[00:00:00] [music]
[00:00:04] Welcome to another episode of Tech
[00:00:05] Unhinch, where tech gets human powered
[00:00:07] by Code District. I’m your host Ravia
[00:00:09] Javi. Then joining me today is Wes Bush,
[00:00:12] the person who quite literally wrote the
[00:00:14] book on ProductLled Growth, his
[00:00:15] bestseller, How to Build a Product that
[00:00:17] sells itself, has been read over by
[00:00:20] 150,000 people worldwide. Over the last
[00:00:23] decade, West has worked with more than
[00:00:25] 400 SAS companies and helped them unlock
[00:00:27] over a billion dollars in self-s serve
[00:00:29] revenue. Today, we are talking about
[00:00:31] something a lot of companies are dealing
[00:00:33] with at the moment. Why so much AI
[00:00:35] spending produces so little and what
[00:00:37] actually works. Welcome to the show,
[00:00:39] Wes.
[00:00:39] >> Yes, thanks so much for having me.
[00:00:40] >> You’ve worked with over 400 SAS
[00:00:42] companies, but let’s start a bit simple.
[00:00:45] What’s a product you personally love
[00:00:47] using and can’t live without?
[00:00:49] >> Oh, this is a fun one. So there’s
[00:00:51] there’s so many these days where I’m
[00:00:53] just like okay, you know, you like pick
[00:00:55] it up and then you’re just like you you
[00:00:57] love using it like on my phone. This
[00:00:59] one’s a fun one. Uh it’s called Dumbify.
[00:01:02] I don’t like it just changes the UI of
[00:01:05] everything. So it’s so simple, right?
[00:01:08] >> I I’m a sucker for apps like that. So
[00:01:10] I’m like, okay, this is this is great.
[00:01:12] So I use that every day. I mean, I
[00:01:13] meditate. So I’m like Insight Timer is
[00:01:15] great as well. Super simple. There’s the
[00:01:18] great free plan on that one. So I love
[00:01:20] that. And I mean for work I’m just still
[00:01:24] loving Claude like just there’s so much
[00:01:26] I just keep learning and then you’re
[00:01:28] like oh and then you can build the
[00:01:30] agents to do a lot of the repeatable
[00:01:31] stuff. So I feel like you can just go so
[00:01:33] deep on that one. So yeah those are
[00:01:35] probably my top of mind ones.
[00:01:37] >> How did your thought process around claw
[00:01:39] change? You know initially it was just
[00:01:40] an agent to write content better.
[00:01:43] >> I know. And then there’s like levels to
[00:01:45] it too where it’s like when you use
[00:01:47] cloud code and then you hook it up with
[00:01:49] Obsidian, you make your second brain and
[00:01:51] then you start breaking everything down
[00:01:54] into like your little business area
[00:01:56] pockets and then you have your life
[00:01:57] pocket and then you create like the
[00:01:59] dashboards and ages for everything. It’s
[00:02:00] it’s pretty wild just how you can do
[00:02:02] that. And then it syncs across all
[00:02:04] devices too if it is in like a I don’t
[00:02:06] know online cloud area which is great
[00:02:09] because yeah I travel a lot so I’m like
[00:02:11] I’m not always on the same device. West
[00:02:12] AI has made it um easy to ship anything
[00:02:15] yet plenty of products still lose
[00:02:17] customers in early stages. Have you
[00:02:19] experienced that as a customer and how
[00:02:21] do you measure value in a product?
[00:02:23] >> Yeah. So I mean with people being able
[00:02:25] to ship anything very quickly I just go
[00:02:27] back to like well what is the area you
[00:02:30] need to differentiate now and so if it’s
[00:02:32] easy to build anything it’s actually
[00:02:33] harder to build great stuff. So I’m
[00:02:36] seeing actually a reassurence of
[00:02:38] companies talking a lot about quality
[00:02:41] taste. What are those hard to copy
[00:02:43] things where like if anybody can just
[00:02:45] for instance uh actually today uh on
[00:02:47] LinkedIn I saw uh my friends from
[00:02:50] Lemlist this year they they launched a
[00:02:52] new app where it’s like you get your 10
[00:02:55] leads every single day identified for
[00:02:57] your business and none of it like the UI
[00:03:00] or anything else like that like you
[00:03:01] could easily copy that it’s not
[00:03:03] something that’s hard. The thing that’s
[00:03:05] actually hard is like how good is that
[00:03:07] output? What is that? Like are those
[00:03:08] actually 10 people worth reaching out to
[00:03:10] that could drive your business forward
[00:03:11] and everything else? It’s like yay or
[00:03:13] nay. And is it 9 out of 10? Is it 10 out
[00:03:16] of 10? Is it 2 out of 10 when it comes
[00:03:18] to the quality? And so that is really
[00:03:20] where I think the most people have to
[00:03:22] think a lot about is just like how good
[00:03:24] is the quality? How can we get more
[00:03:25] people to that bar where they’re just
[00:03:27] like wowed? And the things that used to
[00:03:29] matter like okay the user experience and
[00:03:31] stuff is becoming a little actually less
[00:03:33] important if people don’t even have to
[00:03:35] use your UI but they still want the
[00:03:37] output on like what is your product do
[00:03:40] what is the value prop how can I get
[00:03:42] there in 60 seconds or less
[00:03:43] >> product thinking gets talked about as a
[00:03:46] growth tactic in plain words what does
[00:03:48] it actually change about how a company
[00:03:50] decides to spend on AI
[00:03:52] >> I would definitely say like they go hand
[00:03:53] in hand like you should invest way more
[00:03:56] in AI to get people a personalized
[00:03:58] experience. I think with product growth
[00:04:00] specifically, it was the last 10 years
[00:04:03] were kind of like, okay, do your best to
[00:04:04] create like a like what we would call
[00:04:06] personalized onboarding flows, but it’s
[00:04:09] actually still probably like two or
[00:04:10] three main flows for how people want to
[00:04:12] use your product versus now it’s like,
[00:04:14] hey, what do you want to do in the
[00:04:16] product? And you can have a chat bubble
[00:04:18] or something like that and it it just
[00:04:19] does it in your product. That’s what
[00:04:20] people want. They want that custom
[00:04:22] journey for themselves. So you know one
[00:04:24] of the MIT studies um was found that 95%
[00:04:27] of enterprise AI projects deliver zero
[00:04:29] measurable impact on the bottom line and
[00:04:32] even the critics only have half of that
[00:04:34] number. Is that money dying on the
[00:04:36] technology itself or on everything
[00:04:38] wrapped around it? What do you think?
