Latest / State of the AI Union / Warren Kucker: The Conversation to Action Gap
Transcript
- 0:08Hello, everybody, and welcome to this week's episode of the State
- 0:12of the AI Union. I'm your host, Laura, and joined
- 0:15with me is Warren Cooker CRO at Basic and which is a company
- 0:22built around conversations. I am going to actually let
- 0:26Warren introduce himself and also Basic.
- 0:29Yeah, absolutely. So really briefly, three time
- 0:33founder, 18 years in like shipping and logistics, My first
- 0:37startup was based on an idea that we had at Apple in two
- 0:41years and the distance I had there made sense to go do it for
- 0:43some other companies. 15 pivots later, we had a couple of like
- 0:47decent F sets, a couple of pit stops along the way.
- 0:50I basically converted myself from a founder into a sales
- 0:52leader. I'd say an average sales leader
- 0:56from a founder LED sales motion and then back into the startup
- 0:59game. Basic is yes around
- 1:02conversational intelligence that I'm excited to talk to talk
- 1:04about today. Yeah, fantastic.
- 1:07Warren First, actually tell us about that move to, you know,
- 1:10leading a sales team from being a founder.
- 1:13Yeah, I think it's it's really fascinating when you make the
- 1:17shift from like a founder LED sales model to running like a
- 1:20growth Series A through Series C sales motion, right?
- 1:24You really get to starkly see when you're on the other side of
- 1:27the fence, the differences and the pitfalls when you're making
- 1:31that transition, right? Founder LED sales is all about
- 1:34story, industry expertise, like charisma and very little about
- 1:38sales process. When it's time to hire those
- 1:42like sales leaders who then hire your first like group of reps,
- 1:46translating that to that team could be like really painful and
- 1:49also is a huge failure point for folks, right?
- 1:52Like what's the right amount? You want to keep the founder
- 1:54involved, but you don't want them to be overly involved.
- 1:57So like you still want to keep that industry knowledge, that
- 1:59expertise, bringing that in for that like executive to executive
- 2:02connection. But like the deals need to be
- 2:04run at the team level on top of it, the reps need to be really
- 2:08well versed in the product in the industry as well.
- 2:11That transition is really painful, honestly.
- 2:13So like I've seen both sides of the fence and I think that's
- 2:17generally over the past three years become kind of like my
- 2:18niche and expertise and go to market.
- 2:21Yeah. And and you've talked a little
- 2:23bit about like how sales and conversations come together and
- 2:26where that's sort of like a lot of selling things in the
- 2:31conversation. Tell us a little bit more about
- 2:33that. Yeah.
- 2:34So like when you think about like conversations as a whole,
- 2:38right. So like in general, Forbes wrote
- 2:40that 19, we're taking about 2.2 hours worth a day of meetings
- 2:44just across the enterprise and now we're at about 5.1.
- 2:48So like mini like COVID effect on how often we're meeting all
- 2:51day, right? What's happening is like work
- 2:54has become the meeting, right? So like work product is kind of
- 2:58incidental to those, right? So like leveraging those
- 3:00meetings appropriately is like foundationally important for
- 3:03organizations, right? If you look at this sales in
- 3:06particular is the obvious like culprit here that all sales
- 3:10people do is take meetings with prospective customers and then
- 3:13send emails to those prospective customers in between those
- 3:16meetings. That's how salespeople closed
- 3:18deals, right? This was a bit of a black box
- 3:21prior to LLMS, right? Like it's just you didn't have
- 3:24the ability to have the contacts from that information really
- 3:27easily at your fingertips. Like compare it to like the
- 3:30index of a book. You can like look up a word in
- 3:32your sales meetings like you would in an index in a book and
- 3:35find the page that word was on, but you don't know what was
- 3:38discussed around that word. Now with LLMS in sales ending
- 3:42all things enterprise meetings, you could really like connect
- 3:46your conversations to your workflows much more robustly.
- 3:50Yeah, and I really resonate with that because today one of my
- 3:53meetings wasn't recorded and I was like, Oh my God, I forgot to
- 3:56take notes. You know, I said to go off all
- 3:58of my memory. But anyway, you know, talking
- 4:01about, you know, speaking about like meetings and like having
- 4:04all of the conversations come together.
