← The Ops Experts Club Podcast
The Ops Experts Club Podcast
112. Is AI Going to Replace Your Bookkeeper? Here Is the Honest Answer
00:22:16
Loading summary
Transcript25 lines
- [00:01]
A
Welcome to the Ops Experts Club. If you're at all interested in anything we talk about here in this episode, go ahead and check out the description down below and click any of the links there. Or if you just want to know more about us, click the links below. Now on to the episode Ops Experts Club. What's going on?
- [00:18]
B
Experts Club.
- [00:19]
C
Nice.
- [00:20]
A
Love it. We're gonna pull Tabby into this cool song, this cool jam that Taryn's putting together, that just amazing new soundtrack. Jingle, jingle, jingle, jingle, jingle. I love it. Hey, Ops Experts, I've got some exciting stuff for you guys today. Not only do we get the amazing Tara Turner, who's getting a rock delivery later on this morning. So those of you who don't know,
- [00:39]
B
oh, love a good truck full of rocks.
- [00:41]
A
Mick Turner had a farm and he's pretty excited about getting rock deliveries. That's number two. And then. So not only do we get Taran, not only do you get to talk rock delivery today, we also get to talk to the amazing Tabby Claire. And Tabby is on our team, partner, leadership team or collab team, but also very specifically in the topic we're talking about today pertains to her is finance. Does a lot with the finance side of things. You know, we have picked up quite a few finance clients over the last couple of months and something we're always doing for different entrepreneurs out there in the online space and brick and mortars as well, where we're just coming in with solutions. So, Taryn, I thought it would be first of all, Tabby, thanks for joining us.
- [01:15]
C
Of course, Happy to be
- [01:18]
A
Taryn. I thought it'd be cool today to talk about technology, automations, AI, all things as pertains to how can we, you know, here we talk about here a lot on the show. Delegate and elevate Neos tool. How do we delegate things down so that we can elevate our top people up so we get best bang for the buck out of our people. And I think that there's never been a better time for us to discuss. It's not always delegating down to other people on team or other hires. You know, we help with all those things for clients, but can also be technology. And that's a big part of what Colab Team is doing now is, hey, where does AI fit? Where are some of the mindless stuff that doesn't even need to be people anymore? Because there's great AI agents that you could deploy to do certain tasks and then maybe lean in a little bit extra with Tabby Today on all things finance. Sound good?
- [02:01]
B
Oh yeah, I'm ready.
- [02:02]
A
Cool. I love that Taran. Let me just put Taran on the spot. Taran, Some cool stuff you've seen when it comes to people using AI for streamlining. I know that, you know, we've been in a lot of the different platforms. You know, there's a ton of platforms out there right now. I know pods getting a lot of attention, particularly the last few months. Maybe some cool things that you've seen where just technology and automations has really helped people in this past season. You know, we just started a boxer channel in the Colab team where we're like, hey, share. Yeah, like what did I call it? Something like share. Share your experience, you know, with AI or something. And people were just dropping in experience of how we're using AI and clients. I know. Taryn, you're ahead of the tech here at Cloud from leadership team perspective. Just like Tabby's head of finance. Maybe talk a little bit about some of the cool things you've been seeing with technology.
- [02:51]
B
Yeah, I think one of my favorites so far is relating to organization. I mean, when we come into a lot of these on fire entrepreneurs scaled rapidly, organization is probably the bottom of their list when it comes to where their things are and what they want their people to focus on. And so I, I've gotten to see people utilize AI just to help them organize their files physically, I mean digitally. But you know, just like, all right, we need an organization system for this file structure. How do we build it? What are the top level folders, what are the subfolders? Make it make sense to more than just one person. So I think that's really cool because organization is something a lot of people struggle with. We're really good at doing the work and moving on to the next item and doing it, moving on to the next item without getting a chance to think back about where we stored it, who's it shared with, how could people access it, how are we going to find it a year from now when we're on our 10th iteration and we're somehow coming back around to find this version?
