Charlotte Ward: 0:12
Hello and welcome to episode 318 of the Customer Support Leaders Podcast. I'm Charlotte Ward. Today welcome Brian Levine to talk about who's got time for automation when they've got a queue to work through. Today I'd like to welcome back Brian Levine. Brian, so lovely to have you back. I feel like, well, I would like to say it's been an absolute age since we talked, but only on the podcast. Um we've talked quite a lot recently, haven't we? Haven't we? But nice to see you again.
Brian Levine: 0:47
Yeah, we've been practicing for this. I appreciate it. It's good to be back.
Charlotte Ward: 0:52
Uh it's been quite some time since we recorded. Um and things are moving at a rate of change for everybody, both in the business and on the work in front of them. And, you know, people are moving around a lot. So do you want to just remind everybody
Charlotte Ward: 1:10
who you are and where you are and what you're up to?
Brian Levine: 1:14
Yeah, so uh for those uh hearing me for the first time, uh I'm a support nerd. I've uh run support teams at GitHub and Plaid and Abstract and Meter and a couple of other places in between all that. Uh a few years ago, I left the support leadership world and I founded a company where we were building support tools. We built a help desk, we built um some uh support analytics software, uh company called Yetto with some friends of mine. Uh we shut that down early this year, around like February, March. Uh and I have spent the past you know six to nine months uh building other support tools. And I don't know, it I feel like keeping up with with the tech right now in in tech in general and AI, uh, but also just the support space more specifically, it's like a full-time job. Uh so that's kind of what I've been doing as a full-time job for the past several months. Uh, and only in the past like couple weeks I've decided to go uh just fully uh as a consultant. So now I'm I'm consulting, I'm doing contract work, uh helping people set up their like AI systems and build their internal tools and get everything organized because people who don't have the time to invest uh you know 80 hours a week just learning this stuff, like I can do that for you.
Charlotte Ward: 2:49
Awesome. Awesome.
Brian Levine: 2:50
Um it's exhausting.
Charlotte Ward: 2:52
No, it's it's exhausting. Uh and and it is and it can like it can it can feel like a full-time job. I mean, we're all, I think, I hope, well, and this is gonna be the topic for today, really, to dig into, but we're we all should be trying to build stuff because there's no excuses anymore, right? Um but yeah. Yeah, but I and but the the thing you said about, you know, it is a full-time job, I just find it fascinating, like with how much how one anyone, um, let alone me uh personally speaking, like how I balance the the daily build of doing support operations versus the the daily kind of uh BAU support leadership. You know, still got a team. Yeah.
Brian Levine: 3:44
I think I think this has been uh this has been really interesting to watch over the past uh about the past like 12 months, right? The end of 2025 when a lot of this AI coding tooling, right? You know, plug and codex and all these things are getting good enough that like it really lowers the barrier to entry, it really opens up the gates, like anyone can go in and build stuff. You see a lot of folks who had not been engineers before, right? Who had not shipped software before, getting into building things and automating their work. Uh but then like the cycle the I don't know, this pendulum is going back and forth. Like everyone's investing time all through January and February, right, earlier this year in building, and then they're like, ah shit, like I don't have time to invest in building. I'm gonna ignore that. Like and just get back to like working in the queue. And people are I mean, this has been going on for like a year now, but I feel like people are
Brian Levine: 4:40
still trying to find a balance, right? And you talk to engineers and they're still trying to find a balance between what do people do, what do bots do, how do we ship things. Like for people who are not in the world of shipping software for their job all the time, it's very confusing. It's it's really difficult to manage your time, manage the queue, to manage your team, uh, and also try to like build out stuff to help your team and automate some of this work. Uh it's a lot. It's it's just so much.
Charlotte Ward: 5:12
It's a lot. It's a lot to navigate, it's a lot to find like spend like carve time out in the day or the week to think about where the opportunities are, let alone build it. Um but you've got to make time, haven't you? I I I think like the the entry level of I can't believe actually we're here in late 2026. It is late 2026 at this point, that's based it. Um, and we're still I'm still coming across people who don't have an AI assistant and and are claiming I don't have time for that. Like, how am I gonna find time to like find time? No excuses. It's just it's like literally, you have to be able to put down the queue a little bit because the world is gonna run away with it any day now.