[00:04:40] >> Yeah, I think there’s a lot of skunk
[00:04:42] work little projects and like we’re
[00:04:44] still at the early stages too of like
[00:04:46] okay can you do things useful? Like if I
[00:04:48] even think of myself with how I use
[00:04:50] Claude and the all the scheduled tasks I
[00:04:52] have like yes easily 85 90% of them
[00:04:55] don’t provide any measurable value. So
[00:04:57] even if you think of your own personal
[00:04:59] use there’s a lot of waste but for the
[00:05:01] 10% that does work it’s a gamecher and
[00:05:04] it’s immediately getting people to think
[00:05:07] about how can you actually utilize this
[00:05:09] and implement it right because there is
[00:05:12] a little bit of a learning phase for
[00:05:14] every company too. So like out of the
[00:05:16] companies that do invest and waste 80%
[00:05:18] right now. I’m like good, good, that’s
[00:05:19] great. Like maybe next year you have 20%
[00:05:23] utilization of that stuff. And even if
[00:05:25] you just get 20%, will it pay for that
[00:05:27] 80%. Hell yes. And it will be way
[00:05:30] better. So I don’t actually think it’s a
[00:05:31] bad thing. I think it’s just the natural
[00:05:34] process of learning. There is going to
[00:05:35] be some waste. And as long as you’re
[00:05:39] wasting it in a way where you’re not
[00:05:40] going bankrupt, then I think it’s a
[00:05:43] great thing because u everybody right
[00:05:45] now it’s we’re in the wild west era of
[00:05:47] the AI phase and it’s it’s just going to
[00:05:49] get better and better. So it’s like how
[00:05:51] can we utilize it? Do we have good use
[00:05:53] cases where we can apply this and where
[00:05:56] I think product growth intertwines
[00:05:58] really nicely with this and AI is just
[00:06:00] this is how your business makes money
[00:06:02] and how it delivers value. If you can
[00:06:04] utilize AI to do that better and faster,
[00:06:06] you’re going to make more money. So,
[00:06:08] it’s like it’s uh if you can focus
[00:06:09] anywhere, focus there.
[00:06:10] >> Yeah. So, you know, here’s an
[00:06:12] interesting one, and I want you to dive
[00:06:13] a bit deeper into this one. You live in
[00:06:15] the world of um you know, products, but
[00:06:18] a lot of companies come from services.
[00:06:19] In simple terms, what’s the real
[00:06:21] difference between marketing a product
[00:06:23] and marketing a service? And which one
[00:06:24] is harder? Yeah, I mean services always
[00:06:27] focused on the outcome a little bit more
[00:06:29] than a product would be like. It’s not a
[00:06:31] means to an end for everything, but it’s
[00:06:33] more like it’s a little piece of that
[00:06:35] workflow. And so if you think of uh like
[00:06:38] a CRM tool, great. Okay, you could have
[00:06:40] the best tool to do that to manage your
[00:06:43] pipeline and everything else. But now a
[00:06:45] service is like, “Hey, you’re going to
[00:06:46] hire a sales agent that is going to find
[00:06:50] who you should contact, reach out to
[00:06:52] those people, book those meetings, and
[00:06:54] all you have to do is show up on the
[00:06:56] meeting, and we’ll take care of it.” So
[00:06:58] even after the meeting’s done, it’s
[00:07:00] going to follow up with them, send them
[00:07:01] the proposal, and go through that whole
[00:07:03] thing. And so what you’re buying in that
[00:07:04] case, it’s it’s the whole execution of
[00:07:07] the workflow. thinking about it more
[00:07:09] like an employee versus okay we’re
[00:07:11] selling you this widget called a CRM
[00:07:13] which is nice everybody needs one but
[00:07:16] thinking about it from okay the overall
[00:07:19] output and the work that needs to get
[00:07:20] done I think it’s that’s the biggest
[00:07:22] part where people are headed
[00:07:23] >> yeah how does one pricing decision end
[00:07:25] up shaping how a whole company behaves
[00:07:27] when a business builds by something like
[00:07:29] the number of profiles instead of the
[00:07:32] value people actually get how does that
[00:07:34] ripple through sales support and product
[00:07:36] >> yeah so recently wrote a a nice spicy
[00:07:39] post on this one for customer.io because
[00:07:42] yeah, I was with a tool like customer.io
[00:07:45] that manages your emails, SMS, and
[00:07:47] everything else. They rightly charge
[00:07:49] based on profiles, which doesn’t seem
[00:07:51] like a bad idea. Like a profile is like,
[00:07:53] okay, any accounts or person inside that
[00:07:56] database that you can contact. Now, when
[00:07:59] you think about where does the value
[00:08:01] originate, you know, to a certain
[00:08:02] degree, it is profiles like okay, great.
[00:08:04] Everybody’s all in one place. It’s like
[00:08:05] a spreadsheet. Whoopee. But the real
[00:08:08] value in my opinion is actually reaching
[00:08:10] out to them. It’s contacting them. That
[00:08:13] is where the value originates. So if you
[00:08:15] send an email to let’s say 12,000 people
[00:08:17] and that’s all you send to in the whole
[00:08:19] month but you get build on 80,000 plus
[00:08:22] profiles. It doesn’t make sense. There’s
[00:08:24] a you know a gap between how you’re
[00:08:26] charging and then how you are actually
[00:08:29] uh delivering value. And so the best
[00:08:31] value metrics always focus on that last
[00:08:33] part which is how do we actually deliver
[00:08:35] value? Now, you can’t all be like
[00:08:38] Stripe, which is an amazing business
[00:08:40] model because it’s like, okay, if you
[00:08:42] make a,000 bucks on, you know, a payment
[00:08:44] link that you send to somebody, okay,
[00:08:46] they charge you 2 to 3% like a credit
[00:08:48] card. So, it’s like fantastic, great
[00:08:50] business and it’s super simple. When it
[00:08:52] comes to other tools that that don’t
[00:08:54] have that luxury, how do you do it? So,
[00:08:56] like custom.io, you come up with a
[00:08:58] random metric that you think closely
[00:08:59] aligns, but turns out it’s not always
[00:09:01] super aligned. So when it comes to AI
[00:09:03] tools, this is really where it gets
[00:09:05] interesting. And so one of my my
[00:09:08] thoughts I was like over the last 10
[00:09:09] years um so many people have been like
[00:09:11] bashing userbased pricing rightfully so
[00:09:14] because like every company under the sun
[00:09:16] was like we don’t know how to price
[00:09:18] based on users because there’s value in
[00:09:20] that. But now we’re seeing it’s actually
[00:09:22] just credit based pricing is the new
[00:09:25] userbased pricing. And so, um, why that
[00:09:27] gets me like kind of laughing is because
[00:09:30] it’s so random yet it makes sense. It’s
[00:09:33] like, yeah, uh, AI charges based on
[00:09:36] credits. Uh, all the LLMs charge based
[00:09:38] on credits. So then when you need to
[00:09:40] come up with your pricing, also might
[00:09:42] want to consider credits. But then every
[00:09:45] company, especially as software tools,
[00:09:47] has a different metric of like this is
[00:09:50] how we calculate credits. So then who
[00:09:52] gets left with the short end of the
[00:09:54] stick? It’s always the customer because
[00:09:56] they’re like, “What what is the pricing?
[00:09:58] How much are we going to get charged?”
[00:09:59] You have no idea. And so what you have
[00:10:02] to to kind of go back to the actionable
[00:10:04] part here too, the two things to take
[00:10:06] away here is one, your pricing has to
[00:10:08] align with how do you actually deliver
[00:10:10] value. So going back to what you talked
[00:10:12] about with services versus applications,
[00:10:15] um, think about, okay, is it a certain
[00:10:16] amount of work or an outcome that you’re
[00:10:18] going to get people closer towards?
