SaaS Maturity, AI Hype, and Pricing Models with Ryan Morris and Seth Robinson

[00:00:12] [SPEAKER_01]: Is SaaS slowing down or is it maturing? What's happening to pricing models? Is AI a bubble

[00:00:21] [SPEAKER_01]: about to burst or just going through the regular hype cycle? I invite two analyst colleagues,

[00:00:27] [SPEAKER_01]: Ryan Morris and Seth Robinson, to join me today as we're going to wrestle with these questions.

[00:00:33] [SPEAKER_01]: Welcome to the Business of Tech Lounge, the live version of the Business of Tech podcast.

[00:00:38] [SPEAKER_01]: It's Wednesday, August 21st, 2024, and I'm Dave Solvill. We'll be taking questions and comments

[00:00:44] [SPEAKER_01]: throughout the show so make sure to put them in the chat. We're going to lean in on being

[00:00:48] [SPEAKER_01]: interactive so make sure to throw them into the chat area if you've got comments or thoughts

[00:00:53] [SPEAKER_01]: you want us to react to. Now I want to thank Sales Builder, our Patreon sponsor, whose support

[00:00:59] [SPEAKER_01]: makes this show possible. Focus on your IT sales workflow with the power of automation

[00:01:04] [SPEAKER_01]: and visit them at salesbuilder.com. That's B-U-I-L-D-R dot com. And a reminder,

[00:01:11] [SPEAKER_01]: I am distinctly watching that chat and we want to hear from you. Now let's set the ground

[00:01:17] [SPEAKER_01]: work a little bit. I've got three articles I want to talk about with you, the audience

[00:01:21] [SPEAKER_01]: first, and we'll bring on our experts. First, Vendors SaaS Report, which gave a little bit

[00:01:27] [SPEAKER_01]: of information about what's going on in the marketplace. They're reporting that annual

[00:01:31] [SPEAKER_01]: contract value, or ACV, has increased less than 4% quarter-over-quarter with SaaS spending.

[00:01:39] [SPEAKER_01]: However, the year-over-year increase is significant at 33%. Q2's ACV is the highest

[00:01:46] [SPEAKER_01]: quarterly ACV on Vendor since Q4 of 2022, when net new ACV was $29,000. Now renewal ACV has

[00:01:56] [SPEAKER_01]: remained relatively flat for the past year. This quarter it decreased slightly quarter-over-quarter

[00:02:03] [SPEAKER_01]: and year-over-year, indicating that buyers have an advantage in renewal pricing right now.

[00:02:10] [SPEAKER_01]: And second, from CIO Dive. 9 in 10 survey companies admitted they can't trace at least

[00:02:17] [SPEAKER_01]: 10% of their total cloud spend to the right source. Of the 39% of respondents with cost

[00:02:24] [SPEAKER_01]: management processes in place, just over one quarter have optimized software code and less

[00:02:30] [SPEAKER_01]: than half have enabled chargeback processes that deduct cloud spend from demarc mental budgets.

[00:02:36] [SPEAKER_01]: And finally, from TechCrunch. The net retention reality in the market is evolving in a manner

[00:02:42] [SPEAKER_01]: that appears pretty tough for software companies, both large and small. NDR rates are slipping all

[00:02:49] [SPEAKER_01]: over the software landscape, meaning that a lot of software companies are seeing their

[00:02:53] [SPEAKER_01]: growth rates decline, not due to their inability to sell to new customers, or not

[00:03:00] [SPEAKER_01]: problem, but because their existing customers are not buying as much as they used to.

[00:03:07] [SPEAKER_01]: So let me bring on Ryan Morris, Principal Consultant at Morris Management and co-host

[00:03:12] [SPEAKER_01]: on Killing It, and Seth Robinson, Vice President of Industry Research from CompTIA.

