// transcript — 818 segments
0:00 Coming Up
0:01 There's a reason this book is not called the SAS blueprint. I intentionally
0:05 called it a playbook. SAS at scale, you can see 80 to 90% gross margins and 30
0:10 to 50% net margins. Companies like Zoom went public with more money in the bank
0:15 than they had raised from investors. This is why there's subscription
0:19 underwear and their subscription razors and their subscription snack boxes.
0:23 There's a lot of developers building a lot of [ __ ] in their basement for 6
0:27 months that nobody wants. SAS is everything. SAS is dead. These are our
0:31 two extremes. Guess what? It's somewhere in between. >> You're listening to the Startup Podcast.
0:39 This is an educational episode. >> Hi, I'm Yanv Bernstein
0:42 >> and I'm Rob Wallally. >> For those of you who are on LinkedIn,
0:45 On Today's Show
0:45 you've probably seen all of those silly posts about what this thing taught me
0:49 about B2B SAS and you know what having a child taught me about B2B SAS. Well,
0:53 those are all silly, but the truth is B2B SAS is one of the most powerful
0:57 business models that has ever been created. And today we are going to talk
1:02 about B2B SAS in depth. Not through silly analogies with life, but actually
1:07 with somebody who has lived this for decades, who has founded their own SAS,
1:13 who is running a community of hundreds, thousands of B2B SAS founders, who has
1:17 invested in hundreds of B2C SAS businesses. This person is an OG of the
1:22 space, really really believes in the power of B2B SAS to change lives and to
1:27 be the ultimate vehicle for your startup. So whether you agree with that
1:31 or not, you will get a huge amount out of listening to this episode with our
1:34 very special guest. Find out who it is right after the break. Today's episode
1:38 of the Startup Podcast is brought to you byte domains. One thing about starting
1:42 your startup that should be easy but isn't is naming it. Why is it so hard?
1:47 Because all the good.com domain names are taken. So you end up doing something
1:50 silly like misspelling it or adding random prefixes or suffixes. No more.
1:55 Your domain shows your focus and intent. So you should make it sharp, clean, and
2:00 unmistakably tech with a tech domain. Head to get.te/tsp or a registar like GoDaddy to grab your
2:13 for the startup podcast. Today it's our absolute pleasure to
2:16 welcome Rob Walling back on the show. This is actually his second time on the
2:21 startup podcast. Now Rob is an absolute legend in the startup and SAS space. I
2:26 think he is famous especially for two things. For believing in bootstrap
2:30 startups and for believing that B2B SAS is the best business model in tech. And
2:34 you know what? He makes some really convincing cases for both. Rob is an OG
2:39 content creator. His podcast startups for the rest of us has been going for
2:43 years. It's fantastic. He is the founder and CEO of Microconf and Tiny Seed,
2:49 which is a community and a fund for mostly bootstrap SAS founders. He is a
2:53 font of wisdom. He's been a founder and exited founder himself. So Rob is the
2:58 absolute perfect person to talk about B2B SAS and why you should be thinking
3:02 about building in that space. Rob, welcome back to the startup podcast.
3:06 Why B2B SaaS is the Best Business Model
3:06 >> It is my pleasure. It's great to be back. Rob, you are famous for, amongst
3:11 other things, being the B2B SAS guy and saying B2B SAS is the best business
3:16 model for startups. Now, why is it that you say that? >> It's something I've thought for a long
3:20 time, but really didn't come out and put my put a what do you call it? A stake in
3:25 the sand, a a flag in the on the moon. SAS is incredible because it's recurring revenue. Like,
3:33 that's the that's the core of it. And every business wants to get
3:37 subscriptions. Like this is why their subscription underwear and their
3:40 subscription razors and their subscription snack boxes because one off
3:45 purchases, they just aren't great. And I've run you and I probably both run
3:49 businesses where it's like start of the month zero revenue for the month. And
3:53 that's it's it's tough. And if you if you have MR monthly recurring first of
3:57 the month, you're kind of where you were last month, you know, minus churn and
4:01 and blah blah blah. So that really is at the core of it. And beyond that, um,
4:07 especially B2B SAS because I have long ago stopped talking about B TOC. I did a
4:11 couple of those and they were they were really brutal. But B2B SAS is a lot less
4:17 dependent on luck than most startups that you would, you know, raise venture
4:21 for and go after. There's no blueprint, but there are some repeatable angles of
4:25 like, hey, businesses are willing to pay to solve problems. I got to find a
4:29 problem a business solves. I got to validate it. I got to, you know, try to
4:32 build it. I got to iterate with it. do customer development, blah blah blah.
4:35 So, there's a little more to me, it's it's about repeatability that I like.
4:40 Again, it's not 100% of the time it'll be successful, but compared to starting
4:45 two-sided marketplaces or trying to invent a new type of drone or, you know,
4:49 just whatever startup we can think of, a medtech, a this or that, like B2B SAS is
4:55 boring, but much higher likelihood of success if you're kind of following some
4:59 rules of thumb or best practices. >> Let's dig into those. So the first one
5:02 the recurring revenue you know a lot of founders including myself don't come
5:07 from a business background as such right and we might be product people or we
5:09 might have worked in larger organizations and so a concept like
5:13 quality of revenue might not be the thing that is front of mind for some
5:17 founders but it is so important you know a dollar is not a dollar a dollar
5:21 depending on how you make it and on how predictable it is and on the cash flow
5:26 cycle for that dollar is actually a very different thing and recurring revenue as
5:30 you said instead of every month or every year starting at zero and having to
5:33 climb that mountain again. You're kind of already on the plateau, right? And
5:37 you can climb up the next level. And you know, I' I've got I'm on the cap table
5:41 of this startup that is doing incredibly well. Uh they did $10 million in revenue
5:46 last year, but it was not recurring. And the thing is talking to these founders
5:50 who are doing an incredible job and then they are building a very successful
5:53 business on this non-recurring revenue, but the stress and the anxiety that
5:57 causes them, right? 1st of January, we're at zero. We have got to climb that
6:01 mountain again and not only that but climb higher because of course investors
6:06 everybody expects growth and so even though the business is going well that
6:11 that discomfort uh of of not knowing oh well you know most of my revenue is
6:14 going to roll over from month to month from year to year is is really it's not
6:18 great and you know they're trying to not pivot but they're trying to pivot their
6:21 pricing and their business model into more of a recurring revenue one for
6:25 exactly that reason. everyone is and it's you know I that's what I wanted the
6:30 moment I started I had a a software company it was just me solo in 200 let's
6:36 say maybe six seven eight and when the 2008 financial crisis happened I lost
6:43 80% of my revenue overnight and it wasn't recurring it was it was a
6:47 one-time sale of software product and so I had 20% of the revenue the next month
6:51 and the month after and the month after so I had to just made these huge changes
6:55 now luckily It was a small company because it was just me. But um that's
7:00 where having recurring revenue makes you doesn't make you recession proof but a
7:04 little bit more market resistant. You know, you you lose you can draw down a
7:07 bit but uh >> it dampens those swings, doesn't it? So yes, if you have a major if you have a
7:14 2008 happens again, you won't lose 80% from one month to the next. You might
7:18 lose 20% one month and another 20% the next month, but it sort of smooths
7:21 things out and probably means that the bottom is not as deep because some folks
7:25 just won't cancel. Uh they're not in a position to cancel. >> The other thing that you mentioned, Rob,
7:29 was the repeatability, the sense that there is less luck necessary in building
7:36 a B2B SAS than perhaps other sectors like a marketplace or BTOC. Why in in
7:41 your opinion and in your experience because of course you have seen hundreds
7:46 thousands of these businesses why is it less reliant on luck than these other spaces?