- [03:47]
A
Especially if you're losing people, right, because we do a lot of recruiting here. The truth is your people aren't going to probably be with you forever. You know what I mean? And the danger is if they created it, if they organized it, what happens to it when they leave? Did they communicate those things to you before they left? Does anybody else know what's in their brain? Does anybody know where those things Live. So I think that's. I love super good tabby. Speaking of that, we're in that exact scenario with a brand new client with Colab Team. We saw some proof, pretty sexy stuff that they're doing that I thought would be a great stuff to talk about today when it comes, when it pertains to actually what Taran just described there, which is pretty rad. So Tabs, you want to talk a little bit about some of the things we're seeing on that side?
- [04:26]
C
Yeah, we've run into some interesting use cases I feel like that have been really. It's opened my mind to like what. How I could even use these various technologies. We've seen people, I mean similar to what you guys were just referencing, people transitioning out. They've got all this knowledge that's contextual in their brains and like the organization is left kind of thinking like okay, but what do I do without this person and all the knowledge they had? And we've even. So we saw one client build out an entire like project that was based around every process this person owns, their email archives, their just like any documents, knowledge, information that they could have all built into this one project. So it's like a searchable resource for the company, for whoever's stepping into that position in the future to be able to say like, oh, what did she do three years ago when we ran this same type of project? And like there's probably a template I could pull from or something of that nature. Like I thought that was such a cool way to kind of like ease that burden of somebody that you're is transitioning out for whatever reason. Like such a cool way to use a very, very robust technology and actually make it a robust resource in and of itself for that specific case. That was one that I just kind of woo. That is really cool. That was a smart way of using that. Um, but we see it all over the place. Organization task, like simplification I think is a big one. I use it in what are things. I'm just like manually using my brain power. I don't need to. So I mean I built a, a reconciliation template the other day because it's like we only have so many bodies of people to do things and instead of me figuring out the forward random transactions that aren't reconciling, I can redact any important information, dump this and what I have in my software into. I was using Claude and I built it into a whole project so that whenever, whenever I need to do that, it will just isolate the three Transactions that are causing a reconciliation discrepancy. And I don't have to sit there and pour over data to find that it's going to call it out for me in 12 seconds.
- [06:30]
A
I love that, Love that. Taryn, that that's kind of something that you advocate for, right? What are some end over end things? What are things that we can do? What, what are processes and pain points that people are experienc experiencing and then challenging that and saying is there a smarter way for us to do it where there's pattern recognition involved, where there's something that could be an end over end process like what Tabby's describing, that you could, you know, create a project just like what Tabby's lining out that people could come back to time and time again to roll out within the platform of their choice. I know we've used GROK a lot as well. I know Brandon, things for us from recruiting side of things or from also people reaching out for interested parties, like people that want to work with us and like research projects that are done a certain way the same way every time. Gap analyzer like how are we analyzing all the information that comes back from people when we do these audits and we spot check everybody in the company and exactly what they're working on and then identifying inefficiencies or places where we would suggest that there may be gaps. So maybe talk about that a little bit.
- [07:27]
B
Yeah. What I've been playing around with is just the ability to do something that's clearly repetitive and can be easily structured and built out on its own using code. How can code do this? How can we create a code that can't do this automatically? And so it's not necessarily using AI as it's intelligent to make decisions, but it's using it to be your coder and developer, which used to cost a lot of money to get these custom solutions. Nobody was going to pay a specialist to, you know, solve a pain point just because somebody's spending 30 minutes on it a day. But if somebody can spend a few hours implementing some code that can take over that on their own because they don't have to write it, they just have to direct it. And now all of a sudden this report is being self fulfilled every morning at 3am for the previous day's data, that's a win.