[00:10:20] Maybe it’s closed deals or closed volume
[00:10:22] or something like that. um if you can
[00:10:24] get to that outcome, charge a percentage
[00:10:26] of that if you can actually move the
[00:10:28] needle on that metric. Um but then the
[00:10:30] second thing, and this is like the easy
[00:10:32] part but hard to do, is just does your
[00:10:35] pricing pass the 5-second test. So if
[00:10:38] anybody who doesn’t really understand
[00:10:39] your product that well, can look at your
[00:10:41] pricing page and be like, “Oh yeah, it’s
[00:10:42] that plan. That’s what I would need.”
[00:10:44] Then you passed and you can move on with
[00:10:46] your life as long as it still matches
[00:10:49] the first piece of like it aligns with
[00:10:51] the core value of your product. But most
[00:10:52] AI companies and everything else, I
[00:10:54] don’t see them doing that first piece
[00:10:56] and then the second piece of like, oh,
[00:10:57] how much am I going to get charged? It’s
[00:10:59] still really vague on that end.
[00:11:01] >> Oh, that’s that’s great insight, Wes.
[00:11:03] So, across dozen of AI products, um, one
[00:11:05] pattern keep servicing, uh, that we see
[00:11:08] where teams bolt AI onto a workflow that
[00:11:10] was already broken and just produce the
[00:11:13] errors faster. What is the most common
[00:11:15] broken thing founders try to cover that
[00:11:17] you’ve seen? Yeah, I mean the bolt-on AI
[00:11:20] strategy is super common because they
[00:11:22] are just trying to be like, okay, how do
[00:11:24] we make this current product a little
[00:11:25] bit faster, but it’s a lot like um they
[00:11:29] they’ve studied this a lot in big
[00:11:30] enterprises. It’s like the optimization
[00:11:32] mindset where okay great like we’re
[00:11:34] going to optimize something like 2 to
[00:11:35] 3%. Sure, great. do that every year and
[00:11:38] you kind of keep improving things slowly
[00:11:40] and then eventually you kind of get to a
[00:11:42] point where it gets really hard and you
[00:11:44] hit diminishing returns on that
[00:11:45] investment and you’re like okay we can
[00:11:47] no longer get 2 to 3% gains and so right
[00:11:50] now we’re still in the phase where with
[00:11:52] adding and bolting on AI you can get a 2
[00:11:54] to 3% improvement per year but you’re
[00:11:57] still going to be exposed to a AI native
[00:12:00] company that goes through the entire
[00:12:02] workflow of how to you know any category
[00:12:05] uh let’s call it even closing a new
[00:12:07] business or something like that. You’re
[00:12:08] going to have those like traditional CRM
[00:12:10] companies [clears throat] that are just
[00:12:11] like, “Oh, we got a AI agent for meeting
[00:12:13] follow-ups. Okay, let’s bolt that on. We
[00:12:15] got a AI agent eventually to, you know,
[00:12:18] update all the fields and and register
[00:12:19] contacts. Okay, great. We we do all
[00:12:21] that.” Now, then there’s the AI native
[00:12:23] company that says, “Where does most of
[00:12:25] the time go? Where do most people waste
[00:12:27] time?” And it’s a completely different
[00:12:29] workflow. They’re like, “Actually, let’s
[00:12:30] reimagine how people interact with, you
[00:12:32] know, leads and everything else like
[00:12:34] that. What if they didn’t even need to
[00:12:35] log in to the CRM at all? They just talk
[00:12:39] with Claude and every morning we send
[00:12:40] them, hey, these are the 10 leads you
[00:12:42] should follow up with. Chat within
[00:12:44] Claude. Okay, we got the emails drafted.
[00:12:46] We’re ready to go. And the entire
[00:12:48] workflow changes. It’s like, what are
[00:12:50] the one or two things that you people
[00:12:52] need to do to actually sell a lot? And
[00:12:54] then they just work backwards from that.
[00:12:56] and it would look completely different
[00:12:58] than the company that just says, “Hey,
[00:13:00] we’re going to consistently add some of
[00:13:01] these AI agents to simplify some of the
[00:13:03] workflows because you still have the
[00:13:04] same workflow. It’s just a little faster
[00:13:06] versus let’s blow up the workflow and
[00:13:09] just go through like one of those two
[00:13:11] things or three things you got to do to
[00:13:13] actually close new business and let’s do
[00:13:15] that.” And so the companies that focus
[00:13:17] on disrupting themselves right now uh
[00:13:19] have a huge advantage, although it’s
[00:13:21] harder to do it if you have that
[00:13:22] existing one. That’s why you’re seeing
[00:13:23] all these AI native companies come out
[00:13:25] of the woodwork and be like, “Hey, look,
[00:13:27] we can do that in a fraction of the
[00:13:29] time. Not a 2 to 3% improvement, but
[00:13:31] that’s why we call it disruption.”
[00:13:34] >> Yeah, definitely. So, you know, while
[00:13:36] going through a lot of your thought
[00:13:37] pieces, I found that you’ve said that
[00:13:39] users really complain before they leave.
[00:13:41] They just stop showing up. When a
[00:13:43] company cannot see the people um quietly
[00:13:46] walking out, where should they look
[00:13:47] first to find a reason?
[00:13:48] >> Yeah, I mean, churn data is usually the
[00:13:51] best. And so if you have a couple
[00:13:54] questions when people are canceling
[00:13:56] their subscription, that’s really where
[00:13:58] you can learn a lot pretty quickly. And
[00:14:01] I’m not a big fan of like forcing people
[00:14:03] to hop on a call. But [laughter] I am a
[00:14:05] big fan of like if you can get those
[00:14:07] people on a call, do it. And maybe you
[00:14:10] could give them like, hey, three months
[00:14:12] free if you hop on a quick call to go
[00:14:14] through this whole experience with us so
[00:14:15] we can make it right or understands
[00:14:17] everything else. That could be a great
[00:14:18] option. But what I have seen with some
[00:14:21] other companies that are like, you know,
[00:14:23] simple software rappers, they have a lot
[00:14:25] of people saying like, “Oh yeah, people
[00:14:26] are cancelling just going straight to
[00:14:27] claude.” It’s like, if your moat is just
[00:14:30] that, if people can just do it in
[00:14:32] another tool very quickly, you have a
[00:14:34] very weak moat and even if it’s, you
[00:14:37] know, 80% as good and claude versus your
[00:14:39] tool, it’s it’s good. They’re going to
[00:14:41] catch up and they can already see it and
[00:14:44] they’re leaving proactively. And so you
[00:14:46] had to get to the bottom of why are
[00:14:47] people leaving? And it goes back to the
[00:14:50] previous question you asked of why?
[00:14:52] Well, the approach is better. If you can
[00:14:54] manage something, if you can talk to
[00:14:56] something, it feels faster. It feels uh
[00:14:58] like you have a bit more ownership. You
[00:15:00] don’t have to worry about uh oh, where
[00:15:02] do I click in the UI of this tool to
[00:15:04] figure out that you’re like, “No, I just
[00:15:06] talk.” And most people even in tech
[00:15:09] these days, they’re not even typing.
[00:15:10] they’re just like talking and then it’s
[00:15:12] getting auto AI transcribed and then
[00:15:15] they don’t have to go through it. So it
[00:15:16] just feels like it flows a bit better.
