SaaS Scaling Secrets · 2026-04-07 · 41 min
Key moments - from our scoring
Substance score
45 / 100
Five dimensions, 20 points each
Productive is an AI-enabled platform serving professional services firms - agencies, consulting companies, and similar businesses - helping them manage task tracking, time logging, invoicing, and resource planning in one unified system. Tom Carr shares how the company approached adding AI capabilities by starting with obvious pain points like report building, where customers waste significant time on manual configuration and compilation. Rather than asking customers directly "what do you want with AI," his team frames conversations around business bottlenecks and automation opportunities, then determines how AI can address those specific needs.
The conversation centers on a fundamental shift in software development philosophy. Building AI-native features operates under completely different rules than traditional software: inputs are non-deterministic (users type anything in any format), and outputs from language models vary unpredictably. This requires abandoning conventional automated testing in favor of evals - specialized testing frameworks designed for LLM systems. Tom discusses how the rapid evolution of models (from one year ago to now) has changed their approach: newer, more capable models require less elaborate prompting and scaffolding, allowing for better orchestration rather than heavy instruction-loading. This forces teams to rethink product management, engineering practices, and QA - a mindset shift across the organization that isn't yet well-established as best practice.
AI software involves non-determinism on both the input side (users can type anything in various formats) and output side (LLM responses vary unpredictably), whereas traditional software uses controlled form inputs and deterministic outputs that can be tested with automated testing. This requires completely different mindset and approaches for product, engineering, and design.
Evals are automated tests specifically designed for LLM systems to verify that AI features actually work correctly, replacing traditional automated testing approaches. They've become a new core skill for product managers and engineers building AI-enabled applications.
Instead of asking customers what they want to do with AI, Productive focuses research on identifying business bottlenecks and time-wasting processes, then determines how to apply AI to solve those specific problems. Their first AI feature was AI-powered report building because everyone hates manually configuring reports.
Newer, more capable language models require less elaborate prompting and instruction-loading than models from a year or two ago. This has shifted their approach from heavy instruction-based systems to better orchestration, where models are given tools and context rather than detailed step-by-step instructions.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode has occasional useful tactical observations - particularly about non-determinism on both input and output sides, and the nudge-based AI adoption pattern - but much of the runtime is consumed by meandering host monologues, platitudes about 'the playbook changing,' and generic AI commentary a thoughtful operator would already know. The ratio of genuinely new ideas to filler is low.
what we found that works is just plugging various functionality across the traditional interface that feeds into the chat bot
The biggest problem today with building AI software is the non-deterministic nature of it
The episode leans heavily on circulating AI-era clichés - roles blurring, non-determinism, 'think from first principles,' 'the whole playbook is changing' - without adding a distinctive angle or counterintuitive argument. The MySQL analogy for LLM commoditisation is competent but not fresh, and the billboard advice is pure boilerplate.
The entire playbook I think is changing right now. I don't think a lot of what was worked before in terms of SaaS is gonna work in the future.
I don't know, 20 years ago you would say, yeah, we're all using the same databases, we're all using MySQL, we are all using the same Ruby rails or whatever you're using
Tom Carr is a genuine bootstrapped operator who has built a real multi-continent SaaS business at 150 people with 40% YoY growth and a broader portfolio, and he leads product himself - giving him authentic practitioner credibility. However, the scale is modest and he has limited exposure to the truly hard problems that emerge at hundreds of millions in ARR or thousands of employees.
Tom is a Bootstrap founder from Croatia who's built productive into a global company growing 40% year over year with 150 plus people and customers across six continents
I don't have a VP of product. 'cause that's like my dual role
The episode offers a handful of concrete data points (40% YoY growth, 150 staff, six continents) and one clear product-level example (the AI report-builder button opening a pre-seeded chatbot prompt), but almost no customer outcome metrics, ARR figures, or hard evidence on AI feature adoption rates. Most claims remain illustrative rather than evidential.
you go to our report builder, the traditional one, and there's like a button air reports, and the only thing that button does is it opens up the chat bot with a pre-selected prompt
growing 40% year over year with 150 plus people and customers across six continents
The host asks topically relevant questions and occasionally surfaces useful tensions (pricing vs. adoption, competitive differentiation), but routinely derails with multi-clause self-editorialising that eats guest time, fails to press on vague 'first principles' proclamations, and lets platitudes pass unchallenged. Follow-ups exist but rarely push into uncomfortable or falsifiable territory.
I was listening to a good conversation yesterday and someone made a very astute point of like, people often talk about these, system of record versus system of, workflow or systems of profit. Processes.
Has that changed organizationally, like responsibilities inside the company?
Computed from the transcript - who did the talking, and the words that came up most.
Dan Balcauski interviews Tom Car, founder and CEO of Productive, an AI-enabled SaaS platform for agencies and professional services that unifies task management, time tracking, invoicing, and resource planning to improve productivity and profitability. Tom explains how Productive chose early AI features by focusing on customer bottlenecks rather than asking what customers want from AI, starting with AI-assisted report building. He contrasts traditional software with AI’s non-deterministic inputs and outputs and emphasizes the need for evals, evolving engineering practices, and shifting roles across product, design, and engineering. Car discusses integrating AI into existing tiers to avoid stifling adoption, using in-product nudges to build habits, and considering future usage-based pricing for higher-cost AI workloads. He notes AI affects sales as a “future-proof” signal, while some customers remain skeptical or treat AI as magic, and he advises SaaS leaders to rethink playbooks from first principles as the industry changes.
Transcribed and scored by The B2B Podcast Index.
1 - > tom_1_03-10-2026_171009: The biggest problem today with 2 - > building AI software is the non-deterministic nature of it, 3 - > and there's a joke that like every product manager now thinks 4 - > that are an engineer, and every designer thinks every product, 5 - > everybody's now everything. 6 - > it's, the rules have gotten close to each other. 7 - > And then we would ask them, well, what would you like? 8 - > And they're like, well, we don't know.
9 - > You tell us. 10 - > They know the technology is out there and they want somebody 11 - > that's they can bet on it's gonna execute in the future and 12 - > not gonna become a legacy player in a sense. 13 - > But then the other thing also that's very interesting for me 14 - > is that sometimes people consider AI basically magic, 15 - > which is interesting. 16 - > The entire playbook I think is changing right now.
17 - > I don't think a lot of what was worked before in terms of SaaS 18 - > is gonna work in the future. 19 - > So you need to think everything you know, from grounds up from 20 - > first principles. 21 - > dan-balcauski_1_03-10-2: Welcome to SaaS Scaling Secrets, the 22 - > podcast that brings you the inside stories from the leaders 23 - > of the best scale up. 24 - > B2B SaaS companies.