- 4:07You know, we used to solve this problem with sort of work
- 4:11management with some of these tools like Asana, Notion and all
- 4:15of that. Tell me what you do, You feel
- 4:16like the gaps between what those tools were like pre GBT and now
- 4:22post GBT? Yeah, absolutely.
- 4:24If you think about what's happening in the 30 minute
- 4:27meeting, generally a 30 minute meeting, about 15 pages of a
- 4:30standard sized book of information is being shared
- 4:33regardless of how many people are in that meeting, right?
- 4:35Just speed of chatter, lack of Dead Space, 15 pages of a book
- 4:39for every 30 minute meeting that's had, right?
- 4:42What generally in the past has happened to even we have call
- 4:45recording software is we translate that 30 minute meeting
- 4:50into a couple of tasks and Asana or an e-mail to someone else
- 4:55asking them for advice or telling them about work they
- 4:57have to do. Like it's a supreme condensing,
- 5:02right? It's like the Bluetoothification
- 5:03of meeting information, right? You lose so much context
- 5:07involved with that meeting. On top of it, let's say that
- 5:10meeting had a lot of value, whether it's external meeting
- 5:12with customers, vendors, prospects, or an internal
- 5:15meeting with two teams. You have to decide like what the
- 5:18three things are the most important thing from that
- 5:21meeting and put them into Asana, right?
- 5:22Like generally one person would do that, and Asana is a great
- 5:24tool, but you're losing so much context there.
- 5:27A year later, there could have been something else in that
- 5:29meeting that is now valuable. Think about the transition from
- 5:32traditional computing to LLMS. We don't know what would have
- 5:36been important to know from a workforce standpoint when
- 5:38talking to our customers and vendors from conversations three
- 5:40years ago that it's gone, it's dead, right.
- 5:42So like that, like the it's just an overly condensing the
- 5:46information and there's a selection bias as to like who
- 5:49chose what was most important from that meeting and that we
- 5:51don't have to suffer through anymore.
- 5:53And so that's that gap, that's what you're calling the
- 5:55conversation to action. Gap, yes, I think the best way
- 5:59to close it is to connect those meetings to the workflows that
- 6:03are related to those meetings as closely as possible.
- 6:05Skipping the Asana step and actually getting that context
- 6:09close to the work that needs to be done and closing that gap
- 6:12will be one of the biggest advantages of the next couple of
- 6:14years with the the new new LLMS. Yeah, tell me, tell me your
- 6:18thoughts about this. So in one of my prior episodes,
- 6:21we talked about the AI grind mode versus vibe mode, where
- 6:25grind mode was where AI was creating like more work, right?
- 6:30Like it was creating the, the, the need to read the transcript
- 6:35and read the notes and read more things around what AI could do.
- 6:40And vibe mode is, you know, letting it sort of like do the
- 6:44work for you in a sense. Give me your perspective.
- 6:47Yeah, grind mode verse 5 mode is a really good contextual way to
- 6:50put it. The the problem I think that
- 6:52we're going to run into and grind mode is that great.
- 6:56We have a lot more information. We have a 30 minute meeting.
- 6:59We're not boiling it down to three bullet points anymore.
- 7:01Well, I do I have to process that information now Every
- 7:05meeting is just like a robust like what happened in the
- 7:08meeting. Here's all the work that needs
- 7:09to be done. Here's the all the context you
- 7:11could ever have that's going to create problems along the actual
- 7:16like mental capacity to process that right?
- 7:18No one is leaving at 5:00 and be like, you know what?
- 7:21My brain is so fresh. Like I executed tons of work
- 7:24right now prior to LLMS and I'm just like, I could do so much
- 7:27more work, but I don't need to because work is done right?
- 7:29They were already overloaded with information.
- 7:31We're going to be cramming more information into that thought
- 7:34space. So I think there's a combination
- 7:36of the two, right? When we have to trust that, like
- 7:38on the vibe mode side, that certain types of work can be
- 7:43done by the LLM as well. I think the easiest way to put
- 7:46it is writing. LMS do two things really well,
- 7:49structure unstructured data and write.
