling from tell like a rep telling the the founder story or really living it as if it were their own story? And what what are some of the best ways to continue to relay that information without playing a game of telephone and losing it along the way? Is it enforcement? Is it watching those calls and retelling the story? Like what have you seen that works really well?
Yeah. So the um the one thing that you mentioned there was the the how does the sales reps tell the founder story and I think it's always great to actually tell the history of the company right and and use the founder as a as a story of of um of that they probably saw the problem and um you know this is what they did about it and and aren't the the investors believed in them and we have we we were given money and all this kind of stuff. So you're kind of like co-opting that founder journey and the origin story of the company and you're you're laying that out so that there's credibility of of founder and credibility of investors and and so on and then getting into the actual uh profit itself. So so telling some of that history as to why you should listen to us um even though we may be small but why you should listen to us because we we've got domain expertise etc etc. That's a really really good thing uh to kind of roll with in terms of like is the how to actually follow along and make sure that that the story is happening. Certainly there's um like definitely gong, right? There's Fathom. There's all sorts of different call recording tools you could listen to and you could monitor and and see um see what's happening. I think one of the easiest ways to understand if the story is actually happening the way that you want is that you're actually getting the information back from the sales rep as to how the customer has responded to the story. And so what you're looking for is you're actually looking for the sales rep to do really good discovery because if the story for example if the story is um you know a particular pain like they you know if you talk about Snowflake for example their BI we at Snowflake we solve uh BI latency issues where where your uh reporting is just and dashboards are are not refreshing frequently they're taking too long and so if we say that's one of our main pains that's we want to want to um sell to and then after any sales call hopefully in Salesforce that particular sales rep is saying okay they use Tableau and they do have this pain and it's related to this dashboard and so you're kind of like you can see how you're kind of chasing to make sure that they've uncovered that value you can't the fact that they've said it is you know interesting hoay you've said it but what has happened when you've said it right what has the customer said in response to this this pain hypothes thesis that you've brought to them. I think that's one of the easiest ways to to check and that's like a that's like a first-line leadership thing where you're going through you can go through the calls and and and assess that. But generally when you're doing uh deal reviews, I think that that's important that you're you're assessing that they those questions are are happening on a regular basis. They're asking the right questions. Um, it's pretty easy to find the c the the sales reps that have a have a um instinct for curiosity and and the ones that don't because they just they generally are the are within those two buckets. But yeah, that's probably the easiest way to to validate it.
Hopefully question. So you you know you said 90% of sales cycles will happen inside LLMs within the next 24 months and that sellers are going to be left out of buying cycles because buyers no longer need them for information. You've closed know over 1,800 deals the old way. So if your thesis is right, what does a sales hire even look like for a company launching now?
Yeah, the sales hire is definitely different. There's there's a whole bunch of reasons uh how it's different. Um, so I think one thing that's that's happening from a sales hire perspective is there there is a like a compression of role. So there's this idea of like full stack SDR or excuse me full stack AE. So bringing in account executives that have uh the ability to be the SDR to uh be the sales rep and even potentially to be the customer success uh manager, right? And so I think the the sales hires that we hire today need to show that they can basically run the full cycle of the the customer experience because now we have this AI um along that that full customer experience that does all the work that took three people, right? So it it basically does a lot of the the the sequencing, the meeting generation, right, that an STR would typically do, the research of course that an STR would need to do. Um then it would also uh do a lot of the the the research and like CRM updates and all the stuff that a that an account executive would have to do. And then from a from a CSM perspective, uh AI would help you build your your QPRs. It would flag uh at risk accounts. It would create a a plan to go after those accounts. Again, like you think about it, it's probably like 30 to 40% of all of those roles over the years was about back office research, uh, reporting, uh, coming up with plans. So, I'll give you an example. I was speaking to a customer this morning that's that's using, uh, Ruby. So, basically, um, what you can do with Ruby is is that you can basically say, "Hey, I I work for this company. I want to sell