[00:03:18] [SPEAKER_01]: Guys, thanks for joining me today. Hey Dave, thanks for having me. Thanks for having us on,

[00:03:25] [SPEAKER_01]: let's start with this headline here. SAS, is it maturing, slowing down or do we have a problem

[00:03:32] [SPEAKER_01]: here? Do you guys have a particular take on what we're seeing? I do. In fact, I will say

[00:03:38] [SPEAKER_02]: this is two factors, right? Number one, the customer is maturing as you've been indicating

[00:03:45] [SPEAKER_02]: when customers can't figure out what they're spending, where they're spending it,

[00:03:50] [SPEAKER_02]: why they're spending it. That's not a problem with the model, that's a question of

[00:03:55] [SPEAKER_02]: the buyer not being prepared for the model, not being professional about the way that they

[00:04:01] [SPEAKER_02]: go about it. It is inevitable that the customer will in fact mature and learn not only to

[00:04:08] [SPEAKER_02]: measure what they're doing effectively, but to control. Now when you hear controlling spend

[00:04:14] [SPEAKER_02]: from a customer's perspective, that sounds like value. From a vendor's perspective, I think this

[00:04:20] [SPEAKER_02]: is the bigger issue. It feels like a threat to their growth because they almost believe,

[00:04:26] [SPEAKER_02]: based on my experience, that growth is something that they are entitled to. I run a SAS model,

[00:04:33] [SPEAKER_02]: I can land and I can expand. Expansion should drive more revenue growth than land

[00:04:39] [SPEAKER_02]: and that's just my entitlement because I happen to deploy this business model. That's not true.

[00:04:47] [SPEAKER_02]: Expansion will happen if the customer actually believes that they are getting value from the

[00:04:53] [SPEAKER_02]: tools that they are paying for. If they're shrinking their spend, it's not because

[00:04:58] [SPEAKER_02]: only the economic factors, it's because they don't believe they're getting more value out

[00:05:04] [SPEAKER_02]: than cost that they are putting in. That happens in every economic cycle and if we either

[00:05:10] [SPEAKER_02]: don't deliver value or can't demonstrate that we are delivering value, then the answer is try

[00:05:18] [SPEAKER_02]: harder. That's not the customer's fault for cutting the budget. That's on the vendor for

[00:05:23] [SPEAKER_02]: needing to demonstrate and deliver real value. Yeah, I would agree with that. I think that there

[00:05:31] [SPEAKER_00]: are some issues with the model that are continuing to be worked out by a lot of end

[00:05:37] [SPEAKER_00]: user customers. I think the other thing that I would add is I step back a level and take a look

[00:05:43] [SPEAKER_00]: at this. Ever since we started talking about SAS, there was something that stuck in my brain

[00:05:48] [SPEAKER_00]: and I think it finally came out when we were talking more about emerging technology.

[00:05:53] [SPEAKER_00]: We talk about SAS as its own industry sometimes from the vendor perspective side of it

[00:06:00] [SPEAKER_00]: and I think that there is good reasons for those discussions but from an end user

[00:06:05] [SPEAKER_00]: perspective, I don't think a CIO ever says I have a SAS strategy. They have a workflow

[00:06:13] [SPEAKER_00]: strategy. They have applications that they need to plug in and one thing that I don't think

[00:06:19] [SPEAKER_00]: got pulled out in any of those articles, Dave, that you mentioned but I think is part of the

[00:06:24] [SPEAKER_00]: reality here is the SAS model has led to an explosion in the types of software that people

[00:06:32] [SPEAKER_00]: can get for the various parts of their workflow. We've all seen those pictures of here's what the

[00:06:39] [SPEAKER_00]: MarCom environment looks like and for one, there are 17 categories that you don't even know

[00:06:44] [SPEAKER_00]: there were 17 categories of MarCom and then inside each one of those categories, you've got

[00:06:49] [SPEAKER_00]: so many different vendors that you have to break out the magnifying glass and start looking

[00:06:53] [SPEAKER_00]: at them and it's kind of questionable how much of that the market will bear. So for one,

[00:07:01] [SPEAKER_00]: how much granularity do we need for every piece of the workflow and then how much granularity can

[00:07:07] [SPEAKER_00]: an end user even absorb as they're trying to digitize process, as they're trying to get their

[00:07:13] [SPEAKER_00]: workforce to adopt all of these new ways of doing things. I think there are some realities

[00:07:18] [SPEAKER_00]: in that rollout of all these products and this attempt to digitize every niche part of the

[00:07:26] [SPEAKER_00]: workflow that are beginning to impact the market here and I think very likely impacting these