7:50 I mean it depends on which space we're talking about, right? If we go into like
7:54 medtech or biotech where they are truly trying to discover a new something or
7:58 invent a new thing that is you and I just both know that's really hard to do,
8:04 right? So SAS building a a project management app for park rangers is just
8:09 not that it's not that hard to do, right? The risk there is more of the
8:13 market risk like will anyone pay for it? Is there actually a need? But then
8:17 there's other segments where you know think about starting any social network
8:23 for example from 2003 to tw to now right even Tik Tok came out what a couple
8:27 years ago those think of the amount of luck and how many things have to line up
8:33 for those to take off because there's no revenue for now because you need to get
8:36 to hundreds of millions of daily activives before you really get any
8:40 revenue. So you have to push that off. You don't have any money to actually
8:43 market the thing. So it's all virality. It's like great, I hope we go viral. And
8:47 you can engineer virality, but it's not as repeatable as anyone would would
8:50 believe. It's not as nearly as repeatable as cold email outreach, SEO,
8:55 content, integrations, events. You know, that that's the thing is B2B SAS since
8:59 businesses are paying for it. It's like how do you reach your customers? And
9:03 it's, you know, it's pretty um there's only 20 B2B SAS marketing approaches and
9:06 you you kind of go down the list. >> So that there is more of a playbook, I
9:11 guess you would say. >> Wow. Oh my goodness. It's nearly like I
9:14 set that up. Folks, if you are planning to build a SAS, you should read Rob's
9:18 book, The SAS Playbook. Uh, and it will go in a lot more detail than we can
9:23 cover in a podcast episode about all of these things. So, yeah, you've mentioned
9:27 the fact that because you're marketing to businesses, you can get a higher
9:31 ticket value. There can be more mature business relationships, right, about
9:35 pricing and that sort of thing. That's one side of it. The other thing and I
9:39 know you talk about this is about the ability to define a niche and go for it
9:45 and perhaps the ways in which that is easier in B2B that you have a more
9:51 well-defined ideal customer profile and I know you've you've mentioned vertical
9:55 SAS horizontal SAS and a term which I think you may have invented orthogonal
10:00 SAS maybe talk about all of these and the ways in which these are part of what
10:05 makes B2B SAS more predictable perhaps than other business models. Yeah, I like
10:07 that way of thinking about it. Horizontal SAS obviously or any
10:12 horizontal business is where kind of any business could use it, right? So,
10:16 whatever sector you're in, if you're a construction firm or if you're a tech
10:20 firm or if you're a consultant, you could potentially use this. That's
10:23 horizontal. Vertical is when you focus on a usually a specific niche or
10:27 industry. It's like I am a CRM for construction firms or a CRM for
10:32 municipalities or, you know, some very specific vertical. And then
10:36 orthogonally, you're right. I did I coined it because I I was looking across
10:41 all the companies. I'm invested in 224 SAS companies and I was trying to
10:46 categorize them vertical and horizontal. And what I realized is there was this
10:49 third category that is kind of a weird mix of both. It's where they're
10:53 horizontal technically horizontal in the sense of they can be used by any company
10:58 in any vertical, but they target a specific role or title at a particular
11:01 company. And that gives them an advantage in their marketing because
11:05 then they know HR directors buy my software even though any company any HR
11:10 director at any company can do it. Um you know then that if you're going to do
11:13 cold outreach you're looking for HR directors. If you're going to go to
11:16 inerson events then you're going to go to events where HR directors are. And
11:20 the idea of knowing who your customer is and where they are is such a huge
11:25 advantage in any B2B SAS. And this is why horizontal SAS is harder. we
11:29 actually see headwinds you know across our again 224 investments it's a minor
11:33 definitely a minority that are horizontal and I don't know the exact
11:37 number if it's 20% is horizontal or 25 or 30 it's probably somewhere in that in
11:41 those numbers and they do see headwinds at a certain point they usually grow
11:45 pretty well and then they hit about depends if it's half a million or 2
11:48 million somewhere in that range and they they start releasing headwinds because
11:51 the horizontal players have these big funded competitors uh that they're comp
11:55 competing against whereas vertical and orthogonal can still have that. Vertical has it
12:02 less. We see lower churn often net negative if you really get traction as a
12:05 vertical SAS. Nothing's easy. It's all hard and complicated, but to me it's
12:09 just easier than a lot of the other models that I've either tried myself or
12:13 or see other entrepreneurs that I know trying. >> Yeah. It's funny when I talk about
12:16 things being easy. This is a little bit of a sidebar, Rob, but quite often when
12:20 I say, "Oh, this is easy. This is an easy business model." And people can get
12:22 a little bit defensive. They're like, "This is not easy. This is really,
12:25 really hard." And it's like, yes, it's it's all really really hard. And and
12:28 what I think about is when you're watching the Olympics, in gymnastics or
12:32 in diving especially, they have this concept of degree of difficulty, right?
12:35 These are Olympic divers. All of these dives are really hard to pull off, but
12:39 some of them have a higher degree of difficulty than others. And you probably
12:43 want to calibrate to a degree of difficulty that makes sense for your
12:47 business. And I guess what you're saying here really in this context is that B2B
12:51 SAS, it's not easy, but the degree of difficulty of that compared to
12:56 bootstrapping a two-sided marketplace is quite a lot lower. And when you're doing
12:59 something really difficult and with a quite high likelihood of failure anyway,
13:05 consider whether you have the intention, the skills, the experience, the
13:08 patience, the risk tolerance to do something with a high degree of
13:11 difficulty. One of the things you talk about, Rob, is often this is a very
13:14 personal thing for a founder. what sort of business are you trying to build? uh
13:18 because you know you're famous for B2B SAS you're also famous for bootstrapping
13:24 or near bootstrapping companies and I think those two things they are they are
13:27 very closely connected to each other right the predictability of B2B SAS and
13:32 the ability to bootstrap it these are not separate things right if you want to
13:34 build a marketplace if you want to build in TOC you much more likely to need to
13:39 raise significant amounts of venture capital to do it why because the degree
13:43 of difficulty is higher so you start to see how all of these dots connect
13:47 >> that's Right. And that was another one. There were like three other points about
13:51 why SAS is so great. And one of them I was going to say is you don't need
13:54 funding to do it like you do with some other business models. Funding can help.
13:58 I mean, I'm a freaking investor in 220 SAS companies. So, I I believe in
14:02 funding as an accelerant and venture capital going into SAS is is great if if
14:06 you're using it to fuel fire, if you got the unit economics worked out. Um, but
14:11 you don't need it. You can bootstrap it and and the reason really is what is
14:16 because it's high margin, right? So SAS at scale you can see 80 to 90% gross
14:21 margins and 30 to 50% net margins on SAS. There are companies like Zoom and I
14:26 don't remember if it's GitLab but Zoom definitely went public with more money
14:31 in the bank than they had raised from from investors. They had more money in
14:35 the bank than because it's that profitable. There are several companies
14:37 actually mentioned that >> I believe at Lassian was the same.
