Most founders try to expand margin from the top line. Bart Bartlett doubled his margins from the cost side in a single year — and he’s still pushing. In this session, the founder of DemandZen walks through the operating changes that took his monthly margin to 26%, with a line of sight to 30%+, while running 50% growth. The playbook is unglamorous and copyable: kill the tools that don’t pay, route every non-payroll dollar through a single procurement system, close the books in a day, and arbitrage every vendor renewal against the market.
What you’ll get from this session:
- How to consolidate spend under a single approval and virtual-card system so every dollar is tracked, categorized, and controllable
- The cadence and tooling Bart uses to get to a one-day month-end close
- How to run vendor pricing as an ongoing arbitrage, not a once-a-year renewal exercise
Why it matters:
- Cost-side margin gains are available to any founder right now; top-line moves take years
- A disciplined procurement and close process directly raises exit multiple through higher EBITDA and cleaner financials
- Founders who don’t tighten operations in 2026 will watch AI-native competitors undercut them on price
TRANSCRIPT
Greg Alexander: Hey everybody, this is Greg Alexander. You’re listening to the Pro Serv Podcast, brought to you by Collective 54. If you’re new to this show, we aim to do three things. We want to help you make more money, make scaling easier, and make an exit achievable. This show is for founders of boutique professional services firms, so if you market, sell, and deliver expertise for a living, this is for you. On today’s episode, we’re going to talk about margin expansion, but in particular, how expense management might be the way to go for some of you. And we’ve got a long-standing, well-respected, well-liked member here to share what he’s done recently on this. His name is Bart Bartlett, he is the founder and CEO of DemandZen. Bart, it’s good to see you again. Welcome to the show.
Bart Bartlett: Thank you.
Greg Alexander: So, for those that haven’t heard from you in a while, since you’re an OG, as we like to say, would you mind, reintroducing yourself and your firm?
Bart Bartlett: Sure, yeah, so my name is Bart Bartlett, I’m the CEO and co-founder of Demand Zen. We’ve been around for about 12 years. We are a professional services firm that does outbound demand generation, primarily for software and SaaS companies, with a little bit of ProServe.
Greg Alexander: Alright, sounds great. Alright, so Bart, when we last spoke on office hours, you told me that you doubled your margin. So, give us the headlines. So, what changed over the last year or so?
Bart Bartlett: So, we’ve worked really hard at expense control, expense reduction, SaaS reduction, and then we’ve basically expanded the business without adding headcount. Or minimal headcount, I would say.
Greg Alexander: Okay? Now, when people go after margin expansion, they can do it one of two ways. They can do it on the revenue side, which is usually price increases. or they can do it on the expense side, or a combination, I should say, of the two, which sounds like what you did. But I wanted to spend some time specifically on the expense side. So why did you decide to improve margins by going after expenses?
Bart Bartlett: Well, I heard about this thing called Sasspocalypse, and I figured I should become a participant. So, in practice, right, I’ve always been a huge fan of SaaS tools. I’ve commonly referred to myself as a SaaSaholic. And when I looked across the business, I think we had 139 different tools.
Greg Alexander: Wow.
Bart Bartlett: order or another, between the ones that I knew about and the ones that I found. And just based on kind of what I’ve seen in the market in the last year, I was just like, okay, you know, this is too much, and what can we get rid of? And what can we, you know, repurpose or approach in a different way?
Greg Alexander: So how did you decide, you know, you mentioned to me. sass murder, I think was the word that I wrote down. So how do you decide which SaaS tools stayed and which got killed?