[00:15:18] So you have to have a much better
[00:15:20] approach for people to even consider
[00:15:22] using your UI unless you just become a
[00:15:24] connector. So after 2 years of research
[00:15:26] across hundred of companies, you found
[00:15:28] nine essential elements that separate
[00:15:30] the productled growth success stories
[00:15:32] from everyone else. Without giving the
[00:15:34] whole playbook away, what are one or two
[00:15:37] of those elements that founders most
[00:15:39] often miss?
[00:15:39] >> Yeah. Uh the biggest three is always
[00:15:42] comes down to the foundation. So a lot
[00:15:44] of people will look at productled growth
[00:15:46] and they’ll think oh it’s just a free
[00:15:47] trot. That’s all we need to do. We need
[00:15:49] to you know update the homepage, make
[00:15:51] sure we have a free model that gives
[00:15:53] people access to the product. I’ve heard
[00:15:56] onboarding is important. So let’s figure
[00:15:57] out how to get those people to value a
[00:15:59] little bit faster. And then self-s serve
[00:16:01] pricing. Okay, we need to to have that.
[00:16:02] So like on the surface, I call it like
[00:16:04] the product the growth iceberg. That’s
[00:16:06] what people see. That’s what they do.
[00:16:08] and then they start there and then it
[00:16:12] usually doesn’t work. So that’s like the
[00:16:15] it’s disappointment and they hit the
[00:16:17] iceberg and hopefully they don’t sink
[00:16:19] like the Titanic. But the next part is
[00:16:22] like the foundational elements are
[00:16:24] strategy is the first one. So like how
[00:16:27] does this even align with your business?
[00:16:29] How does it make you stand out as a
[00:16:31] company in your market? Is this part of
[00:16:34] your vision for your company? If you
[00:16:36] think of Canva, can you actually make
[00:16:38] design accessible to everybody without
[00:16:41] it being productled? Hell no. Like there
[00:16:43] there’s no way you’re going to make that
[00:16:45] model work, you know, approaching people
[00:16:47] in all countries, all walks of life, all
[00:16:50] budgets. Uh with the salesled model,
[00:16:52] you’re like, no, you the saleserson is
[00:16:54] going to be working for free most of the
[00:16:56] time to even onboard people and get them
[00:16:58] on like it’s impossible. So it has to
[00:17:00] come from there. And ideally your
[00:17:02] founder and leadership team has to all
[00:17:04] be aligned on this and get the
[00:17:06] conviction from the founder and CEO the
[00:17:08] commitment of the leadership team
[00:17:10] because if they don’t get that locked in
[00:17:12] guess what month three comes around with
[00:17:14] that productled motion or that free
[00:17:16] trial that didn’t work that well and you
[00:17:18] scrap it. So that just goes back to the
[00:17:21] old sales notion most times. So that’s
[00:17:22] the first thing and then the uh the
[00:17:24] second is do you understand who your
[00:17:26] ideal user is? This isn’t the buyer.
[00:17:28] [clears throat] uh sales like companies
[00:17:29] are really good at understanding like
[00:17:31] who is that ideal buyer the decision
[00:17:33] maker let’s go after them because they
[00:17:35] make the the big budget decisions but in
[00:17:37] a product it’s all bottoms up for the
[00:17:40] most part so it’s like if you think of
[00:17:42] Slack somebody might start off with it
[00:17:44] share with their team and then if it’s a
[00:17:46] big company they have multiple teams
[00:17:48] they share with another team and then
[00:17:49] they’re like hey why don’t we like roll
[00:17:51] this out throughout the whole company
[00:17:52] now they could do that self-s serve if
[00:17:54] it’s you know a small company under 50
[00:17:56] or 100 people um but eventually if it’s
[00:17:58] like a company like Microsoft although
[00:18:00] they would use their own team product
[00:18:01] but if it is you know a bigger team then
[00:18:04] they would be able to like work with
[00:18:06] sales and actually roll it all out and
[00:18:08] and get it hopefully a better deal or
[00:18:10] something like that. So that’s how it
[00:18:12] would work. You got to understand those
[00:18:14] users who are those people that are
[00:18:15] going to like champion your product
[00:18:16] initially and get to value and really
[00:18:19] what is value for them to to really uh
[00:18:21] share this with their boss their other
[00:18:23] team members and get that organic
[00:18:25] virality inside the company. And so you
[00:18:27] got to understand those users. And then
[00:18:28] the third thing is the model. So the
[00:18:31] free model is so important to think
[00:18:34] about what do we need to give away for
[00:18:36] free. Some companies just slap on and
[00:18:39] I’ve done this too a free trial for 14
[00:18:41] days on an existing product without
[00:18:43] thinking too much about what is like
[00:18:46] what has to be true for a product to
[00:18:48] work. So one of the biggest things is
[00:18:49] does this solve like the beginner
[00:18:51] problem in our market. So I’ll give you
[00:18:53] an example here too. When we launched a
[00:18:56] free trial at a company called Vidyard,
[00:18:58] it was a like, you know, a video
[00:19:00] marketing analytics tool. So, you upload
[00:19:02] the video, then you had to put it on
[00:19:04] your website, integrate it with your
[00:19:05] marketing automation platform, and then
[00:19:07] I could see, oh yeah, look, you watch
[00:19:09] 50% of our our video. Amazing. And yet,
[00:19:12] nobody did that because most people that
[00:19:14] got there, they they didn’t have a video
[00:19:16] actually upload. And this was like 10
[00:19:19] years ago. So, it wasn’t as common or as
[00:19:21] easy to create these great videos.
[00:19:23] Nowadays, it’ be easy. He’s like, “AI
[00:19:25] generate that stuff and just throw it on
[00:19:26] your site.” But back then it was hard.
[00:19:29] And so what we decided, we were like, we
[00:19:31] got to solve that beginner problem of
[00:19:32] creating a video. So we created a Chrome
[00:19:34] extension, made it super easy. In a
[00:19:36] couple clicks, you could record your
[00:19:38] screen in Chrome, send it to anybody,
[00:19:40] and then you would see, oh yeah, wow,
[00:19:42] they watched 50% of that video or 100%.
[00:19:45] If it’s a prospect, you’re like, that
[00:19:46] tells you a lot about uh especially if
[00:19:49] it’s going through a proposal or
[00:19:50] something like that. you have a lot of
[00:19:52] data now to be like hey I saw you
[00:19:54] weren’t that interested or hey you watch
[00:19:56] 100% uh is it good time to like lock
[00:19:59] this in or or something else you you
[00:20:01] know you have signals that you didn’t
[00:20:02] have before that’s the capability of the
[00:20:05] product that you can experience in a
[00:20:06] smaller scale and so that’s the first
[00:20:09] thing is understanding the model is the
[00:20:11] 8020 because a lot of times you can
[00:20:13] spend so much time on like improving
[00:20:15] your onboarding your offer but then it’s
[00:20:17] like oh you didn’t even tackle the thing
[00:20:19] everybody in your market has the problem
[00:20:20] with Well, that’s not going to work. You
[00:20:22] just wasted so much time. So, those are
[00:20:24] the first three that I would say most
[00:20:26] people just skip over and don’t focus
[00:20:28] too much on, but they they they matter
[00:20:30] the most.