[00:07:31] [SPEAKER_00]: numbers that we're seeing. All right, now you both said it in slightly different ways and I want to

[00:07:36] [SPEAKER_01]: ask about this. Seth, you just go there's issues with the rollout. Ryan, you said that the

[00:07:41] [SPEAKER_01]: buyer isn't ready. Am I blaming the solution provider here? Let's break this down. Whose

[00:07:48] [SPEAKER_01]: job is it to get the buyer ready? Is it a matter of the SAS provider, the software

[00:07:54] [SPEAKER_01]: provider has to shoulder all of that? Isn't that what the integrator, deliverable managed

[00:08:00] [SPEAKER_01]: services provider, you kind of both said it in different ways. I'm saying this because

[00:08:08] [SPEAKER_01]: I think there's an opportunity but where's the problem here and who should own the problem?

[00:08:13] [SPEAKER_02]: Well, I think in every example of a value chain that you've ever examined,

[00:08:18] [SPEAKER_02]: it is a shared responsibility where your interest as the vendor and my interest as the solution

[00:08:25] [SPEAKER_02]: provider cannot be separated because they can't be achieved independently of one another.

[00:08:31] [SPEAKER_02]: If the software is good but the implementation is poor, neither one of them is getting paid.

[00:08:37] [SPEAKER_02]: If the vice versa is true, software is poor but there's a brilliant implementation,

[00:08:42] [SPEAKER_02]: eventually neither of those participants are actually going to get paid. It is on the seller

[00:08:49] [SPEAKER_02]: to teach the customer how and why to buy the things the way we want to sell them and

[00:08:55] [SPEAKER_02]: to enable them to do so successfully because otherwise, I mean it's not the customer's job

[00:09:00] [SPEAKER_02]: to figure out how to buy what you sell or they'll just go buy something from somebody else.

[00:09:05] [SPEAKER_02]: Right? It's our job if we want them to pay us that we need to execute. Yes, that is on

[00:09:12] [SPEAKER_02]: the vendor at the market and at the go-to-market strategy levels but it is also on the implementer,

[00:09:18] [SPEAKER_02]: on the solution provider to educate the customer, to support the customer,

[00:09:23] [SPEAKER_02]: to actually demonstrate to them this is what you bought, this is how much you paid for it

[00:09:28] [SPEAKER_02]: and then here's how much value you got in return. If we can't demonstrate that,

[00:09:34] [SPEAKER_02]: don't blame the customer for saying I don't see the value here. That's our job.

[00:09:39] [SPEAKER_00]: Yeah, I think there are maybe two different problems going on here, right? And one is

[00:09:45] [SPEAKER_00]: if you're a vendor and you're creating this feature-rich piece of software for recruiting

[00:09:52] [SPEAKER_00]: talent, is there a market for that? Is there enough of a market for you to have a growth

[00:09:59] [SPEAKER_00]: path? I think that all of the different niches that we're trying to carve out and all the

[00:10:07] [SPEAKER_00]: niches, I struggle to believe that there's going to be enough appetite for all of these things

[00:10:13] [SPEAKER_00]: even in the medium to long term. And then when it comes to the integration, I think that's a

[00:10:19] [SPEAKER_00]: joint problem, right? So you're not only selling your piece of software, you have to know

[00:10:25] [SPEAKER_00]: how that interacts with the other pieces of software that are being sold by your competitors

[00:10:30] [SPEAKER_00]: in the same space or other software providers that aren't competing in the same space but

[00:10:35] [SPEAKER_00]: part of that workflow. And I think some of that is on the vendor to be aware of that and to be

[00:10:40] [SPEAKER_00]: able to kind of draw that out for the end user so that they understand that total cost of

[00:10:45] [SPEAKER_00]: ownership or the total cost of implementation across that workflow. I would say maybe the

[00:10:52] [SPEAKER_00]: lion shared that they could defer to a managed service provider or something like that. A vendor

[00:10:57] [SPEAKER_00]: is not going to necessarily step in and be a consultant across that entire workflow, but

[00:11:03] [SPEAKER_00]: some awareness of being able to point the customer in the right direction is probably