- 7:51I think we're going to have to get past the fact that like when
- 7:53you're writing an e-mail or a one page document or creating
- 7:57like APDF guide, that we need to read everything that was done
- 8:01and we can't just like send that on its way.
- 8:03At some point the writing piece will help reduce that contextual
- 8:05space. But I think the second piece,
- 8:07it's going to be up to the the software providers of LLM
- 8:12technology. Let's say that what's called
- 8:13them the LLM wrappers, which is a disparaging term right now.
- 8:16But the software to figure out how to surface the right
- 8:19information at the right time in a digestible way to help the
- 8:24user to reduce the complexity rather than increase it.
- 8:28That's going to be the key. Simple, simple software
- 8:31leveraging back end technology. So like grind mode with a lot of
- 8:35vibe underneath the like kind of water level there.
- 8:38Yeah. And you know, so what's
- 8:40interesting is I've been having a lot of this conversation with
- 8:43my SDRS recently. I actually also have the SDRS
- 8:46under me, and a lot of it is actually teaching them how to
- 8:49have good conversations instead of having the LLM have the
- 8:53conversation like for them. What's your perspective on that?
- 8:57Yeah, give me what's an example of something like that, like you
- 9:01teach them about like orchestrating the conversation?
- 9:05Like, for example, things like, oh, pausing, like learning to
- 9:09listen, you know, not going 55 lines, five line pitch straight
- 9:16away, right? Those are things that the LLM
- 9:19can generate for us really well. It can generate a really good
- 9:22like 5 line pitch, but it's the way in which we deliver the
- 9:25conversation that I, I that, you know, I feel only humans can do
- 9:30this, but help me, help me, help me with your.
- 9:32That's a good way. Let me let me take two points
- 9:34there. Let me let's come to that core
- 9:35point. I think this is a great example
- 9:37of like in the past, what you had to do was spend a lot of
- 9:40time developing that five line pitch, right?
- 9:44So like I have to figure out based on my memory and all the
- 9:47conversations I have and product marketing and my sales
- 9:50leadership, what's the five line pitch?
- 9:53That was an intensive exercise that generally what would happen
- 9:56is you'd spend couple weeks working through it and then you
- 9:59wouldn't touch it for a year, right?
- 10:01So that would be it. And this like gathers dust and
- 10:02then maybe it's like, you know what, this pitch might not be
- 10:04working so well. Let's try something else.
- 10:06I think LLMS can do a lot of the like, say vibe work underneath,
- 10:11like the water level again and then enable it so that like
- 10:14you're taking that pitch and you're doing a better job
- 10:17putting into action, right? How do you use pace, speed, tone
- 10:21on, especially like for an STR who has to do this on, let's
- 10:23say, a cold call, right, to sound confident with a pitch?
- 10:27And so now that's the 70% of work that I'm doing rather than
- 10:30the generation of it. I think secondarily, if you take
- 10:34a broader view of this problem, what you're going to be able to
- 10:37do is iterate better on this, right?
- 10:40So like we actually should orchestrate conversations in a
- 10:42way that like the LLMS will get an advantage from that
- 10:47conversation. A good example is something we
- 10:49may have to do in the future is not topic switch so much.
- 10:53If you can actually stay on topic within a conversation,
- 10:56then the LLM has a better job of knowing that that topic in its
- 10:59entirety is the information set that it should know about that
- 11:02topic, rather than a scattered series of white conversations
- 11:06that is going to parse out kind of incorrectly.
- 11:09It's the same thing. Like I get really explicit about
- 11:12to do's on calls now, hey, by next Wednesday I'm going to do
- 11:15XY and Z, and then by the following Wednesday I'm going to
- 11:19do A because I know the LLMS can then take that batch and create
- 11:23my task for me automatically. As a reminder.
- 11:25We probably have to get a little bit more skillful with how we're
- 11:28orchestrating meetings and conversations to leverage the
- 11:31technology appropriately, and I think the games are going to be
- 11:34X compared to what they are right now.
- 11:36I actually love that term that you use, conversation,
- 11:39conversation orchestration, right?