14:40 >> Yep. High margin which means you don't need funding. Again I'm not saying you
14:44 shouldn't. I'm not anti-funding at all, but like that you have options at least.
14:48 Um, and all of that leads to the highest exit multiples I know of. Friends of
14:54 mine run e-commerce websites and they will get two times EBIDA for an ecom
15:00 website or one times revenue or you know some a consultant a guy who running an
15:04 agency sold it for like one times you know trailing recurring revenue and with
15:12 SAS I mean yeah have I seen 7x topline revenue exits have I seen a 15x topline
15:17 revenue exit for a SAS doing $3 million? So yes, that's a $45 million purchase
15:21 price for a SAS 3 million era. I have and that's because of all the things
15:25 we've just said. It all rolls up to the if you're trying to change your life and
15:29 if you have investors like pay investors back the enterprise value you create
15:34 with every dollar of MR is unmatched. I don't know of another space where it's
15:36 that predictable >> and that comes down I think to a couple
15:39 of things Rob. One is like you mentioned the quality of revenue. So high margins
15:45 and also recurring revenues. So if if I'm a private equity buyer looking to
15:49 add this thing to my portfolio, I can model this out. I can say the margins
15:54 are really high and the revenue is relatively predictable and so I'm
15:58 willing to pay more for the that annuity and that growth potential. The other
16:03 thing I'd mention though is that part of all SAS, not just B2B SAS, is that it is
16:09 pure software. Now you mentioned ecom, my last startup was a leasing platform
16:16 for consumer electronics. So ecom plus fintech on steroids. And one of those
16:20 cliches that really hit home to me is that atoms are harder than bits, right?
16:27 where if one of the advantages of tech of software is it is generally high
16:32 margin and it it is generally very easy to scale especially these days of course
16:38 uh with all cloud elastic everything you can scale from one customer to a million
16:42 customers and of course you will have technical challenges along the way but
16:46 nothing that you can't manage with the right team whereas if you're in ecom
16:52 then you have logistics and inventory and procurement and also all the cash
16:56 that gets tied up in that. Uh if you are in hardware or biotech, you have all of
17:02 these clinical trials or or manufacturing and all of those issues
17:05 and all of those drag you down. Building a billion-doll business in consumer
17:10 packaged goods is something that takes decades and only a couple of companies
17:14 ever do it. Whereas in software, it can happen in a matter of years and is done
17:18 relatively frequently. It is this incredible business cash machine. All
17:24 pure software has this advantage of scalability and margin. Although AI
17:28 might be eating into the margin side of things and we we'll come back to that in
17:31 a little bit, Rob, but I emphasize this because I came to it late as a founder
17:35 and as a practitioner, Rob, coming from the product and the technology side,
17:40 understanding the the business, the way the finance and the cash interacts with
17:44 what you're building is actually really important. Even if what your passion is
17:47 is building. >> That's right. And I stumbled upon it because I was a software engineer. Um, I
17:52 taught myself to code when I was I learned how from a book when I was 8
17:56 years old. And so I got out of college with an electrical engineering degree
18:01 and um I wor for 18 months. I was going to become a professional engineer and
18:05 realized man I just code is code is cool. I really enjoy it. I enjoy making
18:09 stuff. I'm a maker you know and so yeah I started building software for other
18:13 people. I was a consultant and then W2 and uh so I looked into it you know and
18:17 I ran an e-commerce website. We didn't manufacture. It was more drop shipping.
18:21 I've sold information products and courses and books and kind of run the
18:25 gamut. I haven't done a hardware startup. I've never done manufacturing
18:28 or anything like that, but uh and I've done one-time software and I've done SAS
18:32 and now I run a venture fund and a community and so I've done a lot of
18:35 different things and yeah, when I look at SAS it's like there's a reason I can
18:39 kind of say I think it's the best in the world best one that I know about. So I
18:41 Challenges and Advantages of Vertical SaaS
18:43 actually wanted to go back a couple of steps and explore one of the things that
18:47 you mentioned Rob which was the horizontal vertical orthogonal and you
18:52 know why vertical and orthogonal SAS tend to outperform horizontal which is
18:58 the ability to identify and reach your niche customer and I I actually wanted
19:03 to ask you to contrast that with B to C for a moment because I think there's an
19:07 interesting thing to maybe deconlate here which is no matter what your
19:10 startup is the advice and it's The advice we share on this podcast very
19:15 often is to niche down to start with something narrow, solve a problem for a
19:21 specific group of people and expand from there, right? And so the advantage of
19:25 doing a vertical SAS or an orthogonal SAS is you say these are the people I'm
19:29 solving the problem for. And perhaps over time you can solve it for other
19:33 people too, but it's very clear you've identified those people. And horizontal
19:36 SAS, you've not done that. And so that's why it's more of a money pit. You're
19:39 competing with more people. the people you're competing with are better funded.
19:44 It is harder to reach those customers because you're not speaking as directly
19:48 to them. But I wanted to ask why is that a B2B thing specifically? Couldn't you
19:53 say the same in B toc or in marketplaces to say, hey, you know, we're a social
19:58 network for tall people living in Minnesota or whatever it is, right? And
20:03 say, you know, we have a very tight ICP. We are building for that ICP and we're
20:08 going to reach those people. Why is it that this tends to work better in B2B
20:13 than it does in these other software business models? >> Two reasons come to mind. I've never
20:16 thought about it, but two reasons come to mind off the top of my head. One is
20:23 that in BTOC you can't charge very much if at all and it's purely ad based or
20:29 you can charge 50 bucks a year, $4 a month. You know, I mean the B TOC SAS I
20:34 see $100 a year. So as a result, you need incredible scale. As you niche down
20:38 in B TOC, you don't have the scale to make money in ads or in $5 a month
20:43 subscriptions. In B2B, I mean, the annual contract values of a lot of
20:48 companies I've invested in are 5,000 to $250,000. You don't need that many customers to
20:52 build an incredible business, you know, to build a $5 million ARR business. And
20:57 these are bootstrappers, right? So, this small, we could talk about ventureback
21:00 to hit a hundred million ARR. Um, you know, you don't need that many. That's
21:04 the first thing that comes to mind. And then the the second thing is that with
21:10 B2B since the contract values are high, it's not just about them finding you.
21:14 You can do outreach for examp like cold cold outreach, cold DMs, LinkedIn and
21:19 email because you can afford to with a high annual contract value. And with
21:23 horizontal, it's like you can't really do cold outreach, right? Because it's
21:28 everybody. But if you have vertical or orthogonal, then you know who to
21:31 outreach to. If it's B toC and I'm looking for as you said tall people in
21:36 Minnesota how would you do cold hour you don't have the money to do cold like it
21:39 doesn't matter that you know who your who your customers are all you can do is
21:44 well I I guess I can build the product better for them and I can target my
21:47 positioning towards them my marketing and oh whatever else I'm doing to draw
21:51 me in but there's less of a lever I think for BTOC in terms of finding your
21:55 prospects >> I had one other hypothesis I wanted to run past you which is The ability to
22:03 meaningfully hang your hat on relatively minor differentiation is much easier in B2B.