25 - > I'm your host, Dan Balcauski, founder of Product Tranquility. 26 - > Today I'm excited to welcome Tom Carr, founder and CEO of 27 - > productive an AI enabled SaaS platform that helps professional 28 - > services teams improve productivity and profitability. 29 - > Tom is a Bootstrap founder from Croatia who's built productive 30 - > into a global company growing 40% year over year with 150 plus 31 - > people and customers across six continents beyond productive. 32 - > He's founded, invested in multiple tech companies, 33 - > including Infinum and joint venture with Porsche Digital, 34 - > with more than 600 people working across his portfolio 35 - > today.
36 - > Let's dive in. 37 - > Tom, welcome to the show. 38 - > tom_1_03-10-2026_171009: Hi. 39 - > Thanks for having me.
40 - > dan-balcauski_1_03-10-2026_: I'm very excited for our 41 - > conversation. 42 - > Before we dive into your scaling journey, give us the elevator 43 - > pitch. 44 - > What does productive do? 45 - > Who do you serve?
46 - > tom_1_03-10-2026_171009: So productive is we mostly serve 47 - > agencies, consulting companies, and similar companies that do 48 - > what you would call professional services. 49 - > We offer them like a one-stop shop so they can do like task 50 - > management, time tracking invoicing, resource planning, 51 - > all of that in one place. 52 - > And that's basically how they can save, like they don't have 53 - > to run multiple tools, multiple different processes. 54 - > and basically there are two main aspects that we offer.
55 - > One is like productivity, so making them productive, that's 56 - > where the name came from, and also profitability so they can 57 - > be more profitable because a big part of our platform is actually 58 - > financial aspect of everything that happens with the client 59 - > work. 60 - > dan-balcauski_1_03-10-2026_1: So helping professional services 61 - > manage the work that they're doing day to day, and then also 62 - > making sure that they're making money 63 - > tom_1_03-10-2026_17100: exactly.
64 - > So both those things juggling because in the past you would 65 - > typically have multiple tools doing that, and they didn't talk 66 - > to each other. 67 - > And that's where we came in. 68 - > dan-balcauski_1_03-1: Fantastic. 69 - > Well, I think there's a same problem on a lot of software 70 - > companies as well, where your costs cost and revenue side is 71 - > completely separate from the technology side.
72 - > But so you explicitly focused at professional services. 73 - > I think, in terms of the world of technology every. 74 - > Software company maybe every company is racing to add AI 75 - > features and capabilities right now. 76 - > When you, I'm assuming product if productive is not any 77 - > different from that.
78 - > When you started thinking about this AI world and where it fits 79 - > with productive, how did you decide what to build first? 80 - > tom_1_03-10-2026_171009: So I think the first thing we, I mean 81 - > we obviously looked at the capabilities, like what can we 82 - > do with the technology actually, and then what are the obvious 83 - > low hanging stuff that we can just. 84 - > Dip part those in and see if it works, right? 85 - > So it's not just AI that we just added there.
86 - > We wanted to find something that people would really use. 87 - > The first thing we did basically was just report building. 88 - > 'cause everybody hates, building reports, configuring them, all 89 - > that part. 90 - > And it seemed like an obvious thing, just type something into 91 - > a prompt and you get like a really nice report before, but 92 - > you had to click around and basically know how to use the 93 - > ui.
94 - > And we got the idea obviously from our customers and like all 95 - > the stuff that we do now with AI is coming from our customers. 96 - > But it a different way in a sense that like if you ask 97 - > customers, what do you wanna do with ai? 98 - > They don't really know how to answer that question. 99 - > It's not a straightforward question.
100 - > It's more like, what do you wanna automate? 101 - > What's where are the bottlenecks in your business? 102 - > What are you wasting a ton of time on that, could be faster? 103 - > And then we try to figure out how to use AI to improve that.
104 - > dan-balcauski_1_03-10-202: Yeah. 105 - > I find that when you talk to customers about. 106 - > Out, those type of bottlenecks as well. 107 - > They tend to, obviously depends upon the skill of the researcher 108 - > asking this kind of question, but they also try, tend to think 109 - > within the frame of the solution that you offer.
110 - > So I'm curious, like, how have you thought about that or 111 - > approached it as you've interacted with customers to try 112 - > to say like, okay, we. 113 - > We have a sense of what AI as technologists can do, and then 114 - > we're going and talking to our customers about, what they're 115 - > trying to do with their business. 116 - > How have you navigated? 117 - > Okay, like use an example of like, well we used it to help 118 - > with reporting, right?
119 - > Because currently that's very painful, but that's painful 120 - > within the scope of your solution, versus thinking about 121 - > how do we take this capabilities and really. 122 - > Expand the scope of the problems that we solve for customers. 123 - > How do have you navigated that and like I'm especially curious 124 - > like if that's been a challenge as you've tried to manage your 125 - > internal team as well on shifting that mindset versus 126 - > just trying to take, hey, what are the friction points within 127 - > our existing product and how do we think broader?
128 - > tom_1_03-10-2026_171009: Yeah, I mean there are two, two points 129 - > there that you made, and I think they're really good. 130 - > One is what we do with the product and how do we figure out 131 - > what the customers want and how do we build it and what's the 132 - > best strategy there. 133 - > And then the other is just the internal thing, team thing. 134 - > I'll just, I'll focus on the first one, which is what did we 135 - > do inside a product?
136 - > And you mentioned you researchers, and I think that's 137 - > very important. 138 - > When you're like a smaller company, which we were as we 139 - > were growing, we didn't have any researchers. 140 - > So at the beginning it was hunch driven. 141 - > Then you might hire product managers and then they're like 142 - > moonlighting as researchers.
143 - > And at some point we did hire reaches researchers and that 144 - > kind of really changed the way that we do this stuff. 145 - > It's always like research first, it's always talking to your 146 - > customers, your users, which again, it is a problem when 147 - > you're a smaller company'cause you don't have any users to talk 148 - > to. 149 - > So it's like very, few and far between and it's hard to 150 - > generalize it. 151 - > With our larger customer base, we can obviously generalize this 152 - > stuff a little bit better.
153 - > That can always be better. 154 - > But yeah, we didn't try to approach them with the AI angle. 155 - > With all of those research sessions we're more around, 156 - > again, what do you want to improve in your business? 157 - > And yeah, you're right, it does boil down back to our product, 158 - > but that's also good for us because.
159 - > don't wanna go too far. 160 - > We want to improve what we can with the data we have with the 161 - > product. 162 - > We have that was at least when we started doing this, maybe a 163 - > year and a half ago, today. 164 - > We can expand more on this stuff.
165 - > But yeah, then the customers are just, you start, you get them 166 - > talking and they're like, yeah, I can do this. 167 - > I spend a lot of time on, I know compiling reports. 168 - > Not like data reports, but project updates, all of this 169 - > stuff, and I need to build tasks from notes from a meeting, blah, 170 - > blah, blah. 171 - > All these like typical things.