- 11:41It's like talking with the expectation that the LLM is
- 11:44listening and structuring it so that they can actually, the LLM
- 11:49can actually do a good job of making sure that the work
- 11:52translates into more vibing rather than more grain for you.
- 11:56And it's almost like, you know, as an executive, you might go
- 11:59into a meeting having a really good structure of what that
- 12:02meeting looks like. And it's your responsibility to
- 12:04hold that meeting with that structure.
- 12:07But now everybody has to be the executive because because the
- 12:10LLMS are the people attending the meeting really.
- 12:14Yep, exactly right. That's a really good
- 12:15installation of that. Yeah, yeah.
- 12:17And and you know, going back to conversational intelligence,
- 12:21right, So a lot of conversational intelligence
- 12:23right now focuses on content and there there are some things
- 12:27around like, well, did you pause?
- 12:29How much did you speak? How do you think this
- 12:32conversational intelligence will change to pick up on the things
- 12:36that I was talking about, for example, coaching the STRS on?
- 12:39Like, are you picking up tone? Are you listening well?
- 12:43Are you responding to the right questions?
- 12:46You know, did they really ask this thing?
- 12:48Yeah. That's a good, a good way to put
- 12:50it, right. So if you look at, it's kind of
- 12:52like the index example we talked about the index of a book
- 12:55searching for a word and finding it.
- 12:56In the past, you had like a number.
- 12:59Hey, good. Reps talk 40% of the time on
- 13:03their calls and this is their pace.
- 13:05And you had that kind of like dashboard as to how you were
- 13:07speaking, right? The LMS I don't think are quite
- 13:09there yet for this, but I think we're going to get there quickly
- 13:12is that you could actually combined a couple of factors in
- 13:16what you said, the content, the speed, the pace and then the
- 13:21relevancy. Like did you respond to the
- 13:23question asked or did you ask a question that was kind of
- 13:25relevant to the stream of the conversation?
- 13:28And then how you're like conversational counterparty
- 13:31reacted to that, right? So like, what was the reaction
- 13:34you explicit, let's get really fundamental cold calls, right?
- 13:38There are a series on a cold calls, 4 to 8 objections that
- 13:41you should be able to handle on a call, right?
- 13:43And really the answer is more like 4.
- 13:44That's so four, I'm not interested.
- 13:46I'm walking on a plane, which is a common one.
- 13:48Everyone's flying when you're cold calling them, right?
- 13:50Like the reactions to that are more important than what you
- 13:55said because what you're said is repetitive.
- 13:57But you can find those two. You can combine the analytics of
- 14:00how you set it, pace, speed, what you said, and then what
- 14:04drove the interaction and then your response to that
- 14:06interaction. I think we're going to be able
- 14:08to get layered to be like, not like you spoke 40% of the time
- 14:11or at this pace, but what you say generally elicits this
- 14:16reaction and then you respond this way when you book meetings.
- 14:20The appropriate way to do it is a different messaging or a
- 14:23different response, right? Like that back and forth could
- 14:26be really valuable. LLMS aren't quite there yet.
- 14:28I don't think it does a great job of relating this chunk of
- 14:30conversation to this chunk of conversation to this chunk of
- 14:32conversation and seeing how the interaction lens.
- 14:35But I think we're like two or three model improvements away
- 14:38from that type of like how is your information being received
- 14:42in the way you're trying to convey it?
- 14:45Yeah, that's really interesting. So talk to me about what that
- 14:50closing the gap on conversation to action looks like today and
- 14:54then what it may look like and let's say nine months, 9 to 12
- 14:59months. Yeah.
- 15:01So right now what we need to understand is it still comes
- 15:04down to like good old fashioned understanding people score
- 15:08workflows, right. So I'll give a good example.
- 15:11We talked a lot about sales, let's talk about product
- 15:13management, right? The ideal for what product
- 15:17managers should be doing is taking feedback from their
- 15:20internal teams, their prospects and their customers and
- 15:24collating that product feedback, perceived gaps, feature lapses,
- 15:29bugs, those types of things and using it to prioritize road map
- 15:34for the engineering teams. All of that first part happens
- 15:37in conversation. There's no like no one really
- 15:40sends an e-mail right like around this.