22:10 And what I mean by that is if you are making a CRM for lawn care businesses
22:17 versus a CRM for distributors in the warehouse space, 95% of the features
22:23 will probably be the same. But those things that really matter to lawn care
22:27 businesses or to warehouse distributors, they will care about those things a lot
22:32 and so they will choose your product over a horizontal alternative and they
22:36 will pay more for that. Whereas in B TOC, right, you know, perhaps there are
22:40 a couple of features of a social network for tall people in Minnesota that you
22:43 might quite value as a tall person in Minnesota, but you know what?
22:46 Instagram's pretty good, too. Facebook's pretty good, too. Are you really going
22:50 to preferentially choose this new product for these relatively small
22:54 tweaks that are ideal for tall people? I think it's harder to do that compared to
22:59 >> Yeah, I hadn't thought of that, but that sounds right. Lines up with my
23:02 experience. >> Well, I mean, you know, I think one of the questions for someone who who's not
23:08 in this space is how many CRM do you really need? Why do we need to have a
23:13 CRM for every industry niche? Why do we need to have a practice management tool
23:17 for every different type of practice? And I'd be interested in your thoughts
23:20 on that. I've sort of sketched that out at a very high level framework, but what
23:26 is it that makes vertical B2B SAS actually a really viable way of of
23:32 working and a way of winning when you're competing against someone like HubSpot,
23:35 for example, if you're building a CRM, who is going to have way more features
23:37 than you. It's going to be more polished. So, why does your little rinky
23:43 dink twoperson bootstrapped startup that is making CRM for lawn care businesses,
23:47 why do you win against HubSpot? >> That's a really good question. The funny
23:51 thing is is if you want to go big, venture capital flows into horizontal
23:55 markets. Not not solely, but a lot of it does. And so, HubSpot's a great example.
24:00 That's a horizontal CRM, Salesforce, and they're huge, huge companies. But for
24:05 the hundreds or thousands of founders who don't want to raise venture, maybe
24:08 they raise small amounts of money, maybe there is 100 grand or 500 grand, you're
24:13 not going to have too much luck competing against HubSpot these days.
24:17 And so the idea is that if you focus on a vertical, as you've said, you can
24:24 build the handful of features that they want different. It's a highly configured
24:29 version of HubSpot that is focused on lawn care providers and they don't want
24:34 to try to figure it out and pay um consultants to get their HubSpot set up.
24:38 It's complicated. So you can add services as a vertical SAS of like hey
24:42 let's get you on board and let's get you trained up on it. You have a tool that
24:46 speaks to their language and uses all the terminology that they want to speak.
24:49 In addition, you know who your market and your ICP is. So all your marketing
24:54 is less expensive uh your marketing and sales because you're only talking to
24:58 longare people and not all the the other verticals. In addition, you might just
25:04 be able to charge more if you're niched down as a vertical. You can often charge
25:08 a premium which which is counterintuitive. If you had Microsoft
25:12 Word, in fact, there there one exists. It's it's basically a word processor
25:16 used for I think it's like biblical languages and it's called noten or
25:19 something like that. I'm mispronouncing it. But like so you can do Hebrew in it
25:23 really well >> ancient Greek all this stuff right that
25:27 word doesn't handle so well I had this example I think in my first book but
25:30 it's like Microsoft Word was charging $75 you know this is before SAS but it
25:35 $75 for a license and not ben was $895 for a license it was the same thing but
25:40 it just did this thing well right so that's the that's the idea too is you
25:42 can have some pricing power and then the retention is is also quite good you can
25:46 have that negative turn easier >> right and you know put yourself in the
25:50 shoes of a business operator. They don't really want to be in the business of
25:55 configuring and working out how to make their SAS work. They're just trying to
25:58 run their business. The SAS is there to help them. And so, if they are able to
26:02 use a tool that that kind of gets them, right, I'll stick with my lawn care
26:06 example. If my CRM understands all the little quirks of running a lawn care
26:10 business that nobody outside of the industry really understands, but man,
26:14 they hurt me, you know, the the seasonality, I don't know, whatever it
26:17 is. I'm not a lawn care guy. and then I can just install that CRM and it gets me
26:21 and I can get back to running my bloody business. Well, that's what I want. I'm
26:24 willing to pay extra for that. Especially if it saves me hours or or
26:28 you know, potentially weeks at the point of onboarding because it just works for
26:32 me straight away. That's worth paying for. >> I mean, throughout my career, I've paid
26:37 for software that was more expensive but just removed the headache because it
26:41 understood me better as the ICP, whatever that was. And I think there's a
26:45 lot of folks out there that want that. >> That's right. And you know, a tool like
26:49 HubSpot is is quite overwhelming, Rob. Like I'm going through this at the
26:53 moment with our business. We want a CRM. I I think we are actually going to go
26:56 with HubSpot, but it was a very close call. And the reason is because it can
27:00 do everything. You don't really know how to make it do the thing that you want it
27:04 to do. It It's not quite the right analogy, but it's close enough. You
27:07 know, the difference between a Swiss Army knife and just one really good
27:11 knife that is designed for the task that you are actually trying to perform. You
27:16 don't want all of these things. So, it's it's really valuable. Um, I know we're
27:20 going back and forth a little bit, but something else that I I wanted to talk
27:22 about, you were talking about these fantastic exit multiples, but when we
27:27 were talking about vertical SAS, another thing that occurred to me is probably
27:31 more exit avenues. And I'd love to ask you about this because you would know
27:35 better than me. But if I think about some of these rollup companies like
27:39 Constellation or or private equity firms and so on, vertical SAS is very
27:43 attractive to them not just because of the quality of revenue but because of
27:47 the ways they can often get synergies and you know bundle them up into suites
27:51 and and get more pricing power that way. So if I think about when you're building
27:56 a startup and eventually you want to get to an exit, obviously you've got the
28:01 giant unicorn IPO, the stuff that gets in the news, you have strategic
28:04 acquisitions and so on, but one of the challenges often when you have a
28:09 reasonably successful startup and you want to get to the exit is finding a
28:13 buyer. And so I wonder if if it is the case and if you can enlighten us a bit
28:18 more about is it the case that B2B SAS has a wider range of buyers potentially
28:22 and what have you seen in terms of getting these businesses to exit? As far
28:28 as I know, yes, B2B SAS is is is a category that is the predictability is
28:32 huge. As we've said, quality of revenue, the recurring usually the growth is is
28:36 better because if if you're compounding uh each month, but I can look at almost
28:44 any B2B SAS company and I instantly know based on the churn and the lifetime
28:47 value, the average revenue per account, the price points, the this and that, I
28:51 get a sense of the quality of the business and and private X. So, it's
28:54 very um numbers driven, right? And private equity is very numbers driven.
28:59 But let's say that you and I were to look at Uber's financials in the early
29:03 days versus Facebook's financials in the early days versus like Google's versus I
29:08 don't know Door Dash, you know, let's just pick a bunch of successful
29:11 startups, but they're not SAS. It's like how do I even I don't know if that's a
29:14 healthy business. Like I it's a two-sided market. There's a lot of
29:18 complexity. It's like there's no uniformity maybe is a way to say it. And
29:23 so if I'm private equity and I want to get I want to get good at buying
29:25 startups for example, it's like can you get good at buying two set of
29:29 marketplaces plus social networks plus biotech plus medtech plus you know all
29:31 these other things or is it's like boy if I focus on B2B SAS you get good at it
29:35 pretty quick you know or at least you understand the businesses pretty quick.