Brian Levine: 5:58
I will I will let all of your listeners know that you have completely got me hooked on the Claude Cowork assistant thing. Uh that was not my not my cup of tea, so to speak, uh until pretty recently. I my automation work has been uh more like custom automations in a queue. I've built a bunch of like open source projects around that. Uh but then you walked me through like a couple of your like blog co-work tasks, and I was like, oh maybe I could play around with that. And it really is like I mean, you and I talked about this last week a little bit. Spent like 10 minutes a day. It's like, what is it that I just did? Right? What did I just spend an hour doing? Like how much of that annoyed me? What do I not want to do again? Uh and just like tell tell Claude, tell, tell your bot, like, I don't want to do this thing again, like come up with a way to automate that. And I I think a lot of folks in support uh tend to look at like two big things is gonna get ranty. Uh the the you look at like two big buckets of where you can like implement not just AI, but like any automation, right? You're you've got a lot of people look at it as like your customer-facing stuff. Like, what workflows can I automate? Someone's like sending in the same five tickets, and I want to just have an answer bot just like answer those tickets, and you can get AI to hook up to all your systems and solve all these problems for you. And that's very cool. Like that really does work in a lot of cases. It takes a lot of time to set up and like weeks of work, and and that's very it's daunting. Right. So then you're like, all right, I'll do internal automation, right? Tickets come in. Tell me what docs I should look at, tell me what's open on GitHub, tell me what Jira tickets are like related to this. And it seems like that's an easier lift. It's not customer facing, so if it breaks, like who cares? But like your team cares and it still takes weeks to set up. And there's this third bucket that a lot of folks just kind of aren't working on as much. That's like for me, right? It's personal. Like, what am I doing? It's not necessarily what the team is, it's like start small. Start like this was your lesson to me, right? Start super small and then like expand, like then decide what can be shared with other people or what other things can we do. But I but yeah, I think we're we're all able to do a lot more uh with the tools available right now than I think we realize. Uh we're giving the tools a lot of credit and capabilities that they don't have yet and ignoring what they do have. Uh and I I think that's it's a really interesting place to be. And sometimes you just need someone to like lecture you and tell you.
Charlotte Ward: 8:50
Like I do apologize for that.
Brian Levine: 8:53
But it but it like it works, honestly.
Charlotte Ward: 8:56
Like that's being that was quite a screen share, wasn't it? It was like that's it, right? But but I think it is start small. Like, I I genuinely believe that. And I've been saying for years before I was doing the AI build, just solve the problem in front of you. And we've all operationally knowledge, workflows, customer relationships, communications, all every part of that um post-sales ecosystem, an operational ecosystem that support touches, has a ton of tiny
Charlotte Ward: 9:30
opportunities in it before you get to the big scary ones that you talk about. And frankly, starting with the tiny ones builds the muscle, it builds some of the groundwork, probably, and um, and the skills necessary to get to the big ones. And the tiniest thing can be life-changing if it's the stone in your shoe. Um, one of the first things I did near like a year ago was a tiny, tiny little thing, the a little skill that took an the awful PDF export out of Notion and just created a much better formatted, branded PDF that we could share with customers for certain document types. It was a tiny little thing, but you know what? And it was only something I had to do like twice a month or something. But I was taking an export of the pages, HTML, going and opening Chrome and DevTools and editing this bit and shoving that bit in there, then exporting that page, that rendered page as PDF. What a pain in the backside. Like just so easy now.