[00:11:07] [SPEAKER_01]: going to go a long way. Now, interestingly, I also and Seth, I'm going to throw this to

[00:11:12] [SPEAKER_01]: you first because in a way you may have already defeated my basic premise here by saying like

[00:11:17] [SPEAKER_01]: because you're right to acknowledge like SaaS isn't its own industry. It's interesting

[00:11:21] [SPEAKER_01]: as you said that as I was processing this thing. Yeah, you know, he's right. We kind

[00:11:25] [SPEAKER_01]: of break it out like it's an industry but it isn't. It's a delivery model of software

[00:11:29] [SPEAKER_01]: in various industries, which makes much more sense. But at the same time, I want to ask

[00:11:34] [SPEAKER_01]: you your thoughts on like some of these feel like symptoms of maturity that we've

[00:11:41] [SPEAKER_01]: gotten. We've been doing this long enough that, you know, numbers can't grow forever.

[00:11:45] [SPEAKER_01]: Like that's one of those basic economic problems is the bigger they get the harder

[00:11:50] [SPEAKER_01]: they are for growth. At some point you have tapped into most of a end user's budget for

[00:11:58] [SPEAKER_01]: solving a particular set of problems. So for example, the idea of I'm not going to

[00:12:02] [SPEAKER_01]: continually buy more on my CRM software if I've solved that problem and it's delivered

[00:12:08] [SPEAKER_01]: by SaaS. Yes, I'm going to see growth from users but I may not necessarily buy see

[00:12:12] [SPEAKER_01]: growth from other things because it's maturity. How much of this do you think is a symptom

[00:12:18] [SPEAKER_00]: of maturity? I think that the cost part of it is definitely tied to maturity. You

[00:12:25] [SPEAKER_00]: mentioned this earlier too, right? That we're still figuring out the cost structure across

[00:12:32] [SPEAKER_00]: all cloud delivery models. You know, SaaS information or infrastructure as a service

[00:12:36] [SPEAKER_00]: platform as a service. You know, we hear a lot. We read a lot about companies that

[00:12:41] [SPEAKER_00]: have to get more involved in FinOps or network administrators that have to get

[00:12:46] [SPEAKER_00]: more involved in FinOps because they have to watch those costs in a much different way than

[00:12:51] [SPEAKER_00]: they used to watch them when it was just capital expenditure, right? So I think that

[00:12:56] [SPEAKER_00]: learning that cost structure is part of the maturing process. And I think another part of

[00:13:01] [SPEAKER_00]: that maturing process is if you're a vendor and you're trying to pack more of these features

[00:13:07] [SPEAKER_00]: in and especially if they're AI features which I think we're going to get to then

[00:13:12] [SPEAKER_00]: you're building more and more cost in the computation, the compute platform that you have

[00:13:18] [SPEAKER_00]: to provide the software and you're probably just not going to absorb all that cost. You know,

[00:13:23] [SPEAKER_00]: you're going to need to pass it along to the end user at some point and they're eventually

[00:13:28] [SPEAKER_00]: going to break. You know, again, if they've gone from 10 different applications to 100

[00:13:33] [SPEAKER_00]: different applications or something like that and then every single one of those 100

[00:13:37] [SPEAKER_00]: applications starts ratcheting up the costs a little bit more, that becomes a huge problem

[00:13:41] [SPEAKER_00]: for the end user, right? And I don't know that the vendors have a lot of levers to pull

[00:13:48] [SPEAKER_00]: to control their costs. I think that's part of this overall market that everyone needs to

[00:13:54] [SPEAKER_00]: understand what is the overall appetite and the overall willingness to bear cost

[00:14:01] [SPEAKER_00]: across our entire architecture and our entire workflow. Well, interestingly, that article from

[00:14:06] [SPEAKER_01]: CIO Dive exactly highlighted what you're just talking about in terms of managing costs. Now,

[00:14:11] [SPEAKER_01]: we could spend a ton of time there, but I actually want to you brought up AI. It was coming

[00:14:16] [SPEAKER_01]: shortly and I wanted to link that trend to what we're talking about. And I want to actually talk

[00:14:21] [SPEAKER_01]: a little bit about pricing models around AI. And we're going to ask the question, is per seat