to these five companies." That's it, right? So you can do that and automatically Ruby loads all of the different uh accounts, finds all the right stakeholders you need to chase, finds the use cases. It just does it all automatically, literally with like two prompts. And then when you have those five accounts loaded, you could literally say, okay, draft me five unique sequences for five people at each of these accounts. And you can just push it into like outreach. And so you can use the intelligence of AI building highly bespoke sequences that go directly into an an engine that then execute those sequences. And so all of that was like 40% of a of an STR's job, right? Was like do find find, you know, the accounts were easy to find. The people were hard to find. um if you wanted to do really intelligent uh sequencing, you'd have to handcraft it uh by on your own and then you'd have to basically put it into into outreach and all that took a ton of time. And so if you think of of that being then you could use other systems for this, but I mean that's something we like to do with Ruby, but if you think about um how much time that saves, a sales rep in theory can actually can actually do that themselves. And so the other piece of this we talk about like a sales rep being able to kind of handle full full spectrum of the of the sales cycle because they now have the tools to enable them. So that's kind of like one thing that's that's changed is like can we test for instincts around pipeline generation passion for that how they've done in the past because it's not just a a check the box oh yeah they can do some pipeline. No it's like they're going to have to really prove that they've been able to do this. Um and then similarly like tell me about a time a customer was failing in the application they were using and what you did about it right like how did how did you save save a poorly adopting customer how do you ensure that whatever you're selling is going to be adopted well like you need to basically test for for that whole whole spectrum and then the other thing is like AI pill right like this is like the term now is like hey have you been AID and uh that means that we just look for examples of of sales people and how they're using AI today. Um, you know, are they using cloud code? Great. That's nice if they are. Um, that isn't necessarily the the must-have thing today. It's just what are the things that you're you're typically using uh AI for. Are you getting into projects? Are you getting into like saved prompts? Are you doing all that stuff? That's the other thing you have to test for. Absolutely. In today's world. Um and then uh and then you know at the end of the day you also have to just make sure that they have curiosity and that they they are going to question the results that they're getting from from uh yes their their AI like it's really scary. Um there's this term called cognitive cognitive surrender right cognitive surrender. And like this is just a a thing that is is going to be um a really bane on society is that we are just all deferring to our AI to be not
what should I do here
right and not just our thinking partner our our thinker it's like I will just listen to this because people are inherently lazy and so it's kind of easy as somebody who just makes decisions for you right um so that is going to be the hardest part to test is going to be curiosity and I think that this comes back to leaders where you just meet with with uh a sales rep uh beforehand and hopefully you guys have you know you and the sales rep have both received a meeting briefing uh from AI but then you just have to ask some questions about like well what do you think is the biggest objection that they're going to raise how are you going to solve it right like what's the what's the top question you're going to ask um how would you say that another way right like trying to find just different things uh to to to query them on so that they are still you're forcing some degree of of curiosity through this, right?
Love it. So, we we both met at Saster. There's ads everywhere, people flying planes saying, "Fire your sales team, hire our our AI robots." Do you see the SDR role going away completely and being replaced by AI? And what about the AE role?
Um, I don't see it going away entirely. I see the the SDR role kind of weak kind of like blending into a little bit of like go go to market engineering. So that that SDR is likely building out a lot of the the the outbound systems that everyone would end up using too, right? So I think that if we think that um a an AE in in organizations is going to exclusively own pipeline generation, I don't think that it's going to swing that much. Um but I definitely think that that the SDR role is going to be a little bit more reduced. So I think just in general span increases, right? So the number of the number of uh account executives and SDR supports goes up, right? Ratio increases. The number of sales reps that a manager manages that increases. Like this idea that it's always insane. This idea that a first-line manager manages six salespeople is like insane. And it happens a in many companies. It's still happening in many companies. The idea that an AVP manages like three or four direct managers, that's insane. like the the the cost um of of operating that those types of management layers is is is really ridiculous. And it also
So that's going away. You see those those getting compressed?