172 - > And then we sit down and see what's, what are the common 173 - > themes or the most important themes, and just get into it and 174 - > figure out how to solve it. 175 - > dan-balcauski_1_03-10-2026_1: So once you've. 176 - > And if it helps to talk about a specific feature or even keep it 177 - > at the general level. 178 - > One thing I'm interested in is that, as companies have started 179 - > to add AI enabled tech, feature sets into their products, it's 180 - > one thing to say, okay, we're gonna go build this, and then 181 - > it's another to actually go and do it with the way that a lot of 182 - > these AI systems interact on the backend.
183 - > How have you seen the difference between, building the, with 184 - > these AI enabled capabilities versus maybe what we might refer 185 - > to as traditional software? 186 - > tom_1_03-10-2026_171009: Yeah. 187 - > Yeah. 188 - > I think that's a good thing.
189 - > Traditional software. 190 - > The biggest problem today with building AI software is the 191 - > non-deterministic nature of it, and it's on two sides. 192 - > You have the input is non-deterministic. 193 - > 'cause typically it's like a chat box, it's a text box where 194 - > people type in whatever they feel like it in whatever 195 - > language.
196 - > And sometimes, we see what people type into and it's like 197 - > somebody types in three words. 198 - > Somebody types in like a paragraph and it has to work the 199 - > same for both of those things. 200 - > And then it goes through a system and it, there's an l lm 201 - > in that system that spits out an output. 202 - > And then the output is also non-deterministic.
203 - > You, it's not always the same. 204 - > So you have done determinism on two sides and if if you're used 205 - > to, building typical, traditional software where kind 206 - > of the input is a form and a button and the output you can 207 - > test with automated testing, this completely different. 208 - > There that's a mindset switch both for kind of the product 209 - > side of the business, product management, but also for the 210 - > engineering side of it and the design.
211 - > And it's all becoming different. 212 - > And it's also very new. 213 - > So there's not so many established practices that you 214 - > can copy paste from. 215 - > You gotta figure this, a lot of this stuff out yourself.
216 - > It's very, it's changing a lot. 217 - > The biggest kind of thing here is evals. 218 - > For everyone who's not familiar with evals, it's automated 219 - > tests, but more geared towards LLMs and, AI systems. 220 - > And it's a different way of testing if your thing actually 221 - > works.
222 - > And figuring out how to build that and get source the data 223 - > from it. 224 - > It's like a new skill, I would say in product management and in 225 - > building these ai, AI enabled applications or AI native 226 - > applications. 227 - > dan-balcauski_1_03-10-2026_1: So as your team. 228 - > The business went through adding these AI capabilities, I guess 229 - > what surprised you in that process?
230 - > I think evals are starting to percolate a little bit more, 231 - > but, I imagine if we went back even 12 months ago, if I'd asked 232 - > anybody what an eval was, they would have no idea. 233 - > So I'm curious like what surprised you, like as you and 234 - > the teams, like, Hey, we got this great idea, we're gonna go 235 - > do it, and. 236 - > Maybe the CTO like, spends a weekend and built a really 237 - > impressive demo, but then that's different from getting set that 238 - > into production.
239 - > What were the areas that, either were surprised or were 240 - > bottlenecks as you guys made this transition to, to build 241 - > these products. 242 - > tom_1_03-10-2026_171009: Yeah, when you start, you do just like 243 - > a vi check. 244 - > Type of thing. 245 - > It's not you.
246 - > You try it out, if it works and you go, oh yeah, this works. 247 - > And then you give it to customers and they just surprise 248 - > you with what, whatever they put in that, from that 249 - > non-deterministic side of the product. 250 - > and then you have to go back and figure out some, a more 251 - > structured approach. 252 - > So that's one thing that surprises also, some of the 253 - > stuff that worked that we built out, I would say a year ago.
254 - > We don't need right now because the way the kind of, the models 255 - > have changed and related to that, or we're doing it a 256 - > different way now like a lot of these like models are very smart 257 - > now and you don't have to instruct them so much more as in 258 - > give them tools. 259 - > So it's an orchestrated in, in, in a better way. 260 - > And you can see that a lot with coding right now. 261 - > So what happens with for example, cloud code and 262 - > everything that happening there, it's more around the 263 - > orchestration and just giving the actual LLM everything it 264 - > needs so that it can, code and not giving so many instructions, 265 - > which was different like a year ago or two years ago.
266 - > You had to build these. 267 - > elaborate prompts and dictionaries and everything, and 268 - > right now it's a little bit different. 269 - > So you know, things are changing, but that's been the 270 - > biggest surprise for me. 271 - > That's something that we built a year ago was, we're doing 272 - > completely differently now.
273 - > dan-balcauski_1_03-10-2026_1: So I'm curious kind of maybe tying 274 - > a couple of things together. 275 - > 'cause I think maybe what I just heard was like people have 276 - > referred to this differently of like, scaffolding or guardrails, 277 - > right? 278 - > Some of these models had very poor, recall or they would go 279 - > off on, in tangents areas, right? 280 - > There are a lot of funny jailbreaks of people, tricking 281 - > ai service chatbots into, giving them a bunch of, money that was 282 - > it, possible or you know, give me a recipe for a tuna salad, 283 - > uh, whether you're chatting with your Salesforce support agent.
284 - > So, you know, the models themselves have got more 285 - > capable. 286 - > So I imagine, you know, that's changed a little bit in 287 - > engineering. 288 - > But then you also talked about, you know, the evals. 289 - > Has that changed?
290 - > I guess either one of those, has that changed organizationally, 291 - > like responsibilities inside the company? 292 - > Um, meaning like, you know, in traditional sort of product 293 - > design engineering, right? 294 - > You've got a spec and sort of the engineers build to that spec 295 - > and they test against that spec. 296 - > But something like an eval.
297 - > Starts to get a very, starts to get squishy, right? 298 - > You have a little bit more because you have so many 299 - > different other dimensions, you know, does that creep into, oh, 300 - > well we actually need the designer or the product manager 301 - > or the customer success people to actually be running these 302 - > evals. 303 - > Like how has that affected the organization in, in, or in your 304 - > view of who is responsible and who needs to have in insight 305 - > into that.
306 - > tom_1_03-10-2026_171009: We're still also trying to figure out 307 - > the best way to move forward. 308 - > And you typically have these three roles. 309 - > You have the engineers, designers, and product managers. 310 - > And right now these roles are changing a lot.
311 - > And there's a joke that like every product manager now thinks 312 - > that are an engineer, and every designer thinks every product, 313 - > everybody's now everything. 314 - > it's, the rules have gotten close to each other. 315 - > I don't think they're still or they're ever gonna be. 316 - > Completely merg into one thing because just some people are 317 - > better at certain things.
318 - > But you as a product manager need to be closer to the code, 319 - > the technology. 320 - > You as an engineer need to be closer to the user, so meaning 321 - > design because as an engineer before maybe you didn't need to 322 - > be so close to what the customers are saying, but not 323 - > what the customers are saying. 324 - > They're saying in a. 325 - > bot chat box that's going directly to you as an engineer 326 - > to figure out how to create an interface out of.