Bart Bartlett: I went through and looked, there’s a… there’s a product called SpendHound from a company called Yippy Data, and basically for free. They will, ingest all your transactions, and kind of surface what your SaaS spend is. So you kind of send them your contracts, you give them access to QuickBooks or Xero. And they come back 3 weeks later, and they’re like, here’s what we found. And then you can rank that by monthly agreements, annual agreements, And so, I looked at, you know, what was coming up annually that we could get rid of. Or just, the other thing is, where can you drive down costs, right? Because part of what’s happening with SaaSpocalypse is there are lower cost alternatives in the marketplace. Or, you can potentially use AI to consolidate from one tool into another tool. So I literally just look at SpendTown probably every week or two, and I’m like, okay, what’s coming up that I can get rid of? And then I look at monthlies, like, anything that’s monthly is on the chopping block if I can get rid of it.
Greg Alexander: Nope. Okay, you also, got to a… One day, month-end close. Which, caught my attention. I’m not quite there yet, I’m not that far away, but you are my hero, because you got to one day close. I’m waiting on my bank sometimes, which is a whole other topic for another day. But, you know, what had to change operationally to get to a one-day close?
Bart Bartlett: Well, I would say it’s usually one day, it’s not always one day, right? But in practice, everything… I mean, 80% of our costs are payroll, right? And so that’s going through Pelosity, and that pushes into QuickBooks. And then… pretty much all of our expenses now go through either RAMP spend cards, or we use RAMP bill pay. Which is awesome. Like, you know, so essentially, if I can pay it on a credit card, I do, because we get, I think, 2% back. Or, if it’s a paper bill, we forward it to RAMP, it ingests it, lets us schedule it, and then we schedule a payment. So everything’s pretty defined.
Greg Alexander: You know, I’m happy to say that we, and we’re going to announce this on Wednesday on the member session, that Collective 54 and RAMP have formed a partnership Where we’re bringing the RAMP products and services to our community in a unique way, applied to ProServe in particular. And, I’m super excited about it, because I think many of the things that RAMP does are highly relevant. to our community. If you wouldn’t mind, would you share maybe 2 or 3 things that you’re… ramp features that you’re using, and how they’ve helped you?
Bart Bartlett: Yeah, so RAMP’s been great, just in the ability to very granularly create temporary or virtual credit cards, and as well as giving physical cards. So, if you have employees in the US, which the majority for us are, you can issue them physical cards, then you can put, essentially, no spend, some spend, or a lot of spend behind those cards. You can turn them on and off from your phone. Or from the web. And then you can also allocate virtual cards per vendor. Which we use a lot, right? Because the easiest way to turn off a vendor is click the card so payment doesn’t go through anymore.
Greg Alexander: Yeah. The other thing that I like about it, particularly in the consulting sector, management consulting, there’s a lot of travel that happens in that segment. And, you know, a founder will have a travel policy. You know, many times they’re using the client’s travel policy, because it’s reimbursed by the client. And then, you know, a consultant goes out and breaks the policy, and now you’re dealing with an exception after the money was spent, which is never a good position to be in. And RAMP allows you to enforce those things at the point of sale, like literally reject the spend before it even happens, which I think is great. There’s lots of other things that they do. One thing that RAMP touts, which is the reason why I got behind it so much, is that they are, you know, AI financial operations, and they’re using AI in all kinds of interesting ways to make our life happy. Have you been using some of their AI features?
Bart Bartlett: I am way too cheap to pay for the area features.
Greg Alexander: Okay. I’m using them like crazy, so I guess I’m not way too cheap.
Bart Bartlett: Do I have different employee numbers?
Greg Alexander: Yeah, yeah, true, I guess so, that’s right, yeah. Yeah, the… some of the things I really like is the fraud detection stuff. I mean, that’s been… that’s been pretty interesting. Also, the, auto-categorization. Yeah. You know, so, it’s saving me real dollars, because… the fractional CFO expense is going down.
Bart Bartlett: Nope.
Greg Alexander: Because they were spending hours, which are getting billed to me, you know, on categorizing expenses, and I don’t know, maybe it’s 90% of my expenses are the same every month, so the categorization should be automated, right? And that’s what they do.