29:38 So I think there's something about the um uniformity is the only thing that
29:43 comes to mind. So I'm a partner at discretion capital which is a sellside
29:49 M&A firm that works with SAS businesses B2B from 2 million to 20 million uh ARR
29:55 right and somewhere around and don't quote me on this but it's somewhere in
29:59 this neighborhood around 70 to 80% of exits kind of in that uh not just
30:03 through discretion but from all the stuff we know about 70 to 80% go to
30:09 private equity and 20ish 30% go to strategics And that tells you a lot that
30:15 private equity is providing liquidity in this market and there must be a reason
30:19 for it. And a big part of it is, you know, net negative churn and that the e
30:23 the unit economics of these businesses are are pretty incredible.
30:28 >> Actually, just as a as a side note for folks who are less experienced listening
30:32 to this, when you talk about ne negative churn, can you give a little bit of a
30:35 definition and also how it happens? >> Yeah, thanks for calling that out.
30:39 Sometimes I get a little deep in in these terms. Um, so obviously churn is
30:43 I'm going to say revenue ch. There's revenue churn and logo churn, but
30:47 revenue churn is is the the dollar amounts of MR that churned in a given
30:50 month divided by the MR at the beginning of that month. Right? So if I start the
30:56 month with $10,000 of MR and $1,000 worth of MR churned that month, then I'm
31:01 going to have 10% uh of revenue turn that month, which is catastrophic
31:04 business on fire. That's that's gross churn. That's just people who canceled.
31:09 But there's also contraction revenue which is existing customers maybe
31:14 getting a little smaller using lower usage less this month than I did last
31:16 month. That's >> contraction. >> And then then there's expansion revenue.
31:21 Like you can imagine I add more email subscribers to my Mailchimp account and
31:26 I bump up from a $25 account to $50 account. Well, imagine there's like a
31:29 million customers that are doing that, you know. So that's called expansion
31:33 revenue. So when your expansion revenue uh is greater than your basically your
31:38 gross churn and your contraction revenue, you can literally add zero
31:44 customers in a month and revenue will grow. So if you had a 3% net negative
31:50 churn, 3% negative churn, your business grows 3% in a month if you added zero
31:53 new customers. >> Exactly. And and one of the metrics I've
31:54 The Importance of Net Revenue Retention
31:58 heard used more more frequently to to measure the same thing these days is net
32:02 revenue retention. It's exactly the same thing as what you're talking about,
32:05 which is to say if the customers you had last month paid you $100 and this month
32:11 they paid you $103. Well, you have got 103% net revenue retention. So that's
32:16 you're making $103 from your existing customers, not your new ones. If it's
32:20 over 100%, then that is what you are calling negative churn, Rob. So yeah,
32:25 and that's another really powerful thing about any recurring revenue model is it
32:29 can actually grow over time. If you're operating a business well, it should
32:32 grow over time. >> Yeah. And it's hard. Not every not every
32:39 SAS, in fact, most SAS don't have net negative terms. Um it is a minority of
32:43 companies that I see. Um, and usually it's where you have your pricing just
32:49 right and it is based on a value metric like oh it's number of seats at a CRM or
32:52 it's number of subscribers in an ESP. It's like you really got it dialed in
32:57 and uh it's and also have incredibly low churn and really high product market fit
33:01 meaning a lot of new people come in almost nobody churns and you have that
33:05 that expansion but I have seen >> yeah 12% net negative churn which is
33:10 just bananas when you think about it. This is one of those business metrics
33:14 that more sophisticated investors and more sophisticated operators look at
33:18 very deeply because it is it correlates pretty strongly, Rob, with the quality
33:22 of the business, doesn't it? Like if you have got high net negative churn, high
33:26 net revenue retention, then that is a sign that your business is incredible,
33:30 that you've got really great product market fit, and that you're just going
33:33 to be able to keep compounding your growth year after year. So, this is a a
33:37 key metric. I I would say if you're a recurring revenue business at all, as
33:42 much as tracking your MR or ARR, tracking this the the NRR, the net
33:46 revenue retention or the net negative churn, that should be one of your real
33:51 key northstar metrics that you look at month after month. >> Yeah. Any buyer who is going to acquire
33:56 you, especially private equity, but even in strategic is going to look at that.
34:01 And I have seen deals where where the net revenue retention is high and you
34:06 know net negative turn is is amazing and they'll get 10x ARR and I'll see a very
34:12 similar business where instead of net negative turn they have 6% monthly churn
34:16 and they'll get three 3x AR. So it's like three versus 10. It's incredible.
34:20 >> Part of the reason is this is a an exponential curve. This is compounding
34:23 in the same way that you get the magic of compounding interest. You get the
34:28 magic compounding of net negative churn. Conversely, you get the problematic
34:32 compounding of high churn, which is why a high churn business is catastrophic,
34:36 Rob. People are like, "Why is it so bad?" It's because it's like your
34:40 revenue is going down month after month, and you have to fight that gravity
34:43 harder and harder. Whereas, if you have negative churn, your revenue is just
34:47 going up month after month, and then you can layer on new customers on top of
34:51 that, and imagine compounding that for a few years. That's an amazing business.
34:54 And it's also a very reliable and predictable business. And so that's a
34:57 really great place to be. All the best.coms are taken or crazily
35:01 overpriced. So you settle on a workaround domain for your website.
35:05 Don't compromise. Get a clean sharp.te domain that instantly says this is a
35:12 tech startup. Grab yours at get.te/tsp for the startup podcast or via domain
35:16 registars like GoDaddy. Hey, before we get back into the show, a quick
35:19 reminder. Please don't forget to subscribe, like, and turn on
35:23 notifications and spread the word about the podcast. It helps us grow the show,
35:26 which in turn helps us help more founders like you. Thanks so much. Okay,
35:31 with that, back to the show. So, Robert, let's let's get a little bit more
35:34 tactical now and and maybe this is a nice time to to mention your book again
35:38 and and folks should read this, but if you want to build a vertical or
35:42 orthogonal B2B SAS, you're listening to this, you're like, "Yeah, Rob's speaking
35:45 my language and it all makes sense to me." How would you recommend people
35:50 approach starting a business like this? >> So, there's no one way to do it, right?
35:53 There's no blueprint. There's a reason this book is not called the SAS
35:56 blueprint. I intentionally called it a playbook because a playbook has things
36:01 that you can do and if you do them well and and everything falls into place, you
36:04 get a little lucky, you know, the the score takes care of itself and you win.