327 - > So it it, it has changed a little bit and we are trying to 328 - > also figure out how to do it properly. 329 - > The whole eval thing, it's mostly around product management 330 - > and it's mostly around engineering, I would say. 331 - > Not so much design, at least for us, I think. 332 - > But the other thing there is, that's interesting to me is 333 - > figuring out to what.
334 - > Degree, your product is a chat box. 335 - > And to what degree it's a application. 336 - > And where do those things blend and how do they work together? 337 - > Because a lot of products today, you take anything like lovable, 338 - > it's here's the interface you're building, here's the chat bot.
339 - > What is your, what is that for your product? 340 - > And doesn't make sense in all situations. 341 - > In certain situations. 342 - > It's it's very interesting and I don't think it's, the same.
343 - > So we're also trying to figure that out. 344 - > dan-balcauski_1_03-10-2026_11: I think that, even over the last, 345 - > two years, right? 346 - > I think when maybe GPT-4 was released, I think the first 347 - > round of AI enabled applications, everyone just 348 - > added a little chatbot to their product. 349 - > It was like, okay, now we have ai.
350 - > But the problem there is that it's like. 351 - > One, it's a blank slate. 352 - > And so anytime, if you give a blank slate to a very technical 353 - > power user, they're like, this is awesome. 354 - > I can do everything here.
355 - > If you give it to, quote unquote normal user, they're like, I 356 - > don't even know what to type here. 357 - > But then even if you look at something as simple as, you 358 - > know, a lot of these, even the. 359 - > Foundation model companies open ai, philanthropic they can now 360 - > create, documents in a workflow. 361 - > Well, okay, now I have a chat and a document, right?
362 - > I mean, you mentioned lovable, right? 363 - > Which is, okay, now I've chat and a user interface. 364 - > Well, what do I change directly in the document versus what do I 365 - > chat about changing in the document, right? 366 - > So even something as easy as like document management just 367 - > in.
368 - > Implies a whole new paradigm of interaction that we haven't had 369 - > to deal with before. 370 - > And so I think a lot of companies, I don't think 371 - > anybody's figured it out yet. 372 - > I've seen some interesting ideas, but I don't think, uh, 373 - > the industry as a whole has aligned on a pattern. 374 - > One thing I wanted to ask about was that, you know, looking at 375 - > your, I was looking at your, uh, your pricing notice.
376 - > You know, you guys have included AI across, all your tiers versus 377 - > having that as a, premium, you know, paywall or add-on. 378 - > I'm curious, how did you arrive at that approach? 379 - > I think there's a discussion going on across, boardrooms in 380 - > every, tech company being like, well, these. 381 - > Large language models, right?
382 - > We could have a large bill going out the door to open AI or 383 - > Google Gemini every month. 384 - > Like, we, we want to, charge customers, more for this because 385 - > it cost us more. 386 - > How did you guys end up at this approach where it's embedded in 387 - > your product across the tiers? 388 - > tom_1_03-10-2026_171009: Yeah, so I think that there are a 389 - > couple of questions there.
390 - > One is do you have a separate tier for your AI functionality? 391 - > And right now for us, that is merged into the basic product 392 - > line. 393 - > 'cause as we were building out the AI functionality made sense. 394 - > But the other thing is more about the credits.
395 - > So usage based pricing, meaning the more you use the more the 396 - > customers pay. 397 - > And that's what all the large language models you mentioned 398 - > the foundation labs are doing. 399 - > So you're paying it, either, it's either through a bundle or 400 - > something, but at the end of the day, you are paying for the. 401 - > Tokens or credits or whatever you call it.
402 - > And that's like also our direction going forward. 403 - > So, we are gonna offer for those functionality that's not very 404 - > cost prohibitive for us, we'll just wrap it up probably in the 405 - > main product. 406 - > But anything that's higher throughput, that kind of 407 - > requires more AI usage I think the only sensible way is to do 408 - > some sort of credits pricing, usage based pricing or outcome 409 - > based pricing. 410 - > If you can do.
411 - > It. 412 - > which are all I think, interesting. 413 - > Some are easier to do in some industries than others. 414 - > dan-balcauski_1_03-10-2026_1: As and I apologize because I maybe 415 - > skipped over this.
416 - > Is the is that sort of credit bundle offer like on in, in your 417 - > price, existing pricing? 418 - > Today? 419 - > tom_1_03-10-2026_171009: No. 420 - > Not the public one, but yeah.
421 - > dan-balcauski_1_03-10-2026_: No, not. 422 - > tom_1_03-10-2026_171009: up. 423 - > dan-balcauski_1_03-10-2026: It's coming. 424 - > Okay.
425 - > So I, my research was correct but we're talking futures. 426 - > As you've gone into it sounds like there was, you're seeing a 427 - > potential change of even more advanced capabilities that are, 428 - > potentially going to drive more costs. 429 - > There's more or tokens generated. 430 - > tom_1_03-10-2026_171009: And also, 431 - > dan-balcauski_1_03-10-: curious.
432 - > tom_1_03-10-2026_171009: value created through that process. 433 - > It's if you're just generating tokens, value, nobody's gonna 434 - > pay for them. 435 - > So that's where we see opportunities. 436 - > dan-balcauski_1_03-10-202: Yeah.
437 - > I'm curious, like going back to like even the stuff that you 438 - > launched originally though, right? 439 - > You know, because I think what's interesting about, you know, 440 - > your approach is that wasn't, it sounds like you're going to the 441 - > go there for at least some features, but I think out of the 442 - > gate, the majority of companies that I. 443 - > C I spent all my time in the pricing and packaging world, so 444 - > this is a pretty thing I pay attention to a lot.
445 - > And so I think the initial sort of gut reaction of a lot of 446 - > companies is like, well, this is gonna cost us tokens. 447 - > We're going to charge it as an add-on. 448 - > I'm curious, what was your thought process around not, 449 - > around the bundling o option, like upfront? 450 - > tom_1_03-10-2026_171009: The thought process mainly was we 451 - > don't wanna, stifle in the beginning.
452 - > We don't wanna stifle usage on, because it's always like you 453 - > have the U and monetization and those are two levers. 454 - > If you pull on too hard on monetization, usage is gonna 455 - > drop off. 456 - > Especially in the beginning when you're building a new feature or 457 - > a new product you can afford it. 458 - > I don't think you should over monetize because you want to see 459 - > what people are gonna do with it.
460 - > It's probably also. 461 - > It has bugs and maybe it doesn't work. 462 - > A hundred percent you want user feedback. 463 - > And a good way to do that is to not over monetize.
464 - > And that was basically our strategy. 465 - > And now that things are progressing there, we're gonna 466 - > work out some monetization strategies that are a little 467 - > bit. 468 - > dan-balcauski_1_03-10-202: Yeah, that's I was pointing it out 469 - > because I think it is unique and, there's always a risk of a. 470 - > Any new product innovation, right?