- 15:42It's the most robust information around the product in terms of
- 15:45happens in meetings that are happening that were previously
- 15:48locked away. Great.
- 15:50What should happen then is like you should only work.
- 15:51You should work on the highest revenue impact, whether it's
- 15:54increasing revenue or reducing churn.
- 15:56That's what a product manager should do.
- 15:58What does a product manager actually do on a day-to-day
- 16:00basis? They are firefighters for
- 16:03engineering teams, right? Like what they're doing most of
- 16:05the time is not that coalition of work from the conversation
- 16:08piece, right? They are like making sure the
- 16:11engineering teams don't go way off base as to what they like
- 16:13build and waste a bunch of time and build product that doesn't
- 16:16make any sense. They're like they're in 30 hours
- 16:18a week of meetings just doing that, right?
- 16:21So two things there. On one hand, the current
- 16:24scenario is we should be that firefighting, we should be
- 16:27taking those conversations and finding better ways to
- 16:30firefight. So like, how do you take those,
- 16:33what they're doing right now and saying great, like I need to
- 16:36analyze what types of tickets are being described that are
- 16:39creating poor code based on the conversations I'm having.
- 16:43Great, that can happen right now.
- 16:44What needs to happen is the stuff that never happens, the
- 16:47customer feedback loop. That's the idealized version
- 16:49needs to be the AI vibe, right? We need to create pieces of
- 16:52software that casually surfaces up to the product manager.
- 16:55Hey, like this feature gap has been discussed three times.
- 16:59The opportunity amounts on those 3 discussions now is about
- 17:03800,000. The weighted opportunity mounts
- 17:05220K based on stage. Do you think this is something
- 17:08that you should work on right? Like and then it helps you work
- 17:11on that discussion. That's the casual surfacing of
- 17:14these problems in between those meetings.
- 17:17That's workflow specific is what needs to happen forward-looking.
- 17:21OK. OK.
- 17:22So you're thinking that the forward-looking piece of it is
- 17:25really that feedback loop and coming around across all of
- 17:31like, like, what is it like joining the conversations?
- 17:33Yeah, And the suggestion of like work, and it's like this is the,
- 17:36this is the context. Like, hey, you have a little bit
- 17:39of free time. Do you think you might want to
- 17:40take a look at this because it might be high priority now?
- 17:43That's not high priority. Cool.
- 17:44Take it off the talk. Take it off the list.
- 17:45Right. It says surfacing that context
- 17:47at the right time because there's so much more
- 17:48information. I think it's going to be how
- 17:51humans need to work with all of this information flying around
- 17:55over the next like 5 to 20 years.
- 17:58Yeah, yeah, yeah. So it's like, it's kind of like
- 18:02like distilling maybe is not the right word.
- 18:03I don't know like unifying, aggregating all of the above.
- 18:07It's like everyone has an EA, right?
- 18:09Like you're an executive assistant, like you run into
- 18:11your office, they're like, hey, don't forget you have like a
- 18:13doctor's appointment this afternoon, right?
- 18:14Like you need to have like a, an AI EA that's kind of like trying
- 18:18to prioritize work for you, surfacing it, but then also
- 18:21helping you achieve that work. It drives you to the doctor's
- 18:23appointment in the scenario, and then you go do the doctor's
- 18:25appointment yourself. What do you think, what do you
- 18:27think when it comes to conversations where most people
- 18:31do have some kind of conversation intelligence right
- 18:33now, right? They're they're recording their
- 18:36calls, transcripts, etcetera. What do you think the biggest
- 18:40sort of opportunity, like easiest opportunity is that
- 18:44we're not doing today? That could just be like a little
- 18:48switch. Yeah.
- 18:49So if you wanted to just like no tech way of doing this, right,
- 18:54in theory, what you could be doing in 15 minutes, but it's a
- 18:57tough 15 minutes to find each day is taking the transcripts
- 19:00from Zoom Teams meet, taking those transcripts and putting
- 19:03them into like clawed projects. Like here's a sales meeting,
- 19:07here's a sale. Maybe you get down to the deal
- 19:09level if you're like an enterprise sales Rep, right?