[00:14:28] [SPEAKER_01]: broken? Now, I'm going to borrow from a newsletter that I highlighted on last week's

[00:14:34] [SPEAKER_01]: Weekend Big Ideas section where I dug into this particular article that talks about the

[00:14:39] [SPEAKER_01]: various ways that we can go to market and focusing on the fact that in a world of AI, it may be

[00:14:44] [SPEAKER_01]: broken. And quoting the article, in a world like this where seats may reduce or in some cases be

[00:14:50] [SPEAKER_01]: completely orthogonal from the value created by SaaS, it's clear there will be at least very

[00:14:56] [SPEAKER_01]: least be some additions if not wholesale changes in pricing models. Now, the premise here is

[00:15:04] [SPEAKER_01]: that if you think about AI, one of the big cost pieces is the fact that compute and the

[00:15:11] [SPEAKER_01]: cost of the query is much more important. And so you can't necessarily let all the users in

[00:15:16] [SPEAKER_01]: and charge per user or per seat because your cost is compute. Now as we infuse AI into everything,

[00:15:24] [SPEAKER_01]: as we make data analytics much more powered by AI, I'm going to ask you both like, are we

[00:15:30] [SPEAKER_01]: about to truly break the pricing models that we've been using in a SaaS world for the past 10 or 15

[00:15:36] [SPEAKER_02]: years? Yes. And in fact, I think those of us who have lived through the evolution of managed

[00:15:43] [SPEAKER_02]: services could see this coming from a while ago. When we were born as MSPs, we looked around

[00:15:50] [SPEAKER_02]: and said, well, how many users do you have? And what work will I do for them? And at A

[00:15:55] [SPEAKER_02]: equals the actual monthly recurring cost. And then we realized, you know what? A lot of the value

[00:16:01] [SPEAKER_02]: we provide is not for the user. It was in the infrastructure. It was in the systems and the

[00:16:06] [SPEAKER_02]: platforms that were under administration. And you couldn't just say this is the cost per server

[00:16:13] [SPEAKER_02]: because a server in isolation doesn't do anything. You successfully managed the server.

[00:16:19] [SPEAKER_02]: Congratulations. But the network was down. It's kind of like having a great car that doesn't

[00:16:23] [SPEAKER_02]: have any gas in the tank. We evolved to look at a different way of pricing. And I believe that is

[00:16:30] [SPEAKER_02]: what will happen with SaaS. We spend a tremendous amount of time in hard goods industries. Think

[00:16:36] [SPEAKER_02]: about manufacturing, industrial production, fuel and oil and energy, all of these kinds of

[00:16:44] [SPEAKER_02]: environments where the value to the organization is not bound in terms of the number of seats

[00:16:50] [SPEAKER_02]: of people who administer those systems. It's the factory floor. It's the OT as well as the IT.

[00:16:58] [SPEAKER_02]: And we will have to be prepared to demonstrate not just that we can provide value, but what

[00:17:07] [SPEAKER_02]: is the unit measurement? We can't just go to a customer and say you've got four admins and

[00:17:12] [SPEAKER_02]: each one of those costs $100,000 a year. They will puke on us. We have to be able to say,

[00:17:18] [SPEAKER_02]: this is your organization. And as Seth described, this is the workflow that is being improved.

[00:17:25] [SPEAKER_02]: Here is the measurable, tangible value to your organization as a result.

[00:17:30] [SPEAKER_02]: This is how much you should pay to receive that value.

[00:17:34] [SPEAKER_00]: Yeah, Dave, you asked the question of whether AI was going to change the pricing model that

[00:17:38] [SPEAKER_00]: we've been seeing in SaaS. But I think Ryan kind of pulled it out a little bit there.

[00:17:42] [SPEAKER_00]: This is the pricing model that we had prior to SaaS just talking about managed services.

[00:17:46] [SPEAKER_00]: It was like, how many seats? How many units am I serving here? And within a managed services

[00:17:53] [SPEAKER_00]: discussion, we've been having a talk for quite a while now about outcome-based pricing.