Brian Levine: 10:34
Like and going through that manually doesn't benefit you in some way. It's not like, oh, but I learn more by doing it myself. Like, you don't. This is just tedious work that you'd rather not do. That maybe it it maybe it only takes 10 minutes twice a month. But like, wouldn't that be great not to do it? And I I think your point there that like it does build a muscle for like figuring out how to automate things, figuring out where you're wasting time or where you just don't want to spend time. It's not even a waste of time, you just don't want to do it. But also it trains you to look for these things. I had this amazing conversation with someone I worked with. This is like, I don't know, 10 years ago. Uh I was building some sort of automated QA tool for a support team, and he was like, Man, that's great that you built that, that you saw the need, and and now the team has this tool. But like, how do you get these ideas? How do you figure out that? And I'm like, man, the ideas are everywhere. The ideas are hard for us all around you. It takes like a little bit of training to like see them or or to like figure out which ones are worth pursuing or not pursuing or just half-assing, like some things you could just put like the most minimum amount of time into and get something back that feels yeah, right? And some you're gonna sink a ton of time into and not really get anywhere. And it takes a little bit of practice to like see where time is well spent and what ideas are worth pursuing and what ideas are just fun, right? Like honestly, I right, so I'm getting into this consultant, going, I'm going full bore, I'm gonna consult, I'm gonna get some clients. Uh, and so that involves like finding uh a marketing channel, right? I'm on LinkedIn. So I'm like, all right, find clients on LinkedIn, find engagement on LinkedIn. Like, how much time do I spend just like trawling through LinkedIn garbage? Why should I do that? Why should I be the one like I could hire like an SDR or I can say, all right, Claude, just like go through my LinkedIn contacts and find the ones that like aren't vendors that have said something about this problem recently. Like, don't do outreach. I don't need AI generated outreach. Like, I'm a person, I could talk to people, but like I don't need to sort through thousands of posts on LinkedIn every morning.
Charlotte Ward: 12:57
Yeah, yeah. It's the little things. I remember, um, and I'm sure you're you're much more of an engineer than I am. Um, there's that um table, which is um how often you do something, how how long it takes, and then how long you should spend automating it as a calculation. So the ROI effectively of the time invested in automation. You mentioned a different word there a few times in in different ways. It's a big, I think we need to make that table three-dimensional. It's missing a dimension, and that is how much it annoys you to do the thing. Um, because that that's not purely time-based. And I think when when you think about how okay, maybe maybe actually trawling through LinkedIn might only take you 10 minutes in the morning, but if you hate it with every fibre of your being, why do it? You know, um, regardless of whether it takes 10 minutes and you know, maybe it took you 15 to automate one
Charlotte Ward: 13:55
day. But but ultimately the um the to your point, the opportunities are there. And this that's really what I mean when I say solve the problem in front of you, because they're there, they're right in front of you. Spotting them is just opening your eyes.
Brian Levine: 14:09
And real, how many people how many people listening to this are going through their queue every like Monday morning or Friday to put together like a voice of the customer report or something for your your C-suite? Like, you want to have like your opinions and your methods in there, but it could also take like 10 minutes to tell Claude, like, here's what I look for in the queue. Like, find me five tickets about this, find me the top three issues. Like, I can go through the queue and like confirm that, like, give me references. So I I think the LinkedIn thing is like a good example of a thing that I hate that was like an easy one to get started. And then looking at the queue is like, all right, that's the next thing. And like, where's the next one after that? I feel like you know, in that in that graph you mentioned, right? The the effort to automate versus like what it saves you, the effort to automate is changing. It's changing like so rapidly right now.
Charlotte Ward: 15:05
Yeah, yeah.
Brian Levine: 15:06
Like if you can spend 10 minutes, 20 minutes, once a week just like getting something off of your plate, like you you can get something out of that now in a way that like a few years ago, like in 2020 or 2021, that would have been like a day's worth of work, two days worth of work.
Charlotte Ward: 15:23
In like dinosaur times, yeah, yeah, yeah. Yeah. And it feels like that, though. It's a really good point. The the effort to automate is is, I mean, it's dropping to almost ground floor level, isn't it, for a lot of things, unless we're talking like building huge ecosystems. The thing I would say then, in addition to that, it's such a good point, is that the opportunities, this comes back to what I was talking about before, like building the skill enables you to find more, as you were saying. Um, like the thing that you build now is often not just one task anymore. So that table kind of then gets blown out of the water in a different way because you might begin by saying, Okay, I can I can automate a Friday summary of ticket, you know, queue sentiment or something. Um you solve that, and and that starts to be delivered to you and your boss by email every Friday afternoon. But but before you know it, you're like, ah, there's something else I can add in there. Yeah, yeah, yeah. That's that's that is my biggest joy, I think, with building with AI is, and that's probably how I ended up giving you the lecture on how many things I was doing with AI was sorry about that again. But the but the um but the point is like it's that moment of like, oh, that I really love. Oh, I can add this. Oh, I can do that now, because once you move that like little rock out of your way, there's the next one and the next one and the next one, and they're all equally they all have an equally low barrier to entry or fixing.