[00:20:30] >> Also, Wes, you’ve held up Gen Spark as a
[00:20:33] model for getting users to value fast,
[00:20:35] and it did cross a $155 million in
[00:20:38] revenue around 10 months. Yet, even
[00:20:40] people who admire it have described
[00:20:42] stalling inside its onboarding process.
[00:20:45] Does the 60-cond rule survive contact
[00:20:47] with real products like these? Yeah, I
[00:20:50] mean it depends what you do in tools
[00:20:52] like GenSpark when it comes to those
[00:20:54] first 60 seconds because right now if I
[00:20:57] look at a tool even like a GenSpark or
[00:20:59] any other new emerging you know AI tools
[00:21:01] that you do try that have like that
[00:21:03] chatbot interface the expectation now is
[00:21:06] quite high of what it could do. It’s
[00:21:08] like okay great you should be able to
[00:21:09] respond to like basic stuff should be
[00:21:11] better than Google for most things and
[00:21:14] then I should get something customized
[00:21:16] to what I want. Now the problem is that
[00:21:19] they’re all running into is why use this
[00:21:22] platform over another especially as you
[00:21:24] get embedded like I was talking earlier
[00:21:26] about claude and how I love it and I’m
[00:21:28] like you know you you get embedded you
[00:21:30] get your workflows it becomes like your
[00:21:31] front like you have all your favorite
[00:21:34] stuff here it’s organized the way you
[00:21:35] want so the switching cost grows because
[00:21:39] you now have to build some context in
[00:21:41] some of those other tools as well uh
[00:21:43] that you don’t have and so when it comes
[00:21:45] to that first 60 seconds what I think is
[00:21:47] going to be the hardest all of these
[00:21:49] tools and this goes for any company even
[00:21:51] if you’re not an LLM is how can we
[00:21:54] showcase our differentiated value
[00:21:56] because it’s not enough to be like we
[00:21:58] just responded to whatever they they
[00:21:59] said whatever their question was we got
[00:22:01] it it’s like that’s not good enough like
[00:22:02] what is the differentiated part now for
[00:22:04] GenSpark they are trying to
[00:22:06] differentiate on look we use all the
[00:22:08] models doesn’t matter which one chache
[00:22:11] or you know Claude Kimmy doesn’t matter
[00:22:13] we use them all in different areas if
[00:22:15] you want to create music great There’s a
[00:22:17] better model for that. It’s not clawed.
[00:22:19] If you want to code, sure, we’re using
[00:22:21] Fable 5. And so, people that like that
[00:22:23] flexibility love it. But people that
[00:22:26] don’t need that flexibility don’t really
[00:22:28] care. So, what are the things we could
[00:22:30] do uh in those first 60 seconds to
[00:22:32] actually showcase that? And that is the
[00:22:34] challenge when it comes to even we were
[00:22:36] discussing earlier with the free model
[00:22:38] is it can’t just be, you know, that hey,
[00:22:41] like look, you create a video or
[00:22:42] something like that. In Vidyard’s case,
[00:22:43] it’s like it had to show the analytics,
[00:22:46] you know, what percentage of the video
[00:22:48] did that person actually watch because
[00:22:50] that was actually the same thing in the
[00:22:52] core product. So, if you like that, you
[00:22:53] love the core product. It’s just you
[00:22:55] roll it out across your entire website,
[00:22:57] not just, you know, the simple Chrome
[00:22:58] video. One of the popular arguments is
[00:23:00] um every one of the 50 fastest growing
[00:23:03] software companies run on a productled
[00:23:06] motion. So, it must be essential. But
[00:23:08] did that approach actually make them
[00:23:10] grow or did they just adopt it because
[00:23:12] they were already winning?
[00:23:14] >> Yeah. So I think it’s a couple factors
[00:23:16] here too. It’s not just any like broad
[00:23:18] claim. It’s like yeah it has to be PG
[00:23:20] motion but the products themselves are
[00:23:22] really good. And this is one of those
[00:23:24] things in like okay your productled
[00:23:27] motion will amplify. It’s a multiplier
[00:23:29] of your product. So if your product is
[00:23:31] one out of 10 better than the
[00:23:33] alternatives, well guess what? Even if
[00:23:35] you have a 10 out of 10 product growth
[00:23:37] motion, you still get 10 as your output.
[00:23:39] Now if your product is four out of 10
[00:23:42] better than the alternatives, which many
[00:23:45] of those products were. So you think of
[00:23:46] like a lovable versus I didn’t even hear
[00:23:48] about like bolt or base 44 or any of the
[00:23:50] other alternatives for a while and it
[00:23:52] was like arguably four out of 10 better
[00:23:54] when it came out. Now, it’s not always
[00:23:57] the case, but when it came out and they
[00:23:59] had a really solid product growth
[00:24:00] motion. It was like literally leading
[00:24:02] with activation on their homepage, which
[00:24:04] at that time was like crazy. And so,
[00:24:06] it’s like you pair that 4 to 10 better
[00:24:08] with a 10 out of 10 product growth
[00:24:10] motion, you get 40 as your overall like
[00:24:12] growth output of how fast you can grow.
[00:24:14] That is what I think is absolutely
[00:24:16] fascinating like how you can unlock this
[00:24:18] crazy fast growth for those companies.
[00:24:20] And if you go through all of them like
[00:24:22] they did have a productled growth
[00:24:23] motion, they did have a great product.
[00:24:25] Now it comes back to was it because of
[00:24:27] the product of growth motion. This is
[00:24:29] where I would argue yes because a
[00:24:32] salesled motion can’t scale that fast.
[00:24:34] Try growing from 0 to 100 million in
[00:24:37] like 6 months or 8 months with a
[00:24:39] salesled team. It’s almost impossible to
[00:24:42] hire people that fast. It’s almost
[00:24:43] impossible to train them that fast to
[00:24:45] get that product understood. Now maybe
[00:24:47] if you’re Microsoft and you have the big
[00:24:49] resources, you have the capability
[00:24:51] inside your company to hire deploy
[00:24:53] people that fast, you could do it. You
[00:24:55] could absolutely do it. But for a
[00:24:56] startup, no. Like [laughter] it’s just
[00:24:58] like crazy talk. It’s not possible. The
[00:25:01] product growth motion actually allows
[00:25:03] you to scale your company at a much
[00:25:06] faster pace than you ever could if you
[00:25:07] had a salesled motion. So that’s my my
[00:25:10] little argument there is there would be
[00:25:12] no other way. Now a lot of these
[00:25:13] companies have now added a productled
[00:25:15] sales motion to scale into enterprise
[00:25:17] but that’s after they got to their first
[00:25:19] 100 million.
[00:25:20] >> So west one of your own rules says that
[00:25:22] um a company should assume its main
[00:25:24] feature will be free within 18 months.