[00:17:59] [SPEAKER_00]: And I think we see this internally for end users. A CIO has to be thinking strategically

[00:18:05] [SPEAKER_00]: now and has to be thinking about what am I delivering because there's more investment

[00:18:10] [SPEAKER_00]: happening in technology. The company, the organization realizes that they want to invest

[00:18:15] [SPEAKER_00]: more to go through digital transformation or to be more on the cutting edge or to accelerate adoption

[00:18:21] [SPEAKER_00]: of emerging technology. But it can't just be that I kept the lights on anymore. I have to

[00:18:26] [SPEAKER_00]: be producing some kind of number, and it's not always going to be an ROI number, especially

[00:18:31] [SPEAKER_00]: with something like cybersecurity. There's not going to be a return on that, but there have

[00:18:36] [SPEAKER_00]: to be new metrics. And I think that that's what we're talking about here is what are the

[00:18:42] [SPEAKER_00]: metrics of technology and the technology purchase and procurement process.

[00:18:49] [SPEAKER_00]: As we're starting to talk about AI, I do think that AI is changing some of that.

[00:18:55] [SPEAKER_00]: I also think that AI is providing pretty convenient air cover for companies who are

[00:19:00] [SPEAKER_00]: looking at it and saying, oh, geez, we went crazy on the number of seats that we handed out

[00:19:04] [SPEAKER_00]: to people. And we're going to ratchet back on that now. We can run leaner, maybe not

[00:19:09] [SPEAKER_00]: reducing workforce, but just reducing the number of people that have to hold these licenses.

[00:19:15] [SPEAKER_00]: And they can say that there's AI and that makes them sound smarter. And maybe there is really

[00:19:20] [SPEAKER_00]: some AI happening. But I think there's also just kind of this right sizing of some investments

[00:19:26] [SPEAKER_00]: that went a little out of control. Interestingly, I literally just covered

[00:19:30] [SPEAKER_01]: some story on the show last week where the use of the phrase AI actually reduces the

[00:19:37] [SPEAKER_01]: likelihood of a buyer because they are become much more skeptical of the product.

[00:19:41] [SPEAKER_01]: So there's a certain degree of like, don't lean too much into AI.

[00:19:45] [SPEAKER_01]: The reason I wanted to bring up this conversation, particularly for listeners is

[00:19:48] [SPEAKER_01]: I don't think this is an imminent problem. I don't think that this is something that's looming

[00:19:52] [SPEAKER_01]: in say the next quarter or two. But I do think we want to collectively as an industry

[00:19:58] [SPEAKER_01]: and some thought leaders here to open this conversation of, I think we're going to end

[00:20:03] [SPEAKER_01]: up in a place where some of this compute power and the way you pay for it won't match

[00:20:08] [SPEAKER_01]: the way the channels put this together. Now, if we think about AI delivered from say an open AI

[00:20:14] [SPEAKER_01]: or an anthropic, the way that it's delivered based on compute and it's billed through as

[00:20:20] [SPEAKER_01]: compute that kind of gets broken the moment you immediately try and turn that into per user

[00:20:25] [SPEAKER_01]: pricing. Somebody like Microsoft, well they can absorb that with full pilot, but your typical

[00:20:32] [SPEAKER_01]: small and mid market service provider is not going to be able to absorb that. And I think they're

[00:20:37] [SPEAKER_01]: going to have to start thinking about a consumption model that looks a little bit more like,

[00:20:42] [SPEAKER_01]: you know, like that compute model. But let me ask then for you guys as you're starting

[00:20:47] [SPEAKER_01]: to gaze into your crystal balls, how much do you think users are going to be kind of

[00:20:54] [SPEAKER_01]: turning up and down the usage of AI in the way that we might think about it from the way

[00:20:58] [SPEAKER_01]: developer might use it? Or how much do you think it's going to be much more of a really static

[00:21:05] [SPEAKER_01]: definable thing that's delivered within a product? Seth, do you want to dive in there?

[00:21:11] [SPEAKER_00]: Yeah, you know, you and I talked about AI, I think on a previous, you know,

[00:21:16] [SPEAKER_00]: lounge show. I've been looking at it for a while. I think that we've had AI, you know,

[00:21:22] [SPEAKER_00]: around for several years, not just the past couple with generative AI.