Brian Levine: 16:59
And I think the more you do it, right, the more you build the muscle of like finding, automating, shipping next, right? Is the more you go through that. You find like the edges, you find the capabilities, like what can these rules do well? Where do they kind of fall off a little bit? Where does their data drift? Like, now I know let's put some guardrails there, send this to me, I'll go do the next thing and I'll pass it back. Right. And then as you start to do that, you start to say, oh, actually, this should be like an app that my team uses.
Charlotte Ward: 17:31
Right.
Brian Levine: 17:32
I was I was talking to someone about some clog skills they made, and they're like, oh, I would really love to operationalize this for my team. Like, how do I get my whole team using this? Well, you can ship your whole team a clog skill, or you can turn this to like an app in your help desk, and this could be like a button that they have. Well, now we're talking about like building something, right? Like we've also lowered the barrier to entry there. Like, how many people could like start building a ZenDesk app with the practice they've now had in like automating their own tasks? And you know, you start to get into all right now there are security issues and there's PII, like there are all these things you're gonna have to start worrying about, but like your support people, you already know about this stuff. Like this stuff is already on your mind in a way that it probably isn't for other people. You know what to talk to your security team about, you know what permissions you're comfortable giving your tools. I feel like starting with these little things, starting with the personal things, sort of builds up the momentum to start tackling the next thing. And then
Brian Levine: 18:35
then, like here's here's where my lecture starts out with people. It's like then your build versus buy conversation when you're talking about oh, I need like a chat bot or I need a co-pilot tool for my team, your build versus buy situation looks a lot different because now you've built a couple of things for yourself already and you know what it's gonna take to build the next thing, you know what you want it to look like. Do you want to buy that? Do you want to trust someone else's product decision? Do you want to maintain that right now? If you're shipping your own product, you're gonna have to maintain it like you have more insight into those uh pieces now.
Charlotte Ward: 19:10
Yeah, yeah, 100%. Um, and you know, uh the layering on the the kind of connective tissue that this uh this um whole process creates between where like tools that you bought versus the thing you build is is a super interesting space to be because we all know that tools don't like to talk to each other. There are whole tools that have been invented to make other tools talk to each other. And so so like and I find that um like whatever tool I've bought does, if I'm lucky, 80% of what I want it to do, more realistically, 50% of what I want it to do. And so to your point, like when we come to uh talking about things like having a button in Zendesk that does the thing I actually want Zendesk to do, well, that was something that I might have been waiting eight years for Zendesk to implement, and now I've got it. I've got it on a Monday afternoon. Um yeah, yeah.
Brian Levine: 20:13
My favorite pre-AI story about this is when I was at GitHub, we had our own help desk that we had built and used, like the GitHub.com web support was totally in-house. And it started because we were using an off-the-shelf product earlier on and didn't like some of the features, so we ended up building our own like uh interface for it. So we built a UI with all these buttons that we needed uh to interact with the API, and then eventually started building other backend services, and then we're like, well, we just basically built a help desk.
Charlotte Ward: 20:48
That's just the entire and that took like I don't know, years, right?
Brian Levine: 20:54
And then we had a team of people maintaining our in-house help desk. And now, like, I don't think that would have taken years. I think it would have taken a whole team to be like, actually, what I only right, I'm only using 50% of the features that the product has, and it only has 50% of the features that I want. So why don't I just build the features into the other tool that I'm using and just be done with it, right? It's yeah, it's a much smaller project now than I think people realize.
Charlotte Ward: 21:25
I I I completely agree, and yet there is kind of a fear factor of getting off the ground. How do you solve that?
Brian Levine: 21:35
Again, start small, right? Just start start building something, right? There's there are so many opportunities now to just get started, right? And you know what your biggest problems are, your most painful maybe there's small like like you said, maybe they're not your biggest problems. Maybe there's the most annoying problems. Like just get started. Just get started building one little thing, and then the next little thing, like Don't start by shipping a help desk. Start by shipping like a a cloud skill that will automatically search Jira for issues related to the ticket that you're working on. Or open up an issue in Linear or look for pull requests in GitHub. Like look for ways to automate these sort of cross-system things that you're always doing anyway. Right? Don't don't worry about making it a permanent integration. Don't worry about like the big picture. Start as small as possible and let that snowball. So all right, now that I've got this working, how do I make this like how do I make it do more? How do I make it do this for more people? Where does the rest of my team start to like hook into that? And you know, before you know it, you've you've built like a tool for your team, right? Like it it could be something like a script that runs once a week to run like QA on your tickets. It could be, you know, your voice of the customer program. It could be like a ton of little things, but those things snowball.