471 - > You put out a new feature, no matter how much, how many, 472 - > customer conversations you have, you build something, you think 473 - > it's gonna be awesome. 474 - > And then, Pareto's principle pers law, like Rule Supreme, 475 - > right? 476 - > It's like, the majority of your feature usage comes from 20% of 477 - > the features and a bunch of thought stuff you thought was 478 - > gonna be, life changing. 479 - > So you always have that risk.
480 - > And I think what you pointed out really well is like as soon as 481 - > you put up a monetization gate. 482 - > You, you increase that risk because like, we didn't get 483 - > usage, just because of the, every feature that we introduce 484 - > has a risk of not being adopted. 485 - > And now we've also made that hurdle even higher for our 486 - > customers to, to get it. 487 - > I'm curious, like, as you have gone down that path I'm 488 - > wondering if you could talk a little bit too, what were the 489 - > things that you were able to learn with that approach?
490 - > Because I think a lot of companies are throwing AI 491 - > features out there and are, just like, I don't know, like, is 492 - > somebody using them? 493 - > Yes or no? 494 - > Like, as you were able to increase that, flywheel of usage 495 - > and adoption, what were the things that you learned that you 496 - > might not have otherwise seen if you had, more gated 497 - > monetization? 498 - > I.
499 - > tom_1_03-10-2026_171009: I think one of the main things I would 500 - > say in general with AI usage across, let's say a product 501 - > that's implementing AI functionality is you talked 502 - > about it before, is like you give people a chat bot and they 503 - > don't know what to do with it. 504 - > So what we found that works is just plugging various 505 - > functionality across the traditional interface that feeds 506 - > into the chat bot. 507 - > So I'll give you, I'll go back to our example of a report 508 - > building.
509 - > So you go to our report builder, the traditional one, and there's 510 - > like a button air reports, and the only thing that button does 511 - > is it opens up the chat bot with a pre-selected prompt, but then. 512 - > People, oh yeah, I could do this in the chat. 513 - > But if we didn't have that there, people wouldn't be like, 514 - > oh yeah, let's open up the chat bot and ask you to build a 515 - > report, because the habit is still not there for most most 516 - > customers, most users.
517 - > So these like little habit building situations all over the 518 - > product, which is what we are doing actually have a really big 519 - > impact on the adoption of the whole thing. 520 - > dan-balcauski_1_03-10-202: Yeah, so there's little, uh, yeah. 521 - > Uh, habit adoption hooks. 522 - > Yeah, nudges.
523 - > Which yeah if, each of those has a, somewhat percent chance of 524 - > being engaged. 525 - > And again, if you have just a smaller pool of people who are 526 - > paying you for that I remember, you know, uh, personal, my 527 - > personal experience, my, my CRM that I use, you know, at some 528 - > point, two years ago, uh, for the plan I was on, they 529 - > introduced, you know, summarize the account history with this 530 - > customer option. 531 - > And they gave like three uses a month for my plan that I was on.
532 - > Right. 533 - > I was on symmetry level plan. 534 - > And I say this to say that I completely align and empathize 535 - > with that point of view you just said, because, you know, it's 536 - > like once I hit that three usage. 537 - > It was like, okay, that was interesting, but it didn't 538 - > become, it wasn't like enough for it to be like, oh, now this 539 - > is taken away from me, and like, I need this for my business, so 540 - > please let me pay you the extra to like, upgrade.
541 - > So I think there's always that that it, there's a lot of like 542 - > art and science to that of like, how much do you give in a 543 - > particular plan before you let someone, upgrade. 544 - > Right. 545 - > And so, that was 546 - > tom_1_03-10-2026_171009: But 547 - > dan-balcauski_1_03-10-202: good. 548 - > tom_1_03-10-2026_171009: yeah, but like our strategy from 549 - > beginning implementing AI was not to make money off of it, 550 - > because we still think it's very early.
551 - > So we want to build a good product, the best product, and 552 - > we want people to use it. 553 - > And then we'll figure out how to make money. 554 - > Kicking out monetization straight through the door I 555 - > think wouldn't be good. 556 - > So like we, we don't, we wouldn't wanna do that.
557 - > Like charge for three uses of a thing that probably doesn't cost 558 - > you too much. 559 - > dan-balcauski_1_03-10-202: Yeah. 560 - > Well, I think a lot of folks are. 561 - > Finding that, these ais, right?
562 - > We run the risk of, uh, too much anthrop authorization or 563 - > treating these ais like a human, but to a certain extent you 564 - > know, a lot of these models have, they become smarter. 565 - > They're now at the level of maybe a junior employee and you 566 - > don't give the junior employee a new, Hey, re-architect our 567 - > software like on your first week, right? 568 - > You build up that trust and capability over time. 569 - > And so it's almost like you need to.
570 - > tom_1_03-10-2026_171009: And you train it, and you figure out the 571 - > various skills and tools and everything. 572 - > Yeah, exactly. 573 - > Exactly. 574 - > dan-balcauski_1_03-10-2026_: And so, you know, for users to adopt 575 - > it, right?
576 - > I think it requires, you know, just for their perceived value 577 - > of that. 578 - > You know, before you say, okay, well now it's time to charge. 579 - > I think that's what I don't wanna gloss over what you said, 580 - > but I think was really important because it sounded like you had 581 - > a defined business objective in mind, which is, it's early. 582 - > We want to, drive value, learn.
583 - > Our goal is not maximizing monetization outta the gate, 584 - > tom_1_03-10-2026_17: absolutely. 585 - > dan-balcauski_1_03-10-2026_: I'm double clicking on that because 586 - > I think that's so important, and it's an area that folks don't 587 - > really spend the time in the executive room discussing what 588 - > is your goal for this, right? 589 - > Because it's very easy for the CFO to say. 590 - > Hey, we got this big bill from OpenAI this month.
591 - > Uh, we need to not, I need to make sure my margins don't 592 - > change. 593 - > Uh, and then for product to react to that versus having that 594 - > discussion at the executive team level, even at the board level, 595 - > saying like, look for the next year or two years, like, we're 596 - > gonna treat this as an investment to learn, to figure 597 - > out what is actually valuable. 598 - > And then, align, you know, longer term monetization goals 599 - > after that. 600 - > So I think that's really smart.
601 - > How has you ob you're in a very competitive space, so I'm 602 - > curious, like, as it pertains to these applications of artificial 603 - > intelligence within the platform how have you viewed that through 604 - > like a competitive lens? 605 - > Because, you know, as I mentioned before, right, like 606 - > everyone maybe ran out and added a chat window to their ai and 607 - > then. 608 - > I think what I saw across the board is everyone's like, we're 609 - > now AI enabled.
610 - > Well, but so all your competitors claim the same 611 - > thing. 612 - > How have you thought about it from do you feel that like. 613 - > These AI capabilities are commoditizing. 614 - > 'cause look, the next person next to you could use open AI or 615 - > aros, most latest model.