Bart Bartlett: Yeah, no, I agree, that’s been great. So we pre… essentially, part of what we do with the spend allocation is we can tie the card to specific accounts, so we’re sort of in the same position. RAMP loads up all the transactions, it pre-categorizes everything, and then our controller looks at it once a day and basically clicks yes, and that’s it.
Greg Alexander: So this SaaS tools sprawl, 139 tools, that’s an unbelievable number of tools. How… what have you gotten it down to?
Bart Bartlett: You know, I should probably count. All I do is kill them at this point, but we’re probably in… we’re probably still in the 40 to 50 range. But a lot of the spend, it’s a long tail of smaller spend. Right? So, my biggest bill at this point is Salesforce, followed by the vendor that we use for outbound calling, and then I spend a lot of money on data, right? Because that’s kind of the lifeblood of what we do. But it doesn’t take that long to where I’m, instead of looking at annualized amounts of $100,000, I’m looking at annualized amounts of $6,000, right? So, some of these are, like, 50 bucks. But I’m literally trying to get rid of everything, right? So, like, we had a… We had a vendor that recorded sales calls, and they were good. It wasn’t… it was way cheaper than Gong, they didn’t have a big platform fee, it was a decent function, but now that’s… there’s a company called Fellow AI that now records all our meetings across the company, and they have a sales function. Right, so that went from $100 a month per sales rep to $16.
Greg Alexander: Yeah. When you replace these. Let’s call them expensive tools, which isn’t necessarily fair in all cases, with these less expensive tools. how do you make the decision on whether or not to subscribe to the next SaaS tool, or just build your own, you know, vibe-coded version of it?
Bart Bartlett: So, we’re doing a mixture of both. They’re… from a… from a SaaS tool and SaaS alternative standpoint, Vendor.com has an AI tool called Roof. And we subscribed to that. It was free, which is how they got me, but then once I ran, like, 30 different queries, because I’m constantly cutting expenses, they were like, hey, you need to actually pay for this now. And so we negotiated a deal, and we do. But I can literally just say, hey, what are the lower-cost alternatives to this tool? So I’ll ask Claude. I’ll ask vendor, and I’ll kind of see what the options are. And then it’s a question of utility versus essentially switching costs, if that makes sense?
Greg Alexander: Yeah. So let’s talk about the switching costs. So you mentioned you’re a Salesforce customer. And I would imagine that’s your system of record, for lack of a better term, and you’ve got all kinds of plugins and workflows built around it. So, is a system like that safe, or are you thinking about replacing something like that?
Bart Bartlett: It is safe-ish. Right? I mean, if I could kill it, I would. I can’t for now, but that doesn’t mean that won’t change.
Greg Alexander: And you can’t because the switching costs would just be too high?
Bart Bartlett: Yeah. I mean, there’s… we’ve got years of development on top of it, all of our data is there, 3 million people… 3 million kind people that we’ve called, right? So that just is a lot to try and move over, and it also is the fundamental business system that we use, so it’s, you know, that’s a big risk.
Greg Alexander: Yeah. So I wanted to call that one out in particular, because some of our members are taking this too far. And they’re trying to replace everything. And, you know, there is… don’t get me wrong, all these fringe tools that you’re just assassinating, I think, is the right strategy all day, every day. But these systems that might have a little bit of a moat around them, you know, because of, you know, all the years of development, and they are the system of record, etc. You gotta ask yourself the question at some point, is the dollars that you’re gonna save, are they large enough to go through the headache, you know, of making the swap? Which takes me to my next question, you know, you’ve got employees, you know, they’re doing their job, they’re used to using all of these tools. And all of a sudden, these tools are going away, or they’re getting replaced by something else. What’s been the learning curve?
Bart Bartlett: Well, our employees are pretty focused on using our dialer as their primary interface. They do limited work in Salesforce, but we’ve been using Salesforce flows and automation and AI and APEX code to reduce what they do, right? So, for us to make incremental margin, we need our callers to spend more time talking to people, setting appointments, following up, and less time inputting data, looking for people to call, sourcing. So we’ve built a lot of incremental functionality around the things that we want them to stop doing, if that makes sense.