Yeah, I think we're going to get to a point where um like every sales leader would have a co-pilot that helps them manage their sales reps that proactively raises uh deal risk that kind of ass is monitoring everything that that's going on in the in the organization. Um, and I think they I think a sales leader does get to like 12 to 14 direct reports even in in like an enterprise sales motion. Um, but going back to the SDR comment, I think it's just the same thing of like higher ratio. So in a world where an SDR used to support like three reps, they probably support six or seven now. And as part of their job, they are helping tea up a lot of the the outbound use of of AI applications, right? that and they're basically becoming like a a bit of a go to market engineering coach for those salespeople and and showing them exactly um how best to to use it. I think that's going to be a major part of it. But I think if if for SDRs today who are going into and I was an SDR for the first year of my career at Oracle. Uh I think for SDRs today to be really really successful, they they do have to be elite users of of AI and um have to just sort of live live levita AI and uh and be able to show people that they are doing that like give them evidence that they're doing it and show the results that they're getting from it. So you think the playbook's going to change where before it was hey hire you know 50 SDRs uh 25 AES so with Ruby and AI now in organizations 100 people let's say they have 50 sales people like what are they going to need in three years is it are they going to be able to do the same with five sales people
yeah I don't know if it'll it'll reduce that much I think probably you I mean either way like you either can have ideally your sales sales people are probably um maintaining relatively the same maybe down like 10% 20%. But quotas are doubling right like I think that that
Do you think quotas are going to double?
I think quotas are going to double. I think they have to double. I think that they I think that um if you're getting all this the sales rep yield has to increase. It is the sales the the yield per sales rep has been so low for so long it's completely unnecessary. And if you think about applications out there today, like you to do a thorough account review, like handcrafted account review to find the right uh stakeholders, uh find out what they care about, find out the use cases that they that that that company cares about. Uh triangulate it with news, all this kind of stuff to really be an intelligent salesperson, not just throw crap at a customer. Like you're generally like four to six hours per account to do that. Um, and then you're like, "Okay, well, well, I need to do this for 50 accounts a year minimum to to try to figure out how to break into those 50 accounts." And so, what's what's the math on that? You're like at 240 uh hours and that's six weeks, six man weeks of just research, like if you think about it. So, that's pretty insane. Imagine a world where you get all that time back. You certainly can handle more customers. You certainly should be able to handle more deals. um your time should be you know obviously very freed up to be able to go after these accounts and thus your number should increase. I think that we can't we shouldn't necessarily just double quotas but we should uh and and maintain comp the same. I think you should double quotas but then still give the people an opportunity to continue to earn more um and uh and not just make it like a one for one, right? We want to pay. We want less reps, higher quotas, paid more. That's it.
Love it. We had uh Mark Shottmire on last episode and or two episodes ago and he talked about how he wants everybody hitting 100% of of quota and you know he comes from a he came from a finance background. So he was discussing, you know, he he said he was able to go and uh argue with his finance counterparts now that he's he's the founder like no, we want everybody hitting 100% and we want those aggressive sales reps. We want to overachieve and hit 150 and 180. Um so it I I think it's a really exciting time for those that want to level up in sales and now are equipped with this AI. I mean I I see it in our team. the ones that are most successful are the ones that are living in Claude and using it every day and not sending the first draft they get from Claude, but really thinking with it and using it as a thinking partner as you mentioned.
Yeah. Yeah. It's it's it's really important to um to question it, right? So I I like I do I do want to talk about that for a moment. So this might sound a little conspiracy theory, okay? So let me lean in and tell you
love a good conspiracy theory. Let's go.