36:08 Blueprints, uh, anyone trying to sell you a blueprint to do a startup, uh, it
36:13 just doesn't exist, right? So there are different schools of thought on this and
36:17 I can tell you kind of the one that I've landed in but the basics of uh the early
36:25 stages of doing it are to find a problem that a business needs solving and try to
36:32 validate that the problem is dire enough that it's painful enough that they are
36:36 actually willing to pay to solve it and then figure out a solution to that. It's
36:41 called finding problem solution fit right everyone talks about product
36:44 market but before that is problem solution sometimes that solution can be
36:48 done with a bunch of virtual assistants in the Philippines collating excel
36:52 spreadsheets or it can be done with some noode automations or make and air table
36:56 The 2-20-200 Framework for SaaS Validation
36:57 or whatever. I have this framework called the 220 200 framework. The two is
37:04 any SAS idea potential youth could throw out to me. I could do about two hours of
37:09 research online and get an idea of if there's demand, if there are existing
37:13 competitors. Um, I would do keyword research in in Google. I would look on
37:17 YouTube. I would look at G2 and and um Capterra and uh probably look for
37:23 competitors. I'd go on Reddit and see what people are complaining about. Kind
37:25 of get an idea of, oh, there might be room in this space for me. Right? That's
37:28 the two hours. So I could take 10 different business ideas and and go
37:31 through all of those and get an idea of like of those 10 there's three where
37:36 there's a lot of demand. Oh there's a lot of competitors too but then I see an
37:40 angle bunch of complaints on capterara specifically competitors not doing this
37:43 right. So that's just in again 2 hours. It's directionally correct, right? It's
37:47 not exactly 2 hours, but the 20. So again, it's 2200. The 20 hours is where
37:54 I set up a landing page with a value proposition that I discovered in my
37:59 research of like tired of your CRM always being really complicated. Uh I
38:04 have the CRM for you. You would write way better copy than this, but you get
38:06 the idea. you know, you put a headline, you put an H1 and an H2 email capture or
38:10 maybe it's a booking link, maybe it's a direct a Savvy Cow link that can go
38:13 right to my calendar and I want to have conversations. So, you have a landing
38:16 page that you start talking about it in podcast, social media, your audience,
38:21 whatever you have. And part of the 20 is not just landing page smoke test because
38:25 that goes back to 20 to 4-hour work week years ago, but you also start doing some
38:30 cold outreach on LinkedIn or to your network, warm network of like who in
38:34 this space could potentially use this, right? Right? So you do about 20 hours
38:37 of research kind of like trying to validate your assumptions that you've
38:41 started to come up with you know in the two and then the 200 hour which with AI
38:49 may now be 40hour but 22200 the 200 is where you build you build an MVP
38:51 basically and you try to get something out there into the hands of folks. Um
38:55 that's not always possible to do in 200 hours but you get the idea. I like to
39:00 see these stages because in my world there's a lot of developers building a
39:03 lot of [ __ ] in their basement for 6 months that nobody wants because they
39:07 didn't spend two freaking hours in a key in Hrefs realizing there's no uh
39:12 interest in this. There's no demand and they don't want to do cold outreach cuz
39:15 they're a developer and they're either scared of it or they don't believe in it
39:18 or whatever because it's sales and it's like well what are you doing with your
39:21 life man? Like do something smarter, you know? So is 200 is 22200 like the only
39:27 way to do it? No. Is it uh a way that I've seen a lot of folks do it and and
39:30 succeed? Yeah. Do I think it increases your chance of of success and reducing
39:34 the chance of just huge misses that you're going to have where you spend six
39:38 months building? I I do. I think it's a a reasonable mental framework.
39:41 >> Yeah, I think so. And this is you know this classic lean startup, right? But as
39:45 as applied to B2B and you know unless you are coming from the exact niche that
39:50 you're building for and even then Rob like you don't want to be overconfident
39:53 in in knowing the shape of the solution and the the language that speaks to it
39:57 but we talked before about why is vertical SAS so powerful and able to
40:02 compete with horizontal competitors is because of that sense of when I use this
40:07 product I'm like ah they get me they get the challenges of running a lawn care
40:11 business and so you you damn well better get it right you You really need to do
40:15 your research and understand what are the challenges that a lawn care business
40:18 faces. Otherwise, you will not be able to cut through to them and you will not
40:23 be able to fulfill your promise that you are building a tool that is just for
40:25 them. >> Yep. That's right. And then I So, my last SAS uh app that I built was called
40:32 Drip and I sold it in in 2016. I was the original target market for it. I was it
40:36 was supposed to be kind of email marketing, you know, for SAS founders
40:39 but really for SAS companies. But I quickly realized that SAS is fine, but
40:44 they don't have the biggest email lists. You know who does? Infoarketers,
40:47 bloggers, like that's where the real money is. And so I had to quickly learn
40:51 I didn't really know how information marketers and and course creators.
40:55 They're called kind of creators in the maker and creator space these days, but
40:58 that wasn't a term back then because we started in 2012. Um, I had to learn how
41:02 they were using it, how they talked about it. And so I quickly became kind
41:06 of one of the ICPS, but I was a subset. like our revenue quickly was outpaced
41:10 our SA, you know, our revenue from SAS companies was quickly outpaced by
41:14 bloggers and kind of course creators. Um, so even if you think you're building
41:19 it for yourself, you often have to learn how other users are going to use it if
41:24 you want to probably do right by the business, so to speak. >> Now, Rob, we need to talk about the
41:29 elephant in the room of every tech discussion these days, and that elephant
41:34 is called AI. I've been seeing a lot of posts on LinkedIn lately that say, you
41:39 know what, B2B SAS is dead, especially niche vertical B2B SAS is dead. Why?
41:45 Because I can just spin up my own custom tool that using AI is not only built for
41:51 lawn care businesses. No, it's built for my lawn care business. It is bespoke. It
41:56 is perfectly tuned to my needs. And why would I go pay someone a high recurring
42:02 fee to use their software when I can just build and operate my own? So, is
42:06 B2B SAS dead? Should we just delete the whole first half of this conversation?
42:11 >> It's all gone. Yep. I'm I'm all in on that boat. Oh, I love the is dead stuff.
42:16 Remember when SEO was dead 15 years ago? I mean, this is just such a classic
42:19 thing. Here's the thing. >> People want to go to extremes and it's
42:25 like SAS is everything. SAS is dead. These are our two extremes. Guess what?
42:29 It's somewhere in between. No code did the same thing. Remember when no code
42:32 started coming up and it's like, well, there's not going to be B2B SAS anymore
42:35 because everyone's going to no code their own thing. And it's like it's not
42:40 that it had no effect because I know for a fact like within my companies, Tiny
42:45 Seed and Micromp. We have three maybe four completely custom line of business
42:49 apps that are basically built in Air Table and they're phenomenal. We used to
42:53 pay 10 grand a year, 15 grand a year for these apps. So, it definitely has
42:56 impacted, but if I were to say the entire SAS market, how much did no code
43:00 kind of chomp out of custom verticals? Maybe 20%, 15%. It had an impact, but
43:05 it's not going to decimate it. I think AI is the same thing. My my brother runs
43:08 a construction firm in the Bay Area. He's very smart. He used to be a
43:11 developer, actually. He's not going to build his own SAS. He doesn't know how
43:15 to do it. It's going to be buggy. It's going to get hacked. He doesn't want to
43:18 host it. You know what I mean? There's there's all these things of Yes, we can
43:22 get a little more bespoke. uh and even as AI gets better over the next several
43:27 years just stuff that is customuilt by builders who are doing it uh who are
43:33 doing it for as a craft let's not let's not Dunning Krueger affect this you know
43:36 so like this is the problem right he's like well SAS is so easy and AI can just
43:40 do it for us and it's like yeah so and marketing is too right and sales and
43:44 running a company AI can just I'm just going to do my founder AI agent then
43:48 have it run my company because isn't this easy it it's a similar thing where
43:52 it's like the complexity of SAS is not not something that uh AI is going to
43:56 solve anytime soon. Now, I'm not saying, you know, in 20 years, yeah, maybe. I
44:00 mean, that's so far out, though. Two, three, five years where we're going to
44:04 have vertical sess for sure. >> And I completely agree, of course. And I
44:10 I think a lot of this AI hype comes from people who do not understand what the
44:14 process of building and maintaining software actually is. They think writing
44:19 the code is the hard bit, not the business logic. Encapsulating that well,
44:25 having a good user experience that actually saves time, handling edge cases
44:31 and corner cases, managing bugs, security, maintenance, being able to
44:36 make changes, database migrations, all of this stuff, right? That is a real
44:42 tax. And so I think it's it's this curse of you can use a tool like cursor or
44:46 lovable or vzero to build something that looks a hell of a lot like that vertical
44:50 SAS that you've just been paying five grand a year for and you're like I built
44:54 this in one day and it meets my needs and it's like yeah it does now but just
44:59 wait just wait until you want to change something or just wait until you realize
45:03 that your private key has been leaked all over the internet or just wait until
45:09 the database is corrupted. um you have all of these headaches and yes yes yes
45:12 you could solve it yes you could maintain it but now you're in the
45:17 business of maintaining a software tool you you have actually built your own SAS
45:21 well guess what now congratulations it's like you made the baby now you got to
45:26 you got to raise it like that's not the business you wanted to be in cuz I'm
45:29 like oh wait I I like to mow lawns I'm I'm not here to maintain a B2B SAS but
45:35 if you buy too much of this AI hype that's exactly what you'll find yourself
45:39 doing and you I I see this all over the place, even with automations, right?