Greg Alexander: Okay, yeah, it does make sense. Alright, now you, as I have known you over the years, you are a, let’s say, forward-thinking technologist. I mean, you were a tech-automated service firm before people even know what that was. You were an AI firm before people were even talking about what that is. How… if you were… if you were speaking directly to some of our members who are not nearly as forward-thinking when it comes to stuff like this. And you would have given one or two, maybe three things to get started on. What would it be?
Bart Bartlett: Oh, I mean, at this point, it’s ask Claude Cowork or OpenAIS Codex application a complicated question and see what you get back. Right? Just try it.
Greg Alexander: What’s the complicated question, you think?
Bart Bartlett: Well, so I read today’s blog post about quality of earnings, and I’m like, I wonder if Clog can do that. And the answer is it can, but not quite, right? I’m just like, okay, I’ve never run a quality of earnings report, and the blog post said that’s $30,000 to $50,000 for a mid-range company, and I’m like, well, I’m too cheap for that, but I can… I can ask Claude if it’ll do it, right? So… At this point, I don’t have any particular AI allegiance. I run OpenAI’s Codex, I run Cloud Cowork, I run Google Gemini, and I use them all for slightly different things.
Greg Alexander: Yep.
Bart Bartlett: But it’s like, it’s kind of like the known knowledge of the world. That will, for the most part, give you the right answer.
Greg Alexander: Yeah. You know, on something like that, too, a quality of earnings, it might not be able to give you the right answer, but if you have a sample quality of earnings. and you load it into Claude, and you say, reverse engineer this for me, then it might ask you a handful of questions, and then presto, bingo, there it is. I mean, it’s quite remarkable, you know, what’s available to us. You know, this is a sign. A glaring red flashing sign that the software industry is completely mature. It has been completely commoditized, because we’re able to swap out vendors for pennies, sometimes for free. Where do you think all this is going? What is the end game for the SaaSpocalypse?
Bart Bartlett: Oh, I mean, I think it’s… where… where is there… where do you have a fundamentally differentiated offering, right? Like, Salesforce has all of my data, and it has for 11 years, right? So the switching cost of the business logic and process around that is very hard to move. And they know it, right? Our dialer is highly optimized, and based on the data that we get back from it, we optimize how we call and how we follow up. Right? So that… that makes that hard to change. So there’s… I think there’s gonna be certain core functions that are differentiated or hard to change, and then there’s gonna be everything else, right? Yeah. Like, we have a knowledge management system, and I like it. It’s really well done, right? It’s got a little thing that we click on, and you can search everything. But I was just like, okay, we’re managing 500 cards, or 500 web pages of knowledge, and it’s how we run the company. But I just asked Claude, hey, I don’t want to pay for this anymore, what can I do? I’ve got Google Workspace, and it’s like, well, how about if I just move it all for you, and we build a Google site on top of it? Right. So then I just assign that to someone to figure out between now and our renewal.
Greg Alexander: Interesting. And then, when you sign up with these new vendors, given this new strategy, and you’re negotiating with them, has this changed your negotiation in terms of contract terms and length and pricing and all that?
Bart Bartlett: Oh my god, yes. That’s why we pay money for Ruth Fender. Ruth Fender, I literally put in, I have a Salesforce contract with this many enterprise licenses and 75 people. Give me the 10th, 25th, 50th, and 75th percentiles on the contracts that you’ve seen recently. Right. So I can tell you that my Salesforce cost for an enterprise license is $2 more than the 10th percentile at our revenue band.