That I think exists. I think that that um hallucinations are a feature, not a bug, of LLMs. And and I'll tell you why. And so um an LLM takes a lot of assumptions, right? It takes a lot of assumptions and it it wants to give you answers that that you want to hear essentially. So it's called motivated reasoning. Uh motivated reasoning means that it it basically listens to your question and it it guesses like what is going to be the most likely answer and it also wants to deliver that really really fast and so it's going to give you these pretty fast answers even if it has to make assumptions along the way to deliver this this this reasoning. Let's assume that we wanted to get rid of motivated reasoning entirely and we want to say, "Hey, AI, don't do assumptions anymore. Just be transparent." Like, if you've if you've guessed on something, tell me. Okay, let's let's assume we want to do that. Well, if we made that decision to have AIS clarify where they've made assumptions, the LLMs and everyone else would burn a ton of tokens. it would get a lot more expensive and you would have an AI that is kind of second-guessing itself and not giving you a highly credible feeling, right? Like if I I'm talking to this AI that's always telling me, well, I guessed on this, I make an assumption here. I make an assumption here, then I'm probably not going to have a high confidence in the veracity of that AI and and how well it's going to be able to support me. And so to clear hallucinations, it's going to cost either a lot of money or degrade brand of AI. And thus, I think that hallucinations are just a permanent thing. Like it's they're not going to go away. Certainly not anytime soon because there's no incentive, no financial incentive for these large LLMs to actually do anything about it. And the only reason why I know this is because I've caught Claude um making mistakes and and uh where I know that's the wrong information. So I'll ask a question about Gong and it'll be it'll tell me this answer. I'm like no I Gong has that feature. I I don't know what you're talking about. Like I'm sorry. Motivated reasoning. That's literally what it is. So if you um catch Claude doing this or Chad Petite, just ask it. Hey, why'd you do this? Will you ever do it again? And so you know what it says when you'll ever do it again? be like, "Yeah, I'll probably do this again because I'll have the memory that you've ever me to do this."
So, it's it's really challenging, but I would encourage anybody just just for fun, go ask Claude today, do you ever use motivated reasoning when you respond to me? I'm going to do that right after this podcast. And and and this is why Ruby exists, right? So, because it doesn't have that motivated reasoning. It has actual data. It's not going and making assumptions.
Yes, that's right. Well, it's we we break all of the the requirements down into multiple agents that have to work together. There's no context drift where an agent is trying to think about far too many things. So, we kind of break break our our um like uh sales uh processes down to a variety of agents. Uh and we also have other agents that are double-checking the work of those previous agents. And we've all always said like do not do motivative reasoning, don't do confabulation, all of these other things that that AI um agents do. And then we have a different model check it. So if we were running Gemini and swapping models in all the time, Gemini is doing the research um and the reasoning, then we would have Claude actually verify are these results sound? Um have we delivered what we we wanted to deliver? So I think purpose-built uh systems like like uh Ruby dramatically reduce hallucinations and the other piece that that purpose-built solutions do is they retain memory right so they retain we retain memory uh and context on account level on a deal level on a stakeholder level so that whenever you go back and you ask a question about that account or that you know deal
we are immediately retrieving all of the memory and context text and not kind of thinking from scratch. And so I think when you're doing this with a with a regular LLM, you generally have to find the same context window, talk talk in that same window potentially in the same uh project, but it's very very difficult to always go back and make sure that that the memory is actually there. You generally have to refeed the context. And so that's the other piece that we do really really well is we basically make sure that the the results are valid because we retain the memory across your whole organization of all your deals and all your opportunities.
Sounds like a go to market team's dream. So Chris, you've been on every wave cloud, CRM, data, call intelligence, ABM, now AI. um how are you thinking about your your first sales hire in in in the future in addition to you know the getting the founder story uh right? How how are you thinking about building your go to market team when you scale more?
Yeah. So I think the the the first thing that we need as as founders is really like a an an SDR, right? So we actually need that SDR first to really tee up meetings for us because the more meetings you can do as a founder and staying close to that that sales process um the better and it's also a lot less expensive if the founder can continue to do you know a lot of selling. So I think that first role is certainly an SDR, ideally an SDR that that is high high intelligence uh uh kind of motivated, fearless, like maybe lean in a little bit more, pay that SDR a little bit more than than you would you would regularly pay an STR because you want them to essentially um be your mentee and grow into hopefully some sales role.
The founding founding SDR, but how quickly do you want them to start taking over your meeting there? It sounds like not not quickly at all. you want your your playbook is to stay as involved in sales as long as possible.