Charlotte Ward: 23:01
They really do. They really do. Yeah,
Charlotte Ward: 23:05
I I couldn't agree more. Um there's another there's another um type of build, which I think I've just, as you're talking, I've come to reflect on is slightly it's it's a different value prop. The process is the same of building it, but I think it's a different value prop. It's not about automation saving you time. It's about I think it's about uh creating things that give you different perspectives. And so I as you were talking, I was thinking like, yeah, there is a ton of operations to automate, and there's a ton of there's a ton of like uh to your point, um smaller scale tools that we can build that perform a function, for instance, you know, ticket. We can build a help desk if we want, we can build a we can build a support agent if we want, we can do all of the things. One thing that um I've just realized is um there's an opportunity for all support people, I think, to just get a different perspective on what's in front of them. And so my assistant at Snowplow, for instance, one of the things it does every morning, aside from connecting and telling me about KPIs across the team and you know, scouring my email and my calendar and telling me where I've, you know, I've got some Slack to do yet more building. Um one of the things it does is look across the whole queue for signals that are really hard to surface when you're just in a help desk queue looking at the queue. Um and it isn't a problem of volume, it's a problem of depth and nuance. And so it will scan the whole queue and it will look for things like buried or implied commitments that my team has made to a customer that it's not sure if we are able to actually meet. Yeah. So did we promise to do something that's just buried in a comment somewhere? I'll get back to you on Tuesday with that thing. And but by the way, like this is actually dependent on some deployment that we know isn't happening till two weeks from now. Like those kind of things, it will surface to me. I there's no way. Like maybe the support engineer wouldn't spot that that dissonance. I couldn't going through hundreds of tickets, and so looking for kind of um uh different perspectives, like look looking for more nuanced um information, um, or just uh, you know, um different priorities. So I think uh I think something that I've often struggled with um both when I was frontline support and leading teams with a big queue is how do you prioritize just inside the queue what to do next? Now, yeah, right. Now most uh help desk software will offer you a view the ordered by creation day or ordered by priority or ordered by SLA and not much else. Right. Um actually there are potential opportunities or let's say um save moments further down the queue that you might not get to till tomorrow because they're only a CEF three, you know, and you've got a bunch of ones and twos that are fighting for your attention. Before AI, particularly before AI, I used to spend a lot of my time trying to just sort by different things on the queue. And I would sort of what can I find by okay, today I'll sort by priority and I'll look at all the sev ones we have open. Tomorrow I'll sort by oldest tickets. The day after I'll sort by last updated date. The day after that, I'll sort by number of interactions that look for the ones that we can do. Do you know what this sounds like?
Brian Levine: 26:59
You know what this sounds like? An automated task. Sounds like a task that you could give to something.
Charlotte Ward: 27:05
It does.
Brian Levine: 27:06
This is what I do. I search for these in these ways, looking for these things. You can just have to do it every morning.
Charlotte Ward: 27:14
It is a task. It is a task. The task is go and look at the queue, but there's a layer on it which is not just those signals. It's not just about priority, and it's not just about created date or SLA. It's about what's deep in the tickets that is a different signal that I should respond to. And that's what I'm talking about, the different perspectives. The tools only give you a certain number of perspectives. A help disk only gives you a certain number of perspectives.