[00:25:26] So why wouldn’t the same thing happen to
[00:25:28] the books and frameworks that teach AI
[00:25:30] strategy and what would make one still
[00:25:33] worth paying for? I mean this is the
[00:25:35] like if you create knowledge or you
[00:25:37] create anything these days you have to
[00:25:39] assume that it will be free essentially
[00:25:42] very soon and so like the top level of
[00:25:45] like I have like five levels of product
[00:25:47] companies the first one’s like you’re
[00:25:48] just a commodity and it’s like okay it’s
[00:25:50] forgettable level two contender like you
[00:25:53] you do something differentiated and you
[00:25:56] stand out a bit in the market but you’re
[00:25:58] still a small player and then level
[00:25:59] three is like you’re on the cusp of the
[00:26:01] market top three to five player in the
[00:26:03] market you definitely are like duking it
[00:26:05] out for like that like you know 20% of
[00:26:07] the market share and then there’s like
[00:26:08] the obvious choice you are the obvious
[00:26:10] choice in the market that’s like where
[00:26:12] you get the majority of the market share
[00:26:14] but in order to stay as the obvious
[00:26:16] choice what I’m arguing is you actually
[00:26:18] have to become a crusader which is like
[00:26:20] you might have set out to become the
[00:26:22] obvious choice let’s say let’s call it
[00:26:24] like design with like Canva or something
[00:26:26] like that but then it’s like you have to
[00:26:27] figure out a way a business model to
[00:26:29] make that completely free because that
[00:26:31] is going to be the expectation people
[00:26:32] just want to be able to do that for
[00:26:34] free. So what are you left with? Okay,
[00:26:36] you kind of go up the value chain of
[00:26:38] okay, if we made design free, how would
[00:26:40] that impact teams uh inside a company?
[00:26:43] Oh, okay. Uh they would need, you know,
[00:26:45] design assets, the brand assets all in
[00:26:47] one place. And you want to always keep
[00:26:50] growing that bigger problem. So like you
[00:26:53] might make your beginner problem
[00:26:54] completely free and then okay, what are
[00:26:57] those intermediate problems that people
[00:26:58] run into? Okay, for teams and then
[00:27:00] enterprise, oh, that’s more advanced
[00:27:02] problems that we have to solve for that.
[00:27:04] And so that’s where I would say every
[00:27:06] company should be thinking about. And if
[00:27:08] you’re not and somebody else does this,
[00:27:10] then you’re in a pretty difficult spot
[00:27:13] because competing with free is really
[00:27:15] hard because it’s like, well, do I want
[00:27:17] to use the free option or pay for
[00:27:19] somebody else? And um there’s very few
[00:27:21] instances where people are like, yes, if
[00:27:23] it’s comparable, I want to pay for it.
[00:27:25] Unless it’s like weird security things
[00:27:27] for big enterprises where they’re like
[00:27:29] they got a little bit better um security
[00:27:31] on some of those pieces. So that’s one
[00:27:32] thing but I think the second part of
[00:27:34] your question was regarding like
[00:27:36] knowledge products like if you have
[00:27:38] frameworks or books or stuff like that
[00:27:41] then it really becomes about the
[00:27:43] application. So one thing we’re we’re
[00:27:45] actively building and exploring at
[00:27:47] product lad 2 is if you look at a book
[00:27:50] you know you have to read it
[00:27:51] unfortunately to gain [laughter]
[00:27:53] any of the expertise and it’s like how
[00:27:55] long does that take well even though my
[00:27:57] books are shorter like 2 to four hours
[00:27:59] to read if you listen to it and stuff
[00:28:01] like that that’s still a lot of time and
[00:28:03] although I would argue it’s still great
[00:28:05] to do that if you’re going to build a
[00:28:06] product company you know most people
[00:28:09] don’t want to become a product expert
[00:28:11] and so whether you do a knowledge
[00:28:13] product or you create a just product in
[00:28:16] general. If it’s going to be productled,
[00:28:18] what it has to do is it needs to give
[00:28:20] anybody the capability to do anything
[00:28:23] without becoming the expert without you
[00:28:26] know fiddling around with your UI
[00:28:28] without like all the pain it takes to
[00:28:31] accomplish that outcome. People just
[00:28:33] want the dopamine hit of like oh look
[00:28:35] this product now enables me to design.
[00:28:38] this product now enables me to think
[00:28:40] better and faster and all these things.
[00:28:43] And so when you remove all those
[00:28:44] capabilities, what you’re left with is
[00:28:46] how can we make this an instantaneous
[00:28:49] capability transfer. And that’s the
[00:28:51] interesting part like what we’re trying
[00:28:52] to work towards now is train AI with all
[00:28:55] our books, all our IP, all our client
[00:28:57] engagements. And now we just will coach
[00:28:59] you and ask you questions and be like,
[00:29:01] “Oh, you want to build out a new
[00:29:02] homepage because we realized your signup
[00:29:05] conversion rate sucks.” Okay, so that’s
[00:29:07] one, we validated this is the focus
[00:29:09] area. Two, let’s go through this
[00:29:11] process. We’re going to ask you
[00:29:13] questions. You don’t have to be great at
[00:29:15] product marketing, but guess what? This
[00:29:17] output will make you look great. It will
[00:29:19] be an amazing home. And if it’s your
[00:29:22] pricing issue or something like that, u
[00:29:24] once again, it’s like we’ll just ask you
[00:29:25] some questions. We’ll bake the cake
[00:29:27] together and then you get to eat the
[00:29:29] cake. That’s the best of both worlds.
[00:29:31] And so that’s where I think a lot of
[00:29:33] knowledge creators need to head towards
[00:29:35] is creating that experience to where
[00:29:38] people can experience your IP, your
[00:29:41] expertise without the long drawn out
[00:29:44] process of like ah I have to read all
[00:29:45] five books. But I do think I have no
[00:29:48] plans for myself of stop writing books
[00:29:51] because it trains AI so well. And if you
[00:29:53] use that with the company context that
[00:29:56] you gain from when people chat with your
[00:29:58] AI chat, you can still get them the
[00:30:00] output pretty quickly. Uh so that’s the
[00:30:03] part I’m I’m more bullish on and excited
[00:30:05] about. But yeah, I still think people
[00:30:06] will buy product growth book because
[00:30:08] they’re like I identify I want a
[00:30:10] productled business and this is going to
[00:30:12] give me the mindset of of that.
[00:30:13] >> Yeah, because you know Wes I also see
[00:30:15] that there is this you know new I won’t
[00:30:17] say it’s a new jargon though but
[00:30:19] productiz SAS has also been in the play.
[00:30:22] So is it like a hard attempt to meet the
[00:30:24] productled growth marketing for
[00:30:26] services-based companies or is it
[00:30:28] something that actually work for
[00:30:30] services-based companies out there? What
[00:30:32] are your thoughts on that?