616 - > How have you thought about it through like a differentiation 617 - > lens to avoids like, yeah, okay. 618 - > We've now added additional cost profile, but you know, we're no 619 - > better off than our competitors in this landscape. 620 - > tom_1_03-10-2026_171009: You have your traditional software 621 - > business and you have your traditional competitors, and 622 - > like you said. 623 - > Some of them added the chat box, some of them did the chat box.
624 - > Some are, more into into the, down the AI rabbit hole and 625 - > summer life. 626 - > But I think, what I at least look at it and what our, I guess 627 - > what our strategy here is we are building these features and we 628 - > are using all of the same lms, but. 629 - > The way that you work with the lms, the data that you have and 630 - > the way you structure'em, I think that's where the gains 631 - > basically are. 632 - > So it's again, it boils down to like a product management and a 633 - > strategy decision.
634 - > It's I don't know, 20 years ago you would say, yeah, we're all 635 - > using the same databases, we're all using MySQL, we are all 636 - > using the same Ruby rails or whatever you're using. 637 - > like the products are gonna be the same, but they're not, it's 638 - > depends on who's building it and what your ideas and what your 639 - > strategy and what. 640 - > What your vision is. 641 - > obviously, in competitive spaces, people always look what, 642 - > look at what everybody else is doing.
643 - > Which sometimes I think is good. 644 - > Sometimes I think it's bad'cause then everybody just starts doing 645 - > the same thing. 646 - > But it is the same game, but with different tools. 647 - > The way that's, at least the way I see it.
648 - > I think that the other question is more around the categories of 649 - > software. 650 - > So are they gonna change? 651 - > For example, you typically had like project management, 652 - > software, resource management, software, CRM, et cetera. 653 - > So we are in that space.
654 - > We are project management, resource management, CRM bundled 655 - > it into one thing. 656 - > What's gonna happen there, like if, like way that people work 657 - > with this software changes and that's like a question what's 658 - > gonna happen with. 659 - > Our customers and the way, their businesses evolve with ai. 660 - > So those categories might change and then that's an interesting, 661 - > I would say, thing to see.
662 - > So are people gonna use, I don't know, Chad, GPT for some stuff? 663 - > So how does that impact us and other products? 664 - > is CHE PT like a gateway thing? 665 - > So they start using it, okay, now I need to connect this to an 666 - > actual database, an actual system.
667 - > Et cetera, et cetera. 668 - > So that's what we're trying to figure out. 669 - > dan-balcauski_1_03-10-2026: Yeah I gave you a little bit of an 670 - > impossible question because I don't know exactly when this 671 - > will come out, but if folks are paying attention to the, what's 672 - > going on in the stock market right now there's many terms for 673 - > it, SaaS apocalypse or the AI tidal wave et cetera. 674 - > And so I think a lot of even, very large public companies are 675 - > getting, revalued of like, what does this all mean for your, 676 - > long-term defensibility.
677 - > And I haven't heard anybody give. 678 - > At least anything that's a hundred percent of a convincing 679 - > answer. 680 - > I think it, it really depends upon very different types of 681 - > software and applications and specific focus to customers. 682 - > I was listening to a good conversation yesterday and 683 - > someone made a very astute point of like, people often talk about 684 - > these, system of record versus system of, workflow or systems 685 - > of profit.
686 - > Processes. 687 - > Because, it's like databases are systems of record, right? 688 - > We haven't really been thinking about those forever. 689 - > It's, when we think about I'll use Salesforce as an example 690 - > love it or hate it, right?
691 - > Salesforce has an advantage because there's a good chance if 692 - > you hire a sales rep. 693 - > They probably know how to work on within Salesforce on the 694 - > first day, right? 695 - > And maybe there's some customizations for your company, 696 - > but and then if there are customizations that defines 697 - > specific workflows for your, for your company. 698 - > And so I, I don't think that companies are gonna come in and, 699 - > vibe code their own version of Salesforce.
700 - > But, there probably is another layer of, challenge to their 701 - > competitive differentiation because, if they come out with a 702 - > new feature, you know, that's super AI enabled, if. 703 - > The hype is to all be believed if, even if you couldn't do it 704 - > now, you know, the software development cycles should be 705 - > getting shorter. 706 - > And so the ability for, uh, their competitors to, to catch 707 - > up, uh, maybe, uh, decreasing over time as well.
708 - > And so, uh, you potentially run into this, uh, race where, you 709 - > know, consumers are the net beneficiaries or consumers, 710 - > either business or, uh, regular. 711 - > Uh, non-business, uh, customers of software because you're gonna 712 - > get more and more value of that differentiation is not gonna 713 - > hold up over time. 714 - > I'm curious how, if at all, has the introduction of these AI 715 - > capabilities changed the conversations that your sales 716 - > teams are having with customers?
717 - > tom_1_03-10-2026_171009: That's really across the board. 718 - > Like some people really expect it and it's a way that they 719 - > gauge your platform as being like a serious future proof 720 - > player. 721 - > It's what are you doing with ai? 722 - > Also very interesting, like we, we had conversations where, 723 - > customers will ask you.
724 - > What are you doing with ai? 725 - > And then we would ask them, well, what would you like? 726 - > And they're like, well, we don't know. 727 - > You tell us.
728 - > It's not like people know, they just know they, they know the 729 - > technology is out there and they want somebody that's they can 730 - > bet on it's gonna execute in the future and not gonna become a 731 - > legacy player in a sense. 732 - > I think that's the biggest way of changed it, but also on a lot 733 - > of topics, people are hesitant, I would say skeptic. 734 - > ai for various reasons. 735 - > Sometimes it doesn't really work, so they're ah, think you 736 - > know what I'm talking about, right?
737 - > That, that, so those situations, people just want a traditional 738 - > piece of software that, does the job that they needed to do. 739 - > And we're very flexible in both of those situations. 740 - > So I think that that's good for us. 741 - > dan-balcauski_1_03-10-2026: Have you, uh, maybe you're not 742 - > directly involved with this conversation, maybe you are with 743 - > some of the, you know, larger customers have you noticed a 744 - > increase in sort of customers.
745 - > Fluency and understanding of the technology such that, I've 746 - > talked to other folks on this program who, you know, depending 747 - > upon the industry that they're in they've had changes to that 748 - > have affected them at like a procurement level where like now 749 - > they have to go through, you know, especially selling 750 - > enterprises. 751 - > Maybe they have an AI technology board who they now have to go 752 - > through. 753 - > Um, have you noticed a change in, your customer base of how 754 - > they think about AI technologies or maybe just'cause you of the 755 - > market you serve that hasn't been as much of an issue for 756 - > you.
757 - > tom_1_03-10-2026_171009: Yeah, that actually hasn't been much 758 - > of an issue. 759 - > We al there's always the whole data governments question and 760 - > that's been around even before AI now is like a little bit 761 - > more. 762 - > So, are you training, the models getting trained on our data, et 763 - > cetera, et cetera. 764 - > So these questions which, we give answers to and that's okay.