Brian Levine: 27:41
One of my one of my favorite things that I built uh this is like a year ago, uh, one of the last things we had at Ghetto was uh this dashboard that you know you could query your your queue and get all this data out of it, but it would also run a bunch of things every, I think I had it running every six hours. Look at the queue and look at all the data it had available. And for every type of query, it would go by your tags, it could go by intent, like whatever. It had a bunch of of heuristics for this. For every type of query, it would say, all right, how many do we normally get like per hour per day? And is our current rate an anomaly? And it would find this it would find like these low volume issues that were starting to spread. You wouldn't notice if you were just answering these tickets. Maybe you know, you only see out of you know hundreds of tickets, you only see five of these. But this week we're seeing like 10, 15, and then 20, and it's still not enough to like notice it, but you know, the bot noticed it, and it could tell you before that sep three becomes a sev one. And like there are all these, right? You could go for depth in tickets, you could go for breadth in tickets, but these are all things that are like hard to do manually. And when you have your the the impetus for building this was that I had these teams who were like were telling me that they they didn't have the data to back up this gut feeling that they had. They're like, I feel like there's something wrong with like credit card processing in this product, and I can't put my finger on why I think it's wrong, but we're getting a couple questions that lead me to believe, and I can't take that to my product team because that's not enough. Like that's this is a hunch. And like you can take that hunch and be like, all right, dig into this and find out like is it a hunch? Is it real? And then go find other hunches, like go find other things that I'm not noticing.
Charlotte Ward: 29:46
Absolutely.
Brian Levine: 29:47
There's so much we're not noticing. We're only human, right?
Charlotte Ward: 29:50
We're only human, and we can only go so far and so deep. And yeah, like we're busy building stuff.
Charlotte Ward: 30:00
Yeah. Um, when we're when we're thinking then, just to bring this back to the start, um, we've talked about, you know, deep diving on the queue and automating the things that are annoying or time consuming or that you just flat don't want to do anymore, um, or raising more insights. Um, let's just bring this right back to the first question, which is how do we find the time? Because um it is start small, but even that's a barrier for some people, like particularly support people, it's a highly reactive role. Um every day of my working life, because I've been in support for 31 years now. Every day of my working life, one of my colleagues has said, I haven't had the time to do XYZ, whether it's you know, call that customer, respond to that email, do their QA for the week. That list never ends up, it never ends. The queue never goes away either, does it? So it just, yeah.
Brian Levine: 30:57
If you're gonna take five or ten minutes out, that queue's gonna be there regardless. Like you're not actually gonna make a dent in those ten minutes. But I think there there are a couple of things here. One, if you're not already, you should be like taking notes. And at the end of the day, take five minutes, and this is like for years I've been I've been telling folks on my teams to do this, take five minutes at the end of the day to like write down the things you did, the problems you had, and like maybe what you would have done differently.
Charlotte Ward: 31:28
Right?
Brian Levine: 31:28
Just like little notes, like I worked with this customer, here's the thing we had to deal with, here's the thing I hated doing, and I wouldn't do that again. Now you have an outlet for that. Right? Previously you would have all these notes, and at the end of the month, you'd be like, all right, what pro what can I correlate? What process should we implement? But now it's like, all right, at the end of every day, I take five minutes and I have these notes, and I can just feed that to Claude and be like, what task can I automate from all this? And then I walk away. Like I'm done. And I can come back tomorrow and see if I have a new task automated or a new suggestion of how to handle it next time. Like, that's just five minutes. The next step is to be like, alright, I'm gonna build these tasks and I'm gonna try a few of them. And maybe that's like ten minutes. And and this is what I did after we talked, right? The next day. It was like Saturday afternoon. I was like, you know, we're going to lunch in a few minutes. Let me just like tell Claude, like, here are the five things that I don't really like doing every week that are right. I have I have this notebook and like here are the things I did, and then I went to lunch and when I came back, it had two tasks already written. And they needed a little bit of work, they needed to be tweaked, and you know, a couple minutes here and there, and like I don't know what's it been, like four days. Uh and now I have five daily tasks running for things that like I didn't really enjoy doing.
Charlotte Ward: 32:50
That's why I'm I feel like I created a monster.
Brian Levine: 32:53
But it's terrible. Every right? And and it also is useful for things that aren't recurrent. Right? If you just ha uh I mean I need to like go through this mountain of data and it's gonna take me like two hours to do it, like may maybe it does require human eyes, right? Maybe I don't trust a bot to do all but some parts of it don't require human eyes, like collating data, data entry is not your job anymore. Like you can have a bot, take all of this like unstructured data and put it into a table, put it in Notion, like format it, give me like a nice HTML page with some buttons to like click through. Like you you can do a ton of like little one-off things that make your life easier that only take like five or ten minutes. Yeah. If you take the five minutes now, it'll save you 30 minutes in the queue tomorrow. Right? And it seems small, but it starts that small.