[00:21:27] [SPEAKER_00]: Companies have been playing around with it, they've been learning. And ultimately,

[00:21:31] [SPEAKER_00]: I think a lot of AI is just going to come as a feature of software that companies are already

[00:21:39] [SPEAKER_00]: buying. They're not going to buy AI specific software. Most companies are not going to be

[00:21:43] [SPEAKER_00]: building their own LLMs or writing their own algorithms, right? AI is going to be baked into

[00:21:48] [SPEAKER_00]: your CRM and your finance software and your business productivity. It already is in a lot

[00:21:52] [SPEAKER_00]: cases, right? And if that's the case, then AI is maybe not just another feature. Maybe it's one

[00:22:01] [SPEAKER_00]: of the biggest best features we've had in a while, but it is another feature, right?

[00:22:05] [SPEAKER_00]: And this almost makes me think back to when we talked about 5G so much and everyone was like,

[00:22:09] [SPEAKER_00]: you know, look at what we can do with 5G. And I always say, well, what are you currently

[00:22:13] [SPEAKER_00]: doing with 4G, right? Like everything that you're talking about is great, but it's

[00:22:17] [SPEAKER_00]: theoretical, right? And it should build on top of what you're doing. And so back to our

[00:22:22] [SPEAKER_00]: previous conversation, if we are already wrestling with feature-rich software packages

[00:22:27] [SPEAKER_00]: where users aren't using all the features, how well are they going to be able to unlock

[00:22:33] [SPEAKER_00]: that AI, right? And maybe the AI does some of it for them so that they don't have to

[00:22:37] [SPEAKER_00]: be reaching in and doing it or prompting it or doing whatever. But if it's just another

[00:22:42] [SPEAKER_00]: feature, we're already dealing with features that are kind of laying by the wayside dead.

[00:22:48] [SPEAKER_00]: So I think it'll be really interesting to see how that plays out.

[00:22:52] [SPEAKER_02]: See, I buy that philosophy completely. You can't say that there's value in AI just like you can't

[00:23:00] [SPEAKER_02]: discreetly say that there is value in compute or in storage or in any other independent

[00:23:07] [SPEAKER_02]: functionality, discrete capability of a piece of technology. The question we must always come back

[00:23:13] [SPEAKER_02]: to is to accomplish what, right? From the business's perspective, what exactly am I using

[00:23:20] [SPEAKER_02]: compute or AI to accomplish? If I'm using that to close my books faster at the end of the month

[00:23:27] [SPEAKER_02]: and I can get more accuracy with less manpower and I can give more confidence to my investors

[00:23:33] [SPEAKER_02]: because I've used AI to enhance the way I analyze and calculate my financials, cool.

[00:23:40] [SPEAKER_02]: That means you got better at financials, not that you got better at AI. It's the same in any

[00:23:46] [SPEAKER_02]: of the functional things that we do. And Dave, to tie these things together for you, it is,

[00:23:52] [SPEAKER_02]: I don't believe compute is the value that people think that it is. Just like electricity and the

[00:23:59] [SPEAKER_02]: original utility model is not in itself a valuable commodity. What are you using it for?

[00:24:06] [SPEAKER_02]: What will be the outcome that it accomplishes in your organization? Are you using it to run

[00:24:12] [SPEAKER_02]: a factory floor? Are you using it to power the lights in the office, to heat the actual facility?

[00:24:19] [SPEAKER_02]: Electricity on its own is necessary, but it does not equal value. The same is true for compute.

[00:24:26] [SPEAKER_02]: The same is true for AI. What are you actually trying to do? You said, I think we need to get

[00:24:32] [SPEAKER_02]: to the idea of a consumption model. I think once and for all, we need to get to the idea

[00:24:38] [SPEAKER_02]: of an outcome model where people are not paying us for the inputs that we deliver,

[00:24:45] [SPEAKER_02]: but rather for the outcomes that we enable them to achieve. There's risk in that. That's why we're

[00:24:51] [SPEAKER_02]: not there yet. People are a little bit afraid of that idea because what if my technology

[00:24:56] [SPEAKER_02]: doesn't produce value? The customer might actually stop paying me for it.