765 - > Other than that, we didn't see a lot of problems there. 766 - > Like I said, it's customers are either, we want this, it's 767 - > interesting, and we are or we don't want it. 768 - > But then the other thing also that's very interesting for me 769 - > is that sometimes people consider AI basically magic, 770 - > which is interesting. 771 - > Like you, when you talk to them and you see some use cases and 772 - > those are use cases that don't have anything to do with ai.
773 - > It's more around some like regular automation or like a 774 - > workflow or something. 775 - > But it's more like, I just want this problem solved. 776 - > And I want AI to do it. 777 - > It's it's very interesting.
778 - > dan-balcauski_1_03-10-2026_1: I, I do wanna transition a little 779 - > bit because we, I've been asking you a lot of product focus 780 - > questions. 781 - > And as I understand, you're still leading product at 782 - > productive. 783 - > Is that correct? 784 - > tom_1_03-10-2026_171009: Yeah.
785 - > Yeah. 786 - > So I lead a team Product managers and product managers 787 - > working on the, on the product. 788 - > dan-balcauski_1_03-10-2026_1: At your scale of about 150 folks, I 789 - > think a lot of folks might be surprised by that. 790 - > May, maybe, maybe not.
791 - > What has been your impetus to still, keep that hat on as 792 - > you've grown the company? 793 - > tom_1_03-10-2026_171009: I think I'm, I've always been a product 794 - > person. 795 - > It's part of my nature. 796 - > What is that saying?
797 - > When you have a hammer, every problem is a nail or something 798 - > like that. 799 - > So basically that's me in my career. 800 - > I've always, when I had a problem, I tried to solve it by 801 - > building something, like a product to 802 - > dan-balcauski_1_03-10-2026: Hmm. 803 - > tom_1_03-10-2026_171009: and that's how actually I ended up 804 - > here.
805 - > So it is something that I think I'm good at, or at least that I 806 - > like doing, which kind of then just motivates me to, invest a 807 - > lot of effort into it. 808 - > Very early on we started building the product and we 809 - > started doing sales. 810 - > My, my basically idea was. 811 - > I don't wanna sell the product.
812 - > 'cause if I have to sell it, and I'm gonna probably do it for 30, 813 - > 40% of my time. 814 - > And it's never gonna be, great. 815 - > So out the door, we hired, salespeople at the beginning and 816 - > I wanted them to sell it. 817 - > And then the other idea was also if they can't sell it, then the 818 - > product is not good enough.
819 - > If I have to come in personally and, convince somebody, well 820 - > that product really is not good enough. 821 - > And then me as the product builder has to make it better. 822 - > it, we started like that, selling it. 823 - > And that was the phase where until you get product market 824 - > fit.
825 - > It. 826 - > as we got it later on, it just the flywheel starts and that was 827 - > good. 828 - > So yeah, that's like my main drive. 829 - > I obviously run the entire company, but like for this 830 - > particular area, I don't have a VP of product.
831 - > 'cause that's like my dual role for all the other areas. 832 - > Marketing, et cetera, et cetera. 833 - > I do have, technology, I have VPs that handle it, but this is 834 - > just like an area that's very close to me. 835 - > The question I always ask myself, obviously is how long 836 - > is, can I keep it up?
837 - > And I don't know. 838 - > We'll see. 839 - > I think right now in this whole crazy AI age where everything is 840 - > changing, you mentioned vibe, coding there's a lot of stuff. 841 - > Ev everything is getting new and the cards are being rearranged.
842 - > I think as a founder. 843 - > very beneficial to have you close to the actual product and 844 - > to the users in a way, just by proxy and to their needs and 845 - > requests. 846 - > I think it's beneficial at this stage, and I guess at any stage, 847 - > but at this stage I think it's even more beneficial than 848 - > before. 849 - > dan-balcauski_1_03-10-2026: Yeah we're seeing that actually 850 - > across even some very mature companies where, founders are 851 - > getting involved even all the way going to Sergey brand at 852 - > Google.
853 - > Right. 854 - > Jumping back in, started writing code again which I don't know 855 - > that anybody had on their bingo card in 2020. 856 - > But I'm curious like. 857 - > Given that role, like, how do you decide, like what does your 858 - > schedule look like?
859 - > How do you decide to split time between sort of the focus on the 860 - > product and the broader responsibility? 861 - > Because I'm thinking, um, that must imply a serious amount of, 862 - > uh, delegation, uh, work on your behalf. 863 - > I'm just curious how you've how you've approached that. 864 - > tom_1_03-10-2026_171009: Yeah, I try to do like a cadence.
865 - > So basically, I don't know, every Monday I have a bunch of 866 - > meetings, like one-on-ones with with my direct reports, and 867 - > that's like the company building part the sales part, the 868 - > business part mostly around that marketing, et cetera. 869 - > I'm focused on the product, so I do syncs with all the product 870 - > people and see where we are with that. 871 - > And Wednesdays and Thursdays, I like to keep open as much as I 872 - > can to do like workshops and to work with actual, people in the 873 - > office.
874 - > Let's sit down. 875 - > There's a problem when people need my help. 876 - > The people don't need my help like all the time, often they do 877 - > 'cause. 878 - > I have the knowledge, the, like, how customers work and also how 879 - > the product works and like decisions we've made in the past 880 - > and why did we make them, et cetera.
881 - > But yeah, so I like to keep a big part of my schedule actually 882 - > open so that people can, pull me into these discussions. 883 - > and that's been working really well. 884 - > dan-balcauski_1_03-10-2026_11: I wish you continued success. 885 - > It sounds like you have a highly regimented approach.
886 - > So probably more so than I could pull off. 887 - > So quite impressed by that. 888 - > tom_1_03-10-2026_171009: So today's Tuesday. 889 - > I like you're my last meeting today.
890 - > I've been like at it since morning. 891 - > Half an hour. 892 - > Half an hour. 893 - > Half an hour.
894 - > I think I had 10, 10 meetings. 895 - > It's very fast. 896 - > It's very, what are problems? 897 - > Okay, try this.
898 - > I don't know. 899 - > Okay, we don't have a solution. 900 - > Let's do a workshop, let's do brainstorming, et cetera, et 901 - > cetera. 902 - > And that's every week.
903 - > That's what I'm doing. 904 - > dan-balcauski_1_03-10-2026_1: Oh Tom, this has been great. 905 - > We're running up on time, so I want to segue out to a couple of 906 - > rapid fire closeout questions. 907 - > Is that okay?
908 - > tom_1_03-10-2026_171009: sure. 909 - > dan-balcauski_1_03-10-2026: What is either a book or podcast you 910 - > find yourself recommending? 911 - > Most other people these days? 912 - > tom_1_03-10-2026_171009: Ooh, our podcasts are like my thing.
913 - > I dunno. 914 - > I like all the landing podcasts of a product. 915 - > I like Pragmatic engineer, which is very interesting to figure 916 - > out, how people are coding today. 917 - > I think those are the two main ones like out, out of the big 918 - > ones that I listen to.