[00:30:33] >> Yeah, I mean I think if you have a good
[00:30:35] service, it should be productized to
[00:30:37] some degree. Now there’s always going to
[00:30:39] be that I call it there was like this
[00:30:41] famous painter I name but he was like
[00:30:44] really good at drawing like the the
[00:30:46] faces whoever the artwork was. So he had
[00:30:49] other artists do everything else around
[00:30:52] it. And then he would spend, you know,
[00:30:53] the one hour of just doing the face. And
[00:30:55] that was his jam. He was like, I’m
[00:30:58] really good at this part. Let me just do
[00:30:59] this. And everything else that took up
[00:31:02] over 80 to 90% of the time was
[00:31:04] delegated. So it’s kind of the same
[00:31:05] thing there, too. It’s like AI can do
[00:31:07] that 80 to 90% of the the work to get
[00:31:10] that everything else around it. But like
[00:31:12] for product le motion as an example
[00:31:14] still like where I could add a lot more
[00:31:16] value than AI’s free model. I don’t know
[00:31:18] why but like I’m really good at deciding
[00:31:20] like what you should give away for free
[00:31:21] and if you give away the wrong free
[00:31:22] thing really hard to know if you give
[00:31:24] away the wrong free thing and then it
[00:31:27] takes months and then you’re like why
[00:31:28] are people not converting and it’s a
[00:31:30] really expensive decision. So it’s like
[00:31:31] yes you can productize to a certain
[00:31:33] extent and there will always be like you
[00:31:36] could have the like auto mode for the
[00:31:37] companies that are really small where
[00:31:39] it’s like just go ahead with this. uh
[00:31:40] it’s good enough but for the companies
[00:31:43] that are you know uh over 5 million or
[00:31:45] 10 million or 100 million it’s like yes
[00:31:47] please like you should absolutely work
[00:31:49] with us at least have us double check
[00:31:52] that because u you know AI does
[00:31:54] hallucinate it does make mistakes so
[00:31:56] where there needs to be a lot of trust I
[00:31:58] think that’s where you still need to add
[00:32:00] that human element because it’s like
[00:32:02] it’s a big decision could go wrong and
[00:32:05] let’s let’s make this right that’s
[00:32:07] actually where I think if one will train
[00:32:09] the AI I in the future, but then two be
[00:32:12] an amazing upsell at the end of the day
[00:32:13] for these service companies to really
[00:32:15] think about, okay, what does that
[00:32:17] journey look like? And where does the
[00:32:19] human add actually a little bit extra
[00:32:22] value than just the AI doing it all on
[00:32:24] its own?
[00:32:25] >> Do do you think that there’s going to
[00:32:26] come a point when AI is not going to
[00:32:28] hallucinate as well?
[00:32:29] >> Oh, for sure. Yeah. I mean you kind of
[00:32:32] seen different models like I mean
[00:32:34] notebook elm is an example like does it
[00:32:36] hallucinate really like within bounds
[00:32:39] like it’s just based on the data you
[00:32:41] give it and so I think there is ways of
[00:32:42] training it so it gets really exact and
[00:32:46] it doesn’t [laughter] in some
[00:32:49] ways if it doesn’t know the answer like
[00:32:51] okay I can’t just pull stuff out of thin
[00:32:53] air here it has to be based on these
[00:32:55] resources or these playbooks um that
[00:32:57] have been based on other companies as
[00:32:59] well. Yeah, I do think the beauty of AI
[00:33:02] is that it does know how to it’s like um
[00:33:05] improv like it knows how to jam and Yes.
[00:33:08] And [laughter]
[00:33:09] and that’s sometimes where you get some
[00:33:11] good ideas like when it has been beating
[00:33:13] a lot of humans at some of like the the
[00:33:15] world’s hardest games. The way it plays
[00:33:17] games is completely different and that’s
[00:33:19] where people are starting to be like oh
[00:33:21] it’s it’s doing things that we haven’t
[00:33:23] done. It’s thinking outside the box a
[00:33:24] little bit more. That’s actually what’s
[00:33:26] going to help us solve some of the
[00:33:28] world’s toughest problems. like whether
[00:33:29] it’s solving cancer, whether it’s
[00:33:30] solving uh you know all these other
[00:33:32] bigger problems that you know go outside
[00:33:34] of tech which are still very important
[00:33:36] to solve. I think it will be really
[00:33:37] really cool to see that part.
[00:33:38] >> Wes, one of your own shows, the
[00:33:40] productled podcast. One of your guests
[00:33:42] mentioned the majority of their new
[00:33:44] signups are now AI agents rather than
[00:33:46] people. So if the buyer becomes a bot,
[00:33:48] what happens to everything the industry
[00:33:50] has taught about getting the users to
[00:33:51] value?
[00:33:52] >> I think the first principles still
[00:33:53] remain the same though. It’s just like
[00:33:56] who is the main person that’s going to
[00:33:57] be using your product? So if you think
[00:33:59] about time to value as an example, do
[00:34:01] humans want to get to value faster in
[00:34:03] your product? The answer is always a
[00:34:04] resounding yes. Like it’s one of those
[00:34:06] first principles like uh Jeff Bezos
[00:34:08] always said with Amazon, it’s like do
[00:34:09] people want to get their packages
[00:34:10] faster? Yes, of course they they do. Now
[00:34:13] maybe there’s a diminishing return when
[00:34:14] it gets to like is it in a minute or is
[00:34:17] it in two minutes where like people are
[00:34:20] like I’m I’m out of I’m not even home.
[00:34:22] It doesn’t matter. Uh so like there is
[00:34:24] some things there. But if you look at
[00:34:26] the other first principle, like do
[00:34:27] people want to get their stuff for
[00:34:30] cheaper and get better quality? It’s
[00:34:31] like of course that anybody will say
[00:34:34] that. Like if I go buy this house for
[00:34:35] like a fraction of what I want to pay
[00:34:36] for it, I’m like amazing. Uh like but
[00:34:39] the owner wants to sell for I would
[00:34:41] totally take it on a discount on that
[00:34:43] end. But u going back to some of the
[00:34:46] first principles, agents uh signing up
[00:34:48] for your product versus humans, it’s
[00:34:50] just the time to value needs to get way
[00:34:52] faster. like a bot is not going to have
[00:34:55] the patience to wait, you know, minutes
[00:34:57] to get to value in your product if it
[00:34:59] takes forever. They want to sign into
[00:35:01] your API. They don’t even want to sign
[00:35:03] up for your product, like go through the
[00:35:05] whole flow using a a fake email or
[00:35:07] anything like that. No, they just want
[00:35:10] to get to the output as fast as
[00:35:12] possible. So, if like you’re a data
[00:35:13] enrichment tool, it’s not going to sign
[00:35:15] up for Zoom info, go through the
[00:35:17] onboarding process, and then finally
[00:35:19] pull a lead list. No, it would love to
[00:35:21] just go in as a connector or MCP or a
[00:35:24] CLI, whatever we decide on the is the
[00:35:27] latest USB connector that everybody
[00:35:28] agrees on and then it is going to just
[00:35:31] pull that info and be like, you want to
[00:35:33] talk to CEOs at SAS companies doing
[00:35:35] these five things that all have free
[00:35:36] trials? Yes. Okay, here’s the top 10 you
[00:35:38] should talk to today. Actually, these
[00:35:40] five are really interesting. You should
[00:35:41] talk to them because they’re actually
[00:35:42] have been on your site in the last 30
[00:35:44] days. So, here’s what you should reach
[00:35:45] out to with them about. That’s what we
[00:35:48] want as humans on the other interacting
[00:35:51] side of it. And we all know like
[00:35:53] whenever like we use cloud or chip or
[00:35:56] something and it takes forever to think.
[00:35:57] We’re like okay twiddling our thumbs or
[00:35:59] doing something else. They’re losing
[00:36:00] engagement on this whole process. So
[00:36:02] it’s like the faster we can get them,
[00:36:04] the faster everything happens and the
[00:36:07] more you win on that level. So nothing
[00:36:09] has changed really. It’s just uh who is
[00:36:11] interacting with it and how they can
[00:36:13] interact with it has changed. But I
[00:36:15] think the first principles are are all
[00:36:17] still the same.