Yeah, I would say that you you try to stay along it's try reasonably to stay in sales as long as possible. Um especially if you have like a strong CTO that can hold a lot of the the the like product reins and then yeah you get that SDR you get them uh getting you meetings and once you have too many meetings then you're like hey you know what I think I need my uh my first sales route. there is a lot of discussion of like should you just hire two right so that's always the debate is like should I hire one sales rep or should I just go for it hire two if you feel like you've got enough uh like a large enough TAM um ICP and customer success that you could you could reasonably if all the the stars align handle two sales reps then you should hire two um it's a little bit more cost obviously to to do that it's a little bit more cost but then you just have a comparison you have multi multiple uh sets of signal coming in of who's doing well who's not doing well, what's actually happening in the market. I think that's a great initial pod to have that SDR and then two sales reps. By the time you need to add a third sales rep, you're probably hiring that third sales rep as um a player coach or converting one of the first two as a player coach to manage that pod and then probably after that you're thinking about um hiring a manager.
Love it. Love it. The um and even so really hasn't changed even with AI. It's just more AI fluent folks that are able to do more. Um, and they have a higher quota. They have more more work to do, but they're able to achieve more because they have tools like Ruby and, you know, other LLMs to help them.
That's right. And and they have to, right? They they they you have to just step up your game so much in terms of the the value that you're providing to your customer. Like take content for example. So content. So, if we're if you're if your buyer is doing cognitive surrender, right? Um, and they all are.
I love that term.
They're they're a good one. They're sending anything you send to your customer is going to their LLM, right? So, latam send you you've got a proposal for me. My LLM, you know, understands my business. Probably doesn't doesn't have a complete memory and context of my business, but I'm going to it's talking. It's confidently talking like it understands my business. Um, hey Claude, what do you think of this proposal? Right? If you send a proposal that is very vanilla and your competitor sends a proposal that's highly specific that deeply, you know, outlines verbosely your understanding of the business and why uh you're a good fit. That claude is going to pick the verbose, highly detailed uh proposal versus a fairly generic proposal. And uh and in fact, you go back to um uh the old world where it was like, hey, just net it out, couple bullets, right? Like make it easy easy to read. You should still do that as the first page, but then you add another two or three pages of notes, understanding our methodology, all that kind of stuff. So that when that that sale or that when that buyer gives it to their LLM, their LLM is going to recommend you because because the LLM also cognitively surrenders, right? So member motivated reasoning, it gives you the answer it thinks you want. So if you give it the answer that is clearly written out for it, that's what it's going to do. It's not going to actually think about it um unless you're like doing an incredible job prompting it and giving it a ton of additional context. Right. So
you gave me an idea there. Is it a simple Sorry.
No, please go ahead.
No, just a hyperpersonalized content is is is everything like like even from a from a data sheet perspective. Like you should not send uh like vanilla data sheets. That data sheet outlining your product and features and benefits should be aligned to what you've learned about the customer or what you think is the hypothesis of what that customer is doing based on the industry or size that that that they're in. Right. Like that's should just be happening all the time.
Yeah. The bar the bar has been raised. uh that there is a new minimum now for sure. You you can't just send uh some generic AI slop when it's so simple to map out their needs, send them a customized uh proposal in, you know, it doesn't take 4 days anymore. Uh you're able to do it in a few minutes if you if you gave it the right context. Uh Chris, we're you know, as we wrap up here, this has been super valuable. Thanks for so much for coming on uh who we hire. It's been a pleasure hosting you. uh for those that want to connect, where where's the best place for people to connect with you uh and find out more about you and and Ruby AI?
Yeah, LinkedIn is a great place. So, uh Chris Lawson, Chris with a K, you can uh send me a connect there, reference this session. I I think it's great. Be happy to connect. You can send me a message there. And yeah, Ruby is PLG. It is it is designed that any company can use it. There's no forward deployed engineer. Um there's no setup uh required other than just connecting into a couple of your systems. Our agents go and learn your business automatically so that you don't have to necessarily train it. Uh we do an incredible job of that and you can just get up and go. And so that basically means that you can try it today. So you can go to the website uh hey ruby.ai uh sign up for it, run an account, run an opportunity uh check it out, see if uh see if it's going to help you. I think it will. Um, and uh, and call me if you uh, or send me a note if you want to talk about it in in more detail. Happy to.
Love it. Let's make it happen, y'all. Chris Lawson here on Who We Hire.
Thank you, sir. Nice to see you, Louis. You