- [08:17]
A
How much time do you think goes in? Yeah, I think that's something probably for visionaries to think through and talk. You know, sometimes I think that the Achilles for a visionary is shiny object syndrome. Right where they hear something and then they just go all in on it, you know, I mean, they just jump on board. And I think that for some visionaries, they're like, AI is the silver bullet. I'm just going to jump on board of this and like, I don't need all these people. I can just go ask AI to do it. And it's this new tool and they're excited about it, but they don't really know how it works and they weren't doing the operational functions anyway. So then they can just jump in and it can start getting a little bit messy on how the actual doing of the doing goes. Like, how much time do you feel like needs to be given to, hey, let's quality control this thing. Let's make sure that it's continuing to do it the way that we want to do it and it's not experiencing. You know, I've heard the term hallucinations. Like, we're not experiencing hallucinations where it's going off on a tangent that actually you're way off downstream because we didn't Course correct. What are your thoughts on that?
- [09:07]
B
Yeah, absolutely. I like to think of it as maybe a general contractor of a house. You know, that's kind of your role in this now. You don't need to know plumbing specifically. You don't need to know everything. You've got a plumber to do that, but you still need to know the big picture. You need to know how it fits in. And so you're kind of directing it from that mindset. You can't go in as a general contractor of a house and have no knowledge of how a house is put up and expect this AI tool to build everything perfectly on its own and just implement it and you're good to go. You're going to have to have some sort of knowledge, you have to have some sort of management. You're kind of just managing a person doing a task. They can just do it within minutes instead of getting an update every day and going back and forth. So, yeah, you do want a quality control. You can build it into it as well, which is great. You can. You can ask for why or how it got somewhere. And once you get into that, it's going to actually show you how it got to it. It's not going to. If it's in conversation mode, that's when it hallucinates. But if it's like, show me the data behind where you got that. That's where it's going to be like, oh, yeah, here's the raw data. This is why I got that.
- [10:12]
A
Which is where I think Tabby that was so cool with what that new client of ours is doing is the reality is the people part of the project that is going to be necessary are those SOPs are the things that they've already put in the work. Right. That this tool is only going to be as smart as whatever it is that you're giving it. And so that was what was great was you know, this finance lady that had been with this client for all these years had all these great sops that they could just dump into Claude or whatever the platform is that they're using that then would extrapolate, interpret, take it. But because it was lined out nice and clean, then okay, it understands what we're asking it to do. It understands where the resources are, it understands what connections need to be made. It understands how it's all going to need to come back together. Instead of just throwing out random conversation pieces where you're only giving it part of the picture. You're. You're going to get kind of messed up results. So I think that realizing that probably Taryn and maybe you can speak to this as well or Tabby, the one where you're going to still need the people is like lining out the SOP of what the process was that we're trying to replace. Like making sure that the, that the AI agent is still going to do what it's supposed to be doing, quality controlling it. Like you were saying from a general contractor perspective, Taran of Is it behaving the way it's behaving, Is it still producing what it's supposed to be producing? And then from there, just like any delegate elevate project that we encourage people to like identify the things that don't require people's brilliance anymore, task those things down, ask yourself who would be the best one to fit that If AI is that solution, spending the time to line out the SOP nice and strong and then quality controlling it so the person can elevate into other function and then the time it takes them from a moving forward perspective is just like touch and base, making sure this thing is still producing the right kind of data that we need it to produce.
- [11:48]
B
Absolutely. Yeah. I built a SOP generator tool for a client and it was never gave you something a hundred percent the first time. It was always 90% which was a huge lift for people. But you always gotta teach people you have to review this. It's. I taught it not to make assumptions, to ask clarifying Questions. But it still might have outdated model information of different apps programs. It still might mix up some of the steps because just how it picked it up in the conversation with you and so you might need to rearrange things, you might need to clarify things. But 90% of the way there is still a huge time saver for sure.
- [12:25]
A
Another thing we had seen a client do that I thought was really cool was that whole thing with their emails. You want to talk about that just for a minute?