919 - > dan-balcauski_1_03-10-202: Nice. 920 - > When you think about all the spectacular people that you've 921 - > had a chance to work with, is there anyone who just pops to 922 - > mind who's had a disproportionate effect about 923 - > the way you think about building or leading companies now? 924 - > tom_1_03-10-2026_171009: Oh, that's interesting. 925 - > I think, my career has been mostly influenced by, people 926 - > that have I've lived with, like, give you an example.
927 - > When I went to high school, I had this guy who was a computer 928 - > science or whatever it's called, the course. 929 - > Teacher and he never actually taught me anything about 930 - > computers or anything, but he gave me so much confidence that 931 - > I could do it all, and that basically labeled my life, up 932 - > until today. 933 - > The interesting thing is, 25 years later, I'm still friends 934 - > with him. 935 - > dan-balcauski_1_03-10-2026_11: I gotta ask the follow up what 936 - > did, what was his method by which he gave you so much 937 - > confidence?
938 - > I think that would be a amazing for other folks. 939 - > tom_1_03-10-2026_171009: It's, he, at some point, like I was 15 940 - > years old and I already knew coding, like there's nothing I 941 - > could actually learn in the actual courses. 942 - > 'cause I basically knew it all because I liked it. 943 - > I was doing it in my spare time or something and he was like, 944 - > You, it was me and my co.
945 - > Co-founder later, but my friend back then, and he was like, 946 - > okay, you guys know more than me. 947 - > You guys know more. 948 - > Here's just, here's computers, here's time. 949 - > You guys are smart.
950 - > Figure it out. 951 - > And constantly when you would maybe dive yourself or 952 - > something, you would say, okay, you're smart. 953 - > Figure it out. 954 - > You can do it.
955 - > Ah, don't gimme that, blah, blah, blah. 956 - > You're building up your confidence that you can actually 957 - > do it. 958 - > I think that's important. 959 - > Like in your life, you need probably just one person to tell 960 - > you can do it right.
961 - > dan-balcauski_1_03-10-2026: Yes. 962 - > We're all lucky to have such folks. 963 - > Glad you guys have remained friends over all those years. 964 - > If I gave you a billboard and you can put any advice on there 965 - > for other B2B Sass CEOs trying to scale other companies, what 966 - > would it say?
967 - > tom_1_03-10-2026_171009: The entire playbook I think is 968 - > changing right now. 969 - > I don't think a lot of what was worked before in terms of SaaS 970 - > is gonna work in the future. 971 - > So you need to think everything you know, from grounds up from 972 - > first principles. 973 - > What is this gonna look like in, five years?
974 - > These, like, technological shifts don't happen that often. 975 - > And I think that, yeah, that's my main advice. 976 - > Everything's getting rewritten, so, think everything through 977 - > yeah. 978 - > If you're reading a book 10 years old, how to build a SaaS 979 - > company it's not applicable today.
980 - > So it just you have to figure it out yourself. 981 - > dan-balcauski_1_03-10-202: Yeah. 982 - > Well, that's a long billboard. 983 - > I think it's the, all the playbooks are changing.
984 - > Figure it out. 985 - > Uh, for the prince of. 986 - > Well, and also the reason that this podcast exists, so, 987 - > continue to tune in to other leaders as we figure out what 988 - > they're learning from the ground up. 989 - > This has been fantastic, Tom.
990 - > For listeners wanna follow you connect with you around the 991 - > internet or learn more about productive, how could they do 992 - > that? 993 - > tom_1_03-10-2026_171009: Yes, probably LinkedIn. 994 - > You can connect with me there. 995 - > Follow me there.
996 - > See what I really up to. 997 - > And yeah, you can also follow me on Instagram if you wanna know 998 - > where I travel and what type of music I listen to. 999 - > And that, that that's the two main places. 1000 - > dan-balcauski_1_03-10-20: Which, what's go-to on your Spotify or 1001 - > iTunes these days?
1002 - > What's what are you listening to? 1003 - > tom_1_03-10-2026_171009: Oh, it's like Croatian, Serbian 1004 - > music. 1005 - > But 1006 - > dan-balcauski_1_03-10-2026_1: If I had to, if I had to listen to 1007 - > one artist, if I had to listen to one artist in that category, 1008 - > who would it be? 1009 - > Gimme a name.
1010 - > tom_1_03-10-2026_171009: Oh, lately, oh, we have voco v 1011 - > That's these an interesting artist. 1012 - > He has a new song. 1013 - > dan-balcauski_1_03-10-202: Okay. 1014 - > tom_1_03-10-2026_171009: And 1015 - > dan-balcauski_1_03-10-2026_: All right.
1016 - > I will, I will, 1017 - > tom_1_03-10-2026_171009: I listened to Blink 1 8 2, which 1018 - > is like music I've listened, has had listened to for 20 something 1019 - > years. 1020 - > It's probably now classic rock or something that you would call 1021 - > it today, whatever. 1022 - > dan-balcauski_1_03-10-202: Yeah. 1023 - > Yeah.
1024 - > Well, I remember when as a child, I used to go to the 1025 - > grocery store. 1026 - > They play all these old old timey songs, and now I go and 1027 - > they play these amazing bangers. 1028 - > So I don't know what that means about about my life. 1029 - > Well, I'll put links into the show notes for of all that for 1030 - > listeners that wraps up this episode of Sask Galy Secrets.
1031 - > Thank you to Tom for sharing his journey and insights. 1032 - > For our listeners, if you found Tom's insights valuable, please 1033 - > leave a review and share this episode with your network. 1034 - > tom_1_03-10-2026_171009: The.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.
Why B2B Brands Are Using AI to Write Sales ProposalsThe Growth Operator with Fexingo · on Prompt engineering85 / 100
Episode 018: Season 2, the $75 Consult and the Frankenstein StackAI Tools for Practicing Lawyers · on Prompt engineering84 / 100
Episode 111: Building Your Defences Against AI MisinformationValue Driven Data Science · on Prompt engineering83 / 100
How B2B Marketers Use AI to Personalize at Scale for EnterpriseB2B Marketing with Fexingo · on Prompt engineering82 / 100
477. The Nitty Gritty of AI From an Attorney and AI Expert with Mike BrownThe Game Changing Attorney Podcast with Michael Mogill · on Prompt engineering81 / 100
Greg Joseph: AI’s Future in Personalizing Streaming Ads10 Minute Martech · on Prompt engineering75 / 100From HBS Dropout to Successful SaaS Exit with Palash Soni, CEO of Goldcast85 / 100Tackling AI's Blind Spots with Abhishek Jha, CEO of Elucidata83 / 100Conquering the Desperation Mindset in SaaS Pricing with Apurv Bansal, CEO of Zenskar82 / 100Scaling a SaaS Selling to Skeptical Scientists with Mike Klein, CEO of Genomenon86 / 100Balancing a Risk-Reward Model in Real Estate Tech with Alvaro Erize, CEO of CINC86 / 100