Charlotte Ward: 33:52
Or it saves it it really does, and it exactly, and these things can live for a moment or they can live forever, and sometimes it's about not reducing time or annoyance, it's about reducing attention, things like that. Um, this the PDF example is a good one for annoyance. The attention thing, I have one personal example, super quick. The um I recently listed a laptop on eBay. So, just for the record, to anyone listening, if anyone thinks my snowplower life is automated, they have not seen my domestic life. Um every single thing. Uh, I listed a laptop on eBay, and I don't want to keep coming in and checking on watches and like dealing with offers and stuff. So I just whacked it up like to the top end of the price range I dare like hope for. I ignored all offers and I just told Claude to come back twice a week and knock five pounds off it or five dollars, whatever you so it just reduced like tiny little and one day it just sold. And I wasn't watching it. I just left it logged and it would just drop, drop, drop, drop, drop. Got a couple of watches at a certain point, and I had an email, and I was like, oh, maybe it's getting to the right price now. I might get and then it was bang, it sold.
Brian Levine: 35:06
I didn't have to do any like any of that tedious human work of like But it's like the it's the work you've already decided what you're gonna do.
Charlotte Ward: 35:15
Yeah.
Brian Levine: 35:16
What what do you what benefit do you personally get out of logging in and doing that every 12 hours? Right? Yeah, you can automate that. And you could have automated it before. Like this is the other thing. Like a lot of the stuff we're talking about could have been automated. And like my co-founder Garn and I, we did automate a lot of this stuff before it was possible, before Cloud Cowork was and now it's so much faster. It would it I would spend a day writing like Chrome extensions to automate a bunch of stuff in the web for me. Now co-work, I can just tell it and then I can leave. And I can tell it to send me a notification when it's done. Or I can just not bother and like I'll come back to it later. I don't need to get a notification during lunch that my Claude finish a task. I'll get back to it. It's fine.
Charlotte Ward: 36:03
Yeah, yeah, yeah. Yeah. I love that handoff of just like do all of this while I'm not here. I'm gonna come back. And I've just got, I think I once shared with you the uh the little series of notifications uh that I had in my little Claude sidebar, right? It was all these things had completed. I just had to sit there and work through different sorts you know, just make sure the status of that one. Oh, you need my input on that one. Okay, you can't sign up. You need me to click sign in because it's an account action or whatever, like so much stuff. It's
Charlotte Ward: 36:30
just doing the doing the grunt work for me.
Brian Levine: 36:32
And honestly, the last point, last point, because we've been rambling for a long time, uh, is like that screen share really is what's holding for. And so if you're listening and you're like, I don't know, this doesn't sound like a thing I want to do, like sit down with someone who's doing it and and look through how they do it, right? DM me. I will walk you through it. Charlotte will walk you through what she's doing because like it's not difficult, it's not crazy involved, it's not I mean, I made fun of I made fun of Charlotte last week. I was like, this day I had psychosis. Uh but it's it's I don't think it's crazy to think like here's a bunch of tedious work I don't want to do, and I can just come back and like a bunch of it was already done, and I can decide whether or not it was done to my satisfaction. I don't have it, I didn't give it my credit card. I did hook it up to email. I d I do sometimes up to email. I'm sorry. Uh but like yeah, just just go see how someone you personally know is doing this because you will get some ideas.
Charlotte Ward: 37:38
Yeah, yeah. Um, I'm gonna confess I've hooked it up to all the systems. Like it doesn't take doesn't take financial decisions on my behalf and buy stuff, but uh email, everything, it's it's it's hooked up. Like I'm I'm wired in at this point.
Brian Levine: 37:54
Yeah.
Charlotte Ward: 37:55
All right, Brian. Listen, thank you so much for coming back um and talking to me for what I think is the fourth time in a week because we got a bit sidetracked a couple of times.
Brian Levine: 38:04
Nobody has to know that.
Charlotte Ward: 38:05
Nobody has to know that nobody has to know. Let's not mention that on the recording. Um would you would you please come back another time in the not too distant future? And uh I know we'll have more to talk about.
Brian Levine: 38:16
I would love that. This is great.
Charlotte Ward: 38:18
All right, thank you so much. Lovely to see you. Thank you. Talk to you soon. That's it for today. Go to customersupportleaders.com forward slash three one eight for the show notes, and I'll see you next time.