- [12:31]
C
Yeah. So that was one, that same client, but they essentially took like one person, this person who's transitioning out. They took their entire email archive, everything they've sent, everything they've received, documents that have passed through and made that a search essentially like her, made her email searchable but external from you know, Outlook or Google or wherever, which I think we all know has some like search discrepancies. Often like you're, it'll kind of miss some of the things that it really should be grabbing and having it into that format where it had all of the information over the, you know, 20 plus year use of this email in searchable catalog. And it wasn't necessarily going to only pull from one place. It was going to reference all of that data and pull everything that pertains to it and let you then sort through, you know, like if the, which, which information is prevalent to like your current situation. So I thought that was very helpful just because again we have context I think is one of the biggest things that it's hard to impart to an AI or like a, a project that you're creating. And so having just like years and years of verbal back and forth and the way that projects progressed, like all the data there I think is just a huge tool. So building something like that and I think that has a lot of different potential uses outside of just someone transitioning
- [13:57]
A
out of the company and even the ability to ask, ask AI hey, how would Vicky have handled this? Like how did Vicky handle this in the past? Where did this like maybe give me a quick, quick synopsis of how this account has been going and what the relationship is. Like you're able to now that it has all the data, ask it very specific questions on how this previous person actually handled it. And I think it was, it was inspiring to me just to think about, okay, so there is still a place for people, obviously there's still a need like, but there is an ability to get some lift out of a tool that would be silly for us not to use because all it's going to do is elevate up your people to get like we talked about, best bang for your buck, right? You're, you're paying people, your people are your highest expense likely. You know, I mean you're, that's one of the biggest things that you're outlaying. And I think a lot of entrepreneurs can feel a lot of grief about that. They're like, man, do we need all these people, like all this money we make, we're putting it back out all back into operations based stuff. So how can we make sure whatever you're, you are paying people to do is actually their highest brilliance and they're not just slogging, they're not just wasting time because your people probably don't like slogging, right? That's just a mis, nobody wants a boring slogging job, right? Your people want to be challenged and your people want good tools. And so I think that one of the best things you can do for them is bring in some experts that can help you identify how could you be using, you know, automations, how could you be using AI? How could you be using these resources to make that person's job more enjoyable for them and more valuable to you? I know Faith on the finance team as well was talking to me about a tool that she's developing. You know, we have a lot, we have a coffee client that has vendors that send in a lot of pictures or kids that work for, for this coffee client that send a lot of pictures of handwritten invoices they receive or places they've outlaid money to. Places that are like physical invoices. They take a picture, they'll drop it into a channel. Well then it's been taking somebody manually keying in some different things based upon what they've seen come through for us to make sure the books stay on, on task, on topic. And I know Faith had been playing with, hey, I want you to recognize here like this invoice what's coming in and then identify for me like you know, there's certain line items that we're going to have to pay taxes on that are different based upon the vendor and allowing it to work on the smarts of looking for certain things within each one of these invoices to pull it up, throw it on a spreadsheet that then she can review and it not having to be somebody that we're manually keying through hundreds and hundreds and hundreds and hundreds of pictures to identify what it was that they're looking for specifically. So I think just cool Things that we can still use the people, but let's use them better, let's use them smarter. Tab, what do you feel like are things that, you know, you and I have had a lot of conversations around this every. Because I think everybody's nervous, right? Everybody's nervous about like am I still relevant? Am I still going to be necessary? Like is there a time when AI could just take over everything? You know, from your perspective looking in, what are the things that you feel like? Yeah, I, I don't think at this point at least that you could hand this part of your finance off to AI.