- 19:11Like or here's my product meeting, here's my internal
- 19:13engineering meetings and you're collating that data in the a
- 19:17knowledge base at least. Then when it's time to like
- 19:20execute work, you could help getting that work executed in
- 19:25that specific project, right? So like I have a specific
- 19:27feature I'm talking about. Here's a transcript, right?
- 19:30So you can start to like push that conversations in there.
- 19:33If you're in five hours of meetings a day, you should take
- 19:36the 10 to 15 minutes to go grab the transcript, paste it in the
- 19:39cloud, right? Like now, like the five hours of
- 19:41work is not just lost, it's like a value part of my ecosystem.
- 19:47That's probably the easiest low tech way to start to get off the
- 19:49blank page. And then really forcing yourself
- 19:51to use that every time I go do work regarding a specific topic.
- 19:55How can I use all of the conversational intelligence I
- 19:57built into a cloud project and to help leverage that work?
- 20:00Yeah. So build your own knowledge
- 20:01piece of call transcripts by putting it to one of the
- 20:04existing LM tools out there. Yeah, agreed, exactly.
- 20:09Right, Super simple, get off a blank page and then you just
- 20:10start to figure out what's the thing that I'm doing
- 20:12repetitively that something might be able to automate for
- 20:14me. Yeah, Let's go back to sales
- 20:16workflow again. That's one of my favorite topics
- 20:19here. What do you think?
- 20:20What are some success stories that you have seen with sales
- 20:23teams really effectively using a conversational intelligence to
- 20:28accelerate their deals? Yeah.
- 20:29So one thing I heard recently, and that's was kind of off the
- 20:32wall and again, this is a great like low software way of doing
- 20:36it, is how sales managers were managing coaching.
- 20:39So the workflow that was described to me was and they
- 20:42actually have a cloud project or an open AI project for every one
- 20:45of their reps. And so in the, in those
- 20:48projects, they will cherry pick, let's say 3 meetings maybe the
- 20:51reps had that week with their prospects, hop those transcripts
- 20:55in into that project. Then they'll record their own
- 20:59coaching calls. And the combination of this does
- 21:02three things for them. One, walking into a coaching
- 21:05call or prepping for a coaching call.
- 21:06You get to look at historic trends with your coaching calls
- 21:09and the calls that they're having, right?
- 21:12What really simply, what things did they say they were going to
- 21:14do in the past week? Something I always forget as a
- 21:16sales manager, I could the reps are always could just be telling
- 21:19me the same task and I'm coaching them the same things I
- 21:21would never remember. So that's that's really
- 21:23functional way to do it. But then second, they looked at
- 21:27great, what could I be coaching my reps on?
- 21:29Based on the conversations we're having in the conversation,
- 21:32you're not listening nearly as much as you should be, right?
- 21:35You're thinking about what you're going to say.
- 21:37The AI is always listening, never misses a beat.
- 21:39So it's really good to be like, great, what trends are we
- 21:41seeing? How can I coach them better on
- 21:42my next call? But this is what was really
- 21:44innovative. Like when you put that that
- 21:46transcript from your coaching discussion, you run it through a
- 21:49series of like pre written prompts as to how you could have
- 21:51done better. When did you get interrupted?
- 21:54What did you coach them on what? Like when did you tell them you
- 21:58should use the Socratic method of questioning when you're
- 22:00coaching? Did you tell them too many
- 22:02things and not ask them enough things?
- 22:04Right. Like, and then that post game
- 22:06could make you really like a better sales manager.
- 22:09And then in kind, what'll be interesting is the Rep should
- 22:11act accordingly as well. They should prep for the call
- 22:13like this is what's going to happen on the call so that we
- 22:15can get like a seamless like kind of like view as to like how
- 22:19can I spend as much time growing as a Rep but also getting help
- 22:23identifying risks and closing deals.
- 22:25Yeah, yeah, I I like that. It's it's like you as a manager
- 22:29can have a coach for yourself too.
- 22:31You know, we always think about like coaching as like the
- 22:33manager coaching the Rep, but we forget that a lot of managers.
- 22:37Also. Agreed.