Greg Alexander: Unbelievable. It’s like strategic sourcing that’s been AI-enabled. I mean, it’s… it’s amazing. Sometimes you wonder what’s going to happen to firms like Gartner Group and others. Let me ask you… I’m going to take the conversation in a slightly different direction, just because of all your experience. Your answer to this question would be interesting to me. I saw an interview the other day on CNBC with the founder of Vista Equity. His name is Robert Smith. And he’s considered by manyDB the sharpest investor in SaaS in the world. And he was talking about the AI factory and what’s gonna happen to his 90 portfolio companies, many of which are SaaS, and are they going to go out of business? And his attitude was, no, they’re not going to go out of business, and of course, now he has a reason for saying that, but the reason for that is this question of, are you okay with a probabilistic answer?
Bart Bartlett: Right.
Greg Alexander: Or a deterministic answer. Do you think that detailed about this? And do you have any concerns around the risk side of this? Or is, you know, good enough good enough?
Bart Bartlett: Well, so there’s… I don’t know if I have a short answer to that, but it’s kind of like, if you can’t beat them, join them at a high level. And then, the issues that AI has, right, is essentially if it’s… I forget what the specific term is, but if the memory that it’s commonly running in GPUs runs out or gets 90% full, it starts to just make stuff up. But if, and what we try and do is gate the problem, make the problem, the thing that you’re asking it, tiny. Right? So, like, we have… we have highly automated sales research that we do for every call before we meet with a prospect. That involves 28 specific searches, and those are discrete, right? And so those are all running against separate agents. But essentially, if you tell it what the input is. And what the expected output is, and you gate that, essentially you put a harness around it, and you’re not asking overly complex questions, then you’re gonna most… you’ll get an answer that’s right, you know, 99 out of 100 times. And then, as you’ve mentioned, and I’m totally on board with, you gotta have a human in the loop, right? Like, someone has to actually look at this before you present it to a client or a prospect.
Greg Alexander: Yeah. You know, and in that interview, he also talked about how the software industry, TAM, is going to go way up. And the reason for that is, historically, you know, if you bought a piece of software, let’s say you spent $100,000 on it over the course of a year, it might cost you a million dollars from a services company to get it implemented.
Bart Bartlett: Right.
Greg Alexander: So with AI now, the SaaS companies are going after that labor budget, that service budget. And he was boldly making the claim that, you know, the Accentures of the world, you know, the big systems integrators, they’re in serious trouble, because a lot of that work is going to get consumed by the AI. What’s your take on that?
Bart Bartlett: I completely agree. I mean, the stuff that I can now do in minutes that, you know. I mean, I wouldn’t necessarily have hired a consultant, but I probably would have had to read the manual and spend hours on it, right? Now stuff takes 5 minutes. It’s just like, how do I do this? And it gives me an answer, and I just work through the steps, and if it doesn’t work, I’m like, hey, this didn’t work, how do I fix it? And then it’s like, well, try this.
Greg Alexander: Yeah, yeah. Yeah, so the lesson there for members that are listening to this, you know, if you’re implementing software the old way. is to have a systems integrator charge you a bunch of money to implement that software. The new way is to just work yourself through it with the AI assistant helping you to do it. that’s gonna work for a lot of applications, particularly for the type of software that our small businesses use. Might get a little bit more complicated, that, and of course, the devil’s in the details when you get into things like, you know, sophisticated enterprise rollouts, but we’ll see. But it’s an interesting time, you know, of how services and software are colliding. All right, well, I think we’re at our time window. This was very interesting, it was extremely practical, and I’m hopeful that those that tuned in are going to put a few more nickels in their pocket, because they listened to your session today, Bart, so thanks for joining.
Bart Bartlett: Yeah, thank you, sir.
Greg Alexander: Alright. Okay, a couple calls to action, so if this topic was of interest to you, and you want to ask Bart some specific questions, maybe about some of those tools and how to use them, attend the member session, which will be Friday. You’ll get a meeting invitation for that, and we’ll have our private Q&A. If you’re not a member, and you think after listening to this you might want to become one, go to Collective54.com, fill out an application, and someone will get back in contact with you. But I appreciate your attention, and I wish you the best of luck as you try to grow, scale, and someday sell your firm.