- [16:50]
C
I think that it's funny because one of the things that people I hear the most specifically when it comes to finance is like bookkeepers are just going to go away. Bookkeepers are never like, you know, that's going to be not something that's needed after AI is developed to a certain point. We are definitely not there yet. Anyone who works in like different accounting softwares will know that that's definitely not to the point where you're getting rid of your bookkeeper. But I actually think maybe not at the bookkeeper level, but I think again back to that context and the knowledge of, of like an indivi. I mean obviously we're working with clients so like there's different knowledge base that I'm taking into each client that we're working with and understanding and doing different things and working with the team that's bookkeeping to like, you know, you're thinking from this framework as you're looking at these transactions. I think until we have like kind of the same way that we do in Claude or Chat or wherever you're working where you can kind of set the context of that project and provide instructions for how to treat different things until we're there. You're never going to get good data and ending actionable insight without like a hands on person that understands all that context doing the bookkeeping or at least going back through it to like audit. I think that's the biggest thing I see is review audit, understanding some of the nuances that are occurring within a business that are just like not explained when it like you can follow patterns, you can follow gap, you can code all of that into a software, but it's not necessarily going to capture what a person who understands the goals of the company understands why this time we're treating this random purchase differently than we did last time. I think there's, there's just a lot of different circumstances in which you need oversight and so kind of Like Taryn was saying, where it's like, like it's not that the person is going away, it's that their job is shifting. Your job is not necessarily going to continue to be the doing. It's going to be a lot more of the, the project management, the under, like the review, the did this, did we get things done on time and is it passing the inspection, to use the contractor terminology? It's, it's very, I think it's shift. It's going to shift to some degree to be that, that you are a knowledge base that's reviewing based on the context of that client corre. And hopefully like I love to picture a world in which you can continue like QuickBooks has like an instructable AI built into it. You know, that you can continue to refine and build processes within that that's different for each set of books that you're in. I think that's, that's one way. I also think that reporting and like getting again, it's like you're, you're driving the tool. I think like that's one way. I've seen huge efficiencies and then also just like huge improvements to what I'm producing is how I'm using it with reporting. But it needs the context and it needs the review and it needs pointing out just the different factors that it, it doesn't know at this point. I mean, I think that's the, that's the always the caveat to all the AI conversations is like today, this is where we're at. In two years, who knows? I'm not a mind reader or a future teller, so we'll see where it goes.
- [20:05]
A
I think it's good. I think just the realization of, you know, some things that we call AI now have been existing in the background. We just haven't been calling them that in the past. You know, I think of like QuickBooks rules where people will just import files and allow QuickBooks rules to define things and then categorize things in your QuickBooks. But anybody that's used QuickBooks and has used QuickBooks rules, like, I don't know a lot of people that actually stick to a lot of the rules because the rules can a lot of times start miscategorizing things and you can get into real messy situations where it just runs off and then you're like, oh, do you know how much time it's going to take me to go back and figure out how long that you've been doing this and miscategorizing and having a problem that when I triple over things. So I think obviously as we come into things, just the realization of it is still important to keep really smart people on your team. It is really smart for you to continue to have people continuing to quality control whoever's doing it. If you've cast it down, if you delegate it down to somebody that's not as skilled, that's great to realize there's still going to be a quality control element that that person that understands how things work are not going to get out of my finance perspective. But I think too Tabs, what you said is really good. I really love the way at the end of months as you tie off a set of books where you'll go through and you'll give them an evaluation, you create a video and you're like, hey, this is where your eyes should go on this profit and loss. A lot of business owners don't understand how the profit loss works and they're too embarrassed to ask an accountant because they feel stupid. I love that you invite that conversation and hey, tell me where you're confused. Nobody here is stupid. Like you're so smart. You've made all this money. This isn't your expertise. How can I help you understand this better? So I love how you'll create a two minute loom video. Walking through a, A, P and L and saying, hey, this is what I see as I look at this. This is what I think is important for you to put your eyes on. Let's do a follow up call if you want to have, you know, some conversation around it or just shoot me a quick box message and let's have a conversation back and forth on it. But I just think that element is the human element that I think is so valuable when you get into a jam of hey, we're here to help you understand we're going to use the right tools, we're going to use the right people and then we're here just to give you ease to continue to drive your business forward. So anyway, Aaron Turner, always an amazing thing. I hope rock delivery goes incredible today. Tabby man, thank you for being willing to jump on the call. I love hearing from Tabs, such a smart lady. And Ops Experts, I hope you enjoyed the show today and we'll see you back here next week on Ops Experts Club.
- [22:14]
C
Sounds good.