- 22:38Yeah. And that's a layer that never
- 22:39happens, right? I never get coaching on coaching
- 22:40like my VP of sales instead of my calls as a sales manager and
- 22:43say you should coach better. Like that's just like a black
- 22:45hole when it comes to like coaching On top of it, I think
- 22:48it's a perfect example of AI. It seems like it's going to make
- 22:51things less personal. This actually made it more
- 22:52personal. Now I have a personalized plan
- 22:55to how I work with like Sarah, my Rep right Like and what we've
- 22:58done in the past. It's a good way to like use the
- 23:02vibe code portion of like AI to then grind out a better personal
- 23:06relationship. This would really help
- 23:08fundamentally improved human to human interactions.
- 23:11Yeah, yeah, yeah, yeah, yeah. And, and the other thing that
- 23:14you said was, you know, the reps can use it as a little prep tool
- 23:18for themselves. And I see a lot of reps coming
- 23:20to, you know, one-on-one. What do you have to do?
- 23:22Nothing. I've got nothing on my plate.
- 23:24But they actually have, you know, 100 things that they, they
- 23:27could be working on. And so it is also like giving,
- 23:30it's actually also giving back some more autonomy and agency
- 23:33back to the Rep to be able to say, OK, these are all the
- 23:35things that you could be talking to your manager about if you
- 23:38want to leverage it, right? Agreed, agreed.
- 23:41As a Rep, you want to it's like you can see over the past seven
- 23:43calls, like there's a common theme that you're like manager
- 23:46has been talking about. You get ahead of that theme,
- 23:48right? Like I'm talking as a manager, I
- 23:50want to stay really stable on how I coach.
- 23:52If something is off the wall, right?
- 23:53Like and it's like, listen, we've never talked about that
- 23:55before, Like we've never talked about the strategy.
- 23:58We've never talked about like this particular level of like
- 24:00questioning this is new. Like you can get ahead of that
- 24:03conversation as well. I try not to surprise a Rep.
- 24:04I try to. The repetition is the plan.
- 24:07It's a really good way to keep repetition on both sides of the
- 24:09fence. Yeah, awesome.
- 24:11OK, Warren, let's switch topics a little bit here.
- 24:14I wanted to ask you, you know, since you're a three time
- 24:16founder, what is a, what is a, what is a business idea you had
- 24:21in the past that you wish AI was there for?
- 24:27That's a really, really good question, a business idea.
- 24:30I've had me think about this a little bit, let's say this.
- 24:34So my background is in shipping, right?
- 24:38The core of shipping, the whole global economy moving goods
- 24:44around is managed in the e-mail. And the reason it's managed in
- 24:48the e-mail is that there is 10s of thousands of vendors all
- 24:52across the world to help execute complex shipping.
- 24:56There are trucking companies in Beijing that are the best
- 24:59truckers to move freight in Beijing and there's no global
- 25:04consolidation of like one trucking company that does it
- 25:06great. Like somebody spends up a new
- 25:07trucking company with four trucks and they are great from a
- 25:10rate and service level on that lane.
- 25:12So it'll be like this for a very long time.
- 25:14You'll never see consolidation in shipping.
- 25:17And so when you have that, you can't have like the centralized
- 25:20ERPSAP to SAP connection to move information around between those
- 25:25counter parties. So e-mail is the best place for
- 25:29shipping and it will be for a very long time.
- 25:32So if you think about that, e-mail is unstructured
- 25:34communication, right? And it's also very reactive and
- 25:38shipping. And as well, my customer sends
- 25:40me a message and I respond. There's very little proactive
- 25:43messaging and shipping because there's too much of it, right?
- 25:46Like there's too much information with like LLMS. What
- 25:50should be happening in the near future is we should be
- 25:52leveraging in that conversation base, right?
- 25:55How can I proactively inform my customer of service problems or
- 25:58rate issues along the way? How can I be getting ahead with
- 26:01my vendors to make sure I'm ahead of issues?
- 26:03If you think about like a typical like operator in a
- 26:07shipping company might be reading like 300 emails a day
- 26:10and responding to 50 of them, right?