45:42 People say, "Oh, I'm I'm just empowering my whole team to use Zapia and Nad to
45:47 automate all of our business processes and and I think to myself, okay, so
45:50 you've just empowered a bunch of people who are non-developers to create
45:54 business critical software that you're now running your business on, and you
45:58 don't see the risks of that." It it is a profound misunderstanding of what
46:05 software actually is as a dynamic system as opposed to being a static system. you
46:08 know, it's easy to build that first version, but over time, you'll you'll
46:13 really see that uh it's it's much harder to maintain. And you know, I was
46:17 thinking people talk about vibe coding a lot. There's a reason why vibe
46:21 maintenance is not a term. Uh and that's because AI is really bad at doing that.
46:25 >> Yeah. And the maintenance thing is such a huge deal, right? There there used to
46:28 be numbers back when I was a contract developer, say 20 years ago, where it
46:33 was like the cost to build a project might be half a million dollars if it's
46:36 this big kind of government thing. And the cost to maintain that for the next
46:42 20 years is like $3 million or something like you know it was this huge it was
46:45 it's the iceberg thing where like this tiny little bit above the iceberg but AI
46:50 cannot maintain code in any form or fashion yet. >> So that's some myth busting but that
46:57 doesn't mean that AI doesn't matter. And in fact I I sometimes come across as an
47:02 AI skeptic but I'm not. I will I think AI is incredible. I use it extensively
47:06 in my own work. And so then maybe to flip the question, if you are in the B2B
47:11 SAS space building a B2B SAS, how should you be thinking about AI and how should
47:16 you be using AI both in building and operating your business and within the
47:20 software tool itself? >> Yeah, I have this framework. I think
47:21 Leveraging AI in B2B SaaS
47:26 there's five and only five categories of ways to use AI in B2B SAS specifically.
47:31 Now I think we can even expand this and I think three of them apply to kind of
47:36 any startup even if you're not a software at all like if you're building
47:40 a drone company or something like I think three of the five do. So two of
47:43 them are based based on you being software. Category one is AI as your
47:49 core business. This is where if you removed AI there is no business left. So
47:54 you're either a foundational model Gemini claw GPT5 or AI is at the core of
48:00 the entire product. So a couple examples of these uh I'm actually invested in
48:04 both of these. One is called fiscal.ai and it's uh financial data. It's all AI
48:10 driven augmented and and the interface it's a chat interface. Um the other
48:14 one's called Rosie and I don't know if it's Rosie.AI but it's an AI answering
48:18 service for think of like service businesses. as if you were a
48:21 construction firm. I call your phone number, you're out on a job, Rosie
48:26 answers, and it's an AI voice that says, "Hello, you know, I've scanned their
48:29 website." And Rosie doesn't say that, but um AI is at the core of that entire
48:35 business. It doesn't exist without AI. That's category one. Category two is AI
48:43 as a feature. So if AI went away, like let's say Notion, you know how Notion
48:48 has the AI writing assistant? Notion still exists if we remove the AI writing
48:52 assistant. It's still notion. A AI is a feature. Zoom has AI meeting summaries.
48:57 Zoom still works and exists without AI meeting summaries. Loom has AI titles
49:02 and chapters, but AI is just a feature. And that's a this is probably the most
49:06 common thing we see uh across my investments is, oh, I need to add AI in
49:10 here to make it easier for my users to do things. But AI is not the core
49:17 product. Category three is using AI for building your product. So actually I
49:21 guess three of these are all for software. I m this calculator but this
49:25 is where your developers are using GitHub copilot cursor wind surf
49:32 you know v 0ero and you are using it to help help make you faster. You write
49:36 more unit tests you write better unit tests make generates code quality blah
49:40 blah blah. Now you and I are starting to see have you seen the research? There's
49:44 been a couple early research where it's like AI actually doesn't make you
49:48 faster. Yeah. And that's a whole other conversation, but because I'm I'm I I'm
49:51 not convinced that it's making people faster, but whatever. So, that's the
49:55 third category. Category four is using AI to grow your business. So, this is
49:58 where it's a growth engine. It's marketing and sales, right? So, you
50:02 might use tools like Clay, very heavily AI powered now for outbound, jasper.ai
50:08 or copy.AI for content. You can AI personalize emails, AI powered ad optimization, right? So
50:13 that's category four. So it's really marketing and sales. And the fifth and
50:17 final category is AI for operating the business. And this is where you're using
50:20 it internally. Maybe you're handling support tickets. Uh you're screening
50:24 resumeumés. It's all internal operations. I mean, [ __ ] we had a we
50:28 had a contract the other day because we run inerson events and we were kind of
50:31 talking internally. Does this contract from this hotel or this venue have any
50:35 gotchas? and we just fed it into chat GBT and said does this have any gotchas
50:38 that are non-standard and we didn't you know it's not that we take what it says
50:41 as gospel but we looked through it and we're like this just saved us like a
50:45 round with a lawyer and we felt so much better about it and it it surfaced a
50:48 couple things that we negotiated with the venue and they actually gave in on
50:51 and we said these are non-standard and they agreed with it. So it's like it's
50:54 it's little things like that you know it's just internal operations. So those
50:57 are my five categories. What do you think? Do you think I've missed any off
51:01 the top of your head or any adjustments you make? I mean, do you do you see
51:05 these as a hierarchy from like most AI to least or is it's more like pick and
51:09 choose between these? >> I think it's pick and choose between
51:12 them, you know, and if you're not a software company, maybe it's just the
51:16 last two. Uh, but >> the the AI as your core business is more
51:21 the exception for us in our space. I mean, there's a bunch of hackers doing
51:24 it, but like we are not funding that many AI as as the core business because
51:28 those ones tend to have really high churn and they have a ton of platform
51:32 risk. Um, so we have tended to fund where AI is a feature. You still have a
51:36 moat without AI. That's actually the business I want to see, right? AI is
51:40 augmentation and it can kind of get you there faster and get your customers
51:43 maybe retained because it can do more stuff, but uh it it really reduces your
51:47 reliance on it. Yeah, I don't think you've missed anything like but I think
51:50 that category too that you just mentioned the AI as a feature I I guess
51:55 I I have a a slightly different lens on it which is AI as an enabler of features
52:01 and and that might sound like I'm I'm playing word games here but what I think
52:07 is that AI when thoughtfully used can actually make features that weren't