[00:36:17] >> So Wes, what’s the hardest lesson you’ve
[00:36:19] learned that you would pass on to a
[00:36:21] founder just starting out or one
[00:36:23] struggling to grow their company in the
[00:36:24] time of AI?
[00:36:25] >> Yeah, this kind of goes back to the
[00:36:27] analysis of the Brexit’s list of 50
[00:36:29] fastest growing companies. When I kind
[00:36:31] of came into this world of software as a
[00:36:34] service, I was really amazed and
[00:36:36] obsessed with productled growth just as
[00:36:38] a goto market model and how amazing you
[00:36:42] could get users to value pretty quickly,
[00:36:44] get them to upgrade. But then, you know,
[00:36:45] after helping over 450 plus companies
[00:36:48] now, when it comes to just how do you do
[00:36:50] product growth, how do you get more
[00:36:51] people to value quickly and self-s serve
[00:36:53] and all that stuff, the thing I realized
[00:36:55] that matters even more than product
[00:36:57] growth is just you got to have a
[00:36:59] magnetic product. You got to have
[00:37:00] something that’s four out of 10 better
[00:37:02] to stand out in this noise. And that’s
[00:37:06] insanely hard. And that’s like the next
[00:37:07] book I’m working on because I’m like,
[00:37:09] this is a tough nut to crack. And this
[00:37:11] is one of the biggest problems we’re
[00:37:13] going to run into. It’s because it’s
[00:37:14] getting easier to build. How do we build
[00:37:16] stuff that actually stands out and it’s
[00:37:19] not always what you think. It’s not
[00:37:20] always you got to have the most funding
[00:37:21] everything else. It’s like we’re going
[00:37:23] back to a period of time where it’s
[00:37:25] about being a little more opinionated.
[00:37:27] When software first came out, it was all
[00:37:29] about, hey, let’s let’s make it easy to
[00:37:31] customize this tech so you can do
[00:37:32] whatever you want. Now it’s like hey
[00:37:34] look this is my way of doing LinkedIn
[00:37:36] for business growth experts whatever you
[00:37:39] want to call it and this is the product
[00:37:40] how you you can use my approach and
[00:37:43] somebody who’s like a writer or somebody
[00:37:46] who has some other kind of business they
[00:37:47] might hate it but for that niche that
[00:37:51] very specific niche they’re like this
[00:37:53] opinionated approach fits me it writes
[00:37:55] like me I love it and so we’re I want to
[00:37:58] see more opinionated products out there
[00:38:00] that really stand out board. They’re
[00:38:03] very specific people. So, I think this
[00:38:05] is going to be the like a bit of a
[00:38:06] revolution in the sense that there’s
[00:38:08] going to be so many more software
[00:38:10] companies out there than ever before in
[00:38:12] the next 5 to 10 years. The the long
[00:38:14] live the micro SAS, there will be big
[00:38:16] SAS that come out of this too, but there
[00:38:18] has never been a more interesting time
[00:38:21] to be a product entrepreneur in my my
[00:38:23] world as well. So, that’s the the
[00:38:26] biggest advice I would give is just like
[00:38:27] really focus on building that product
[00:38:29] that stands out that is four to 10
[00:38:30] better. And what was in this entire
[00:38:32] process that you learned the hard way?
[00:38:35] >> We I was talking to one client uh a few
[00:38:38] weeks ago and he’s like Wes like product
[00:38:40] growth is honestly the like the best
[00:38:41] business model ever ever. He is like
[00:38:45] entrepreneur who has done like many
[00:38:47] hundreds of millions just like sold many
[00:38:49] many companies done all the different
[00:38:50] kinds of business models. Like this is
[00:38:52] honestly the best one but Wes your
[00:38:54] business not that scalable. Like even if
[00:38:56] you, you know, hire a 100 more
[00:38:58] implementers, you’re still going to run
[00:38:59] into the problem. Whereas like you can
[00:39:01] train people maybe up to like 80% of
[00:39:03] your competence in some areas and then
[00:39:04] you’re still going to be better than
[00:39:06] everybody at like maybe one area like
[00:39:07] out of those nine components like
[00:39:08] strategy or your model design and you’re
[00:39:11] never going to have that full
[00:39:12] replacement. So it’s like but what could
[00:39:14] you do? You could create something that
[00:39:17] is a lot more scalable like AI agents
[00:39:19] that are specialized in each of those
[00:39:20] core areas. They could get people there.
[00:39:22] So it’s like how can I falling into this
[00:39:25] is uh I focus so much on helping other
[00:39:27] people’s businesses instead of focusing
[00:39:29] on my own business of like what is the
[00:39:31] scalable version of productled that will
[00:39:33] scale expertise in a way that helps a
[00:39:35] lot more people and creates a insanely
[00:39:37] useful tool.
[00:39:38] >> Well before we wrap up so you know
[00:39:40] coming back to our topic why AI spending
[00:39:42] fails without productled thinking. So a
[00:39:44] do you think that AI spending fails? Do
[00:39:47] you agree or disagree and why? I mean,
[00:39:50] if you’re a sales company using AI, I
[00:39:51] don’t think that that applies. Like, it
[00:39:53] sure you can still get a lot of benefit.
[00:39:55] There’s still a ton of waste. There’s
[00:39:56] probably there’s actually I know this
[00:39:58] for a fact. There’s way more waste of
[00:40:00] time in a sales like company than a
[00:40:01] product like one just because the nature
[00:40:03] of who delivers the value, who monetizes
[00:40:05] the value and everything else. It’s all
[00:40:07] humanled versus a product. So, um you’re
[00:40:09] already at a disadvantage there. But for
[00:40:11] product companies, does AI investment
[00:40:13] fail? No. Like there is going to be a if
[00:40:16] you focus on the core principles like
[00:40:18] okay using AI to acquire customers
[00:40:20] better how can we make that better
[00:40:22] experience using AI to activate users
[00:40:24] better and faster you’re going to see an
[00:40:26] immediate ROI on that using AI to
[00:40:28] monetize people better okay customize
[00:40:30] the offers customize the the outreach
[00:40:32] emails all those things like you’re
[00:40:33] going to see an investment uh and a
[00:40:35] return on that expansion same thing but
[00:40:38] if you don’t have the right product
[00:40:39] principles this is where it gets
[00:40:41] interesting if you don’t know what
[00:40:42] you’re doing in each of those areas is
[00:40:44] you can just like that report from MIT
[00:40:46] you said you can waste 80 90% of the
[00:40:49] spend and that’s where I would really
[00:40:52] recommend
[00:40:53] learning some of those first principles
[00:40:55] whether that’s reading the product
[00:40:56] playbook or the product growth book to
[00:40:58] just understand like how does this
[00:40:59] actually work
[00:41:00] >> well thank you so much for this sharp
[00:41:02] and grounded conversation I think
[00:41:03] there’s a lot that our listeners are
[00:41:05] going to take away from this entire
[00:41:06] conversation thank you for your time no
[00:41:08] >> worries thanks for having
[00:41:11] >> [music]