- 26:13That's a perfect place for LLMS. I started a digital freight
- 26:17forwarding company in 2017. The comparison of what digital
- 26:20meant in 2017 and what digital should mean now is like night
- 26:24and day along the way. So that would be the best ways
- 26:26if I can go back in time and just push LLMS. We push machine
- 26:28learning in there, Machine learning and LLMS just like, you
- 26:33know, not anywhere close from a technology standpoint.
- 26:36So it sounds like a gap that the shipping industry could really
- 26:38like, like use could really mitigate with with LLMS right
- 26:42now is is give more proactive? I have a hard time because I'm
- 26:46so communication focused and this is another communication
- 26:49like piece, right? Like everything in the shipping
- 26:51industry is trying to use the like LMS to like scour data to
- 26:56predict service failures. It just never happened.
- 26:59It's not going to happen that way.
- 27:00Do it. A better way to move all like
- 27:02shipper intermediaries do is move information around between
- 27:05vendors and customers. Move that information better.
- 27:07LMS are great at taking huge unstructured data and coming up
- 27:11with a good. This is what this person needs
- 27:12to know. Do you want me to send something
- 27:15like this along? So that's what I that would be
- 27:17ideal. I might even start another
- 27:19digital freight forwarding company right after this call
- 27:21just based on this conversation. Yeah.
- 27:23And you know, I want the analogy that just popped in my head was,
- 27:26you know, earlier in this call we talked about work basically
- 27:30having turned into like moving conversations from one place to
- 27:33another. And shipping is basically moving
- 27:35goods from one place to another. That's really that's really sort
- 27:38of everything has boiled down to what we move and from whom to to
- 27:42where. Yeah, that's the point.
- 27:44Cool. OK.
- 27:44Last question here is, well, if you were not building this
- 27:48company, what would you be doing?
- 27:50Yeah. So I think the nature of like
- 27:54startup SaaS work is that it's, it's a wonderful like community
- 28:00being it's, it's a wonderful, but stressful, like career
- 28:04choice to make, right? It's by nature, it's a feature
- 28:07on a bug that it's unstable, right?
- 28:09You don't know where it's going to go, which is great.
- 28:10You take it so many different ways, but the same token,
- 28:12there's a stress level to that, right?
- 28:14So it's wonderful. And so I started when I was 30,
- 28:17I'm 40 now, hopefully wrapped up by the time I'm 45.
- 28:20I think it's important. What I've really come to more
- 28:22recently is like got to get back to something that touches the
- 28:25physical world, right? Like for the most part, I've
- 28:27moved around ones and zeros from the past like 10 years, right?
- 28:30So like there's an impact when you really breakdown industry
- 28:34wise, what's touching people's lives, food, healthcare,
- 28:37shelter, right? Like these are like there's
- 28:40actually a limited amount of things that like touch a human's
- 28:43day-to-day life, right? Making an impact in something in
- 28:46that space is definitely on the this is my last software go
- 28:50round, right. Like how do we actually like use
- 28:53just broader technology improvements, maybe not LLMS,
- 28:56but to make an impact on it's not a going to be a consumer
- 28:59product. That would be something like
- 29:00problem solving on the front lines of like, you know what,
- 29:03today my life was a little bit better because I did this thing.
- 29:06I don't know what that is yet. Am I going to give it one more
- 29:08go at software? But like, beyond that, I'm, I'm,
- 29:11I'm out of the SAS game sometime soon.
- 29:13Wow, OK, that's the big claim. We'll, I'll talk to you in two
- 29:16years and and we'll, we'll if. You have any idea?
- 29:18Shoot on my way. I haven't got a blank complaint
- 29:20here. Right now I have to ask Claude
- 29:21what I should do in five years. OK, Perfect.
- 29:24Warren, thank you so much for joining us.
- 29:25Is there anything else that you wanted to leave the audience
- 29:27with? Well, this has been a great
- 29:29conversation, Laura. I've been following your podcast
- 29:31for a few months now and I appreciate it.
- 29:33This is really well done. Awesome, thanks so much form and
- 29:35to our our listeners, if you like this, please drop a
- 29:40comment, send a like and then we can send you more of it.
- 29:42Thank you so much and see you next time.