52:12 possible before that are real value ads. And so, you know, I I guess I mentioned
52:15 that in the context of like, yeah, notion has notion AI and, you know,
52:19 various products have an AI sidebar or whatever. To me, that is most of the
52:25 time that is lazy poor UX. That is CEOs saying, "We've got to put AI into this
52:28 product." And and you know, product managers saying, "Oh, look, we we've
52:31 added AI." But the real potential here is to rethink what is possible now that
52:36 AI is here. And that doesn't mean it's category one. It doesn't mean that you
52:39 have no product without it. But if you say, "Okay, what does my CRM for lawn
52:44 care look like?" Now that AI is a possibility, right? Can it process large
52:50 amounts of data? Can it basically have a feature where you send in a photo of
52:54 your lawn and it can automatically send out a quote? That wasn't possible
52:57 before, right? That sort of thing. Like incorporate it deeply into your product
53:01 thinking and and consider not, oh, let's add a chatbot. I think chatbots are the
53:05 least interesting part of this when you are building into a product. If I want a
53:09 chatbot, I go to chatgpt or claude. I prefer Claude. Um, but in my product,
53:14 how does AI make the possibilities different? In the same way that 20, 30
53:18 years ago, the web made everything different, right? And before that, the
53:21 personal computer, the product, the actual way the product work changes
53:26 because the underlying infrastructure provides new capabilities. And how you
53:31 surface it, well, that is that is really product thinking. >> I like that. I like that
53:36 differentiation. Both of those fit under my category too. But you have made a
53:42 distinction that I had not thought about until now. And I think you are dead on
53:47 that the ability to actually innovate and not just be like, "Hey, there's a
53:51 text box where they have to type text. Put an AI button." So it types that text
53:54 for them. Right? It's like, come on, man. This is this is not creative
53:57 thinking. eBay does this. I sell stuff on eBay and you know what you know the
54:01 only AI on eBay is my description. Ding. I can like put the title of a book or a
54:05 comic book that I'm selling and I click the AI and it's like this was published
54:09 by Marvel in 1968 and like that's neat. That's a neat parlor trick. I can also
54:13 go to Wikipedia and copy and paste that. >> Yeah. My my horistic or hypothesis
54:17 somewhere in between is that the best AI features are not obviously AI, right?
54:22 Because it's not like a little sprinkly magic emoji button that says use AI to
54:27 to do a parlor trick. It's like it is baked into the workflow that is solving
54:33 the business problem that your user has. >> Uh, and it can be quite subtle and it
54:36 just enables new things. >> Yeah, I'm trying to think of examples of
54:40 that kind. Do you have any off the top of your head? I know they're out there.
54:43 I'm just having trouble thinking of apps that are not completely AI.
54:48 >> I can give you an example. Uh, I I am invested in a company that is, you could
54:53 say it is a a vertical B2B SAS for real estate buyers agents. And one of the
54:59 things it can do is based on photos of a property, it uses AI to extract features
55:04 of the property that are not actually in the listing. And then you can search by
55:06 those things. >> That's neat, >> right? And so it's like it's not saying,
55:09 "Oh, I'm using AI." It's just like now you can search by the style of jour,
55:14 right? >> Yeah, that's a neat one. I'm sure there are obviously like five tiny seed
55:19 founders right now that's screaming at the microphone like, "Why didn't you
55:21 mention it?" But nothing's coming to mind. But I uh I do like that
55:24 differentiation you've made. >> Okay, Rob. So, this is a really good
55:26 Closing Thoughts
55:28 place to stop. I know you do all sorts of things to contribute to the
55:31 community. You have a podcast, Startups for the Rest of Us. You have a book you
55:36 can you can show it now once again. So, the SAS playbook by Rob Walling. If you
55:40 are building SAS, you should read this book available where all good books are
55:44 sold. Um you have Microconf and Tiny Seed, which are I believe your your main
55:48 projects. Maybe you can tell us a little bit more about those and also how
55:52 founders can get in touch and and interact and participate in some of the
55:55 communities and things that you do. >> Yeah, Microcom and Tiny Seed are really
55:59 sister companies and it's one ecosystem and it's all for for B2B SAS founders.
56:03 Microcom provides more on the education side. We put out a lot of free content
56:08 on YouTube, youtube.com/microcom. As well as we do onetime fees for
56:13 mastermind matching and we sell a course about early stage B2B SAS, how to get it
56:17 off the ground. The 22200 framework is like one of 27 modules, you know, from
56:22 that course and we do uh inerson events twice a year with microcon and then tiny
56:26 seed is a B2B SAS accelerator where we actually invest. We've raised believe
56:32 it's 5758 million. Uh uh which is which is a lot when you're writing 120
56:38 $150,000 checks. Um and Tiny Seed is an accelerator where we invest in and uh
56:43 and advise founders. And it's also a premium paid coaching program for SAS
56:47 founders above a million ARR. Um and they can come in and be among other
56:52 seven figure SAS peers and get one-on-one mentoring, coaching, and uh
56:57 mastermind groups. So, as I said, it's kind of a big ecosystem of like
57:00 everything over the years. I mean, I've been doing this for almost 20 years now,
57:03 and Microcom's been around for 10. What's the next need? What's the next
57:07 feature and quotes that I need to build to help SAS founders? And each of these
57:10 things comes out of, well, I need a mastermind. Well, I need funding. Well,
57:13 I need inerson events. And we just one by one kind of have rolled these out
57:16 over the years. >> And of course, you're really one of the
57:19 OG content creators in this space. Your podcast startups for the rest of us, I
57:23 think, is approaching its thousandth episode, not too far in the future.
57:27 It's, you know, as someone who's done, we're at a bit over 250 and that's
57:32 already a lot. So, so getting to 1,000. Uh, and you know, folks, if you enjoy
57:36 listening to the startup podcast, I'm pretty sure you'll enjoy startups for
57:40 the rest of us, too. Uh, even at nearly a thousand episodes, you're still
57:43 keeping it fresh and and sharing really great new knowledge and wisdom. So, uh,
57:46 yeah, get on it. And thanks so much for everything you do for founders
57:49 everywhere, Rob. >> For sure. Yeah. Thanks so much for
57:51 having me on, Yanov. I really appreciate it coming back. >> All right. Well, that's the pod. Have a
57:56 good one. Today's episode of the Startup Podcast was brought to you byte domains.
58:01 Naming your startup should be easy, but it isn't because all the good.com names
58:06 are taken or wildly overpriced. No more. You don't have to compromise. Your
58:10 domain shows your focus and intent. So, make it sharp, clean, and unmistakably
58:15 tech with a tech domain. How do you do that? Head to www.get.te/tsp
58:20 tech/TSP or a registar like GoDaddy and grab your tech domain today. get.te/tsp
58:28 for the startup podcast so they know you