[00:25:02] [SPEAKER_02]: Welcome to the real world. If it doesn't provide valuable outcomes that are recognizable,

[00:25:08] [SPEAKER_01]: why would we pay extra for those things? Well, in risk there is the reward. Those that are

[00:25:13] [SPEAKER_01]: going to start figuring that out are going to do something. I'm going to ask this because it's

[00:25:17] [SPEAKER_01]: the question that's dominating the news cycle and I figured I'd let us all the way in on this

[00:25:22] [SPEAKER_01]: one. Let's look at whether or not we're actually about to hit an AI bubble or are we

[00:25:26] [SPEAKER_01]: in the natural hype cycle? I'm going to pull from CNN on this one, which actually had a

[00:25:31] [SPEAKER_01]: great article covering this whole space. Quoting from the article, in short, investors'

[00:25:37] [SPEAKER_01]: fears can be boiled down to this. Is all of this actually worth anything or is it just another

[00:25:43] [SPEAKER_01]: shiny object the industry is chasing to bring back its dreams of endless growth before it

[00:25:49] [SPEAKER_01]: abandons it and moves on to the next big thing? Guys, I got my answer but I want

[00:25:54] [SPEAKER_01]: to hear from you first. Are we hitting an AI bubble or is this the natural hype cycle?

[00:26:01] [SPEAKER_00]: Yeah, I'll jump in first because in our Outlook report at the end of last year,

[00:26:05] [SPEAKER_00]: our very first trend was that AI hype is going to fade but workflows are going to continue evolving

[00:26:11] [SPEAKER_00]: and the first half of that prediction is like the easiest prediction ever. AI hype is going

[00:26:17] [SPEAKER_00]: to fade. Yeah, eventually sometime it will. It always happens. It has to happen.

[00:26:24] [SPEAKER_00]: People have made entire industries on the fact that hype rises and then it falls.

[00:26:28] [SPEAKER_00]: So it is going to fall. I think the question really is what's that plateau at the end?

[00:26:33] [SPEAKER_00]: I think media, who was largely responsible for pumping this thing up in the first place,

[00:26:39] [SPEAKER_00]: is really quick to say, hey, we're not seeing anything out of this. I think when you dig in

[00:26:45] [SPEAKER_00]: and you talk to people that have been using this for years, not just two years ago when

[00:26:50] [SPEAKER_00]: ChatGPT came out, they are finding ways to transform their workflows. They're finding ways

[00:26:56] [SPEAKER_00]: to improve their outputs. They might not be finding all the ways to measure that delta

[00:27:02] [SPEAKER_00]: quite yet but institutionally it's working for them and as a technology team, they can see

[00:27:09] [SPEAKER_00]: that they're making progress. They probably need to make sure that they're communicating that

[00:27:13] [SPEAKER_00]: as well as they need to to the business side of the house but I think that there is

[00:27:18] [SPEAKER_00]: some real value in this. It was always going to be less than what the bubble got pumped up

[00:27:24] [SPEAKER_00]: to. There's a timing question of when do we go over the top and really start having some

[00:27:30] [SPEAKER_00]: disillusionment and then there's an ultimate plateau value of what is this really going

[00:27:36] [SPEAKER_00]: to look like? That plateau even has some timing in it. It's not like we're going to

[00:27:41] [SPEAKER_00]: get to two years down the road and everyone has figured this out. We just talked about the

[00:27:46] [SPEAKER_00]: fact that companies are still trying to figure out the costs of cloud computing and that's

[00:27:51] [SPEAKER_00]: a topic that I think I talked about eight or ten years ago at conferences. These things

[00:27:57] [SPEAKER_00]: take time. As much as there can be rapid change and rapid improvement in compute capability,

[00:28:04] [SPEAKER_00]: there can be really slow change in behavior and I think that's how it's going to

[00:28:09] [SPEAKER_02]: play out for a little while here. Ryan? Again, we have seen this movie before

[00:28:16] [SPEAKER_02]: and we can predict what the second act is going to look like. I remember a day when there was a

[00:28:22] [SPEAKER_02]: headline that said, is there really any value in this newfangled technology called e-commerce

[00:28:28] [SPEAKER_02]: or is it just a shiny object that the industry wants to sell to you? If you'll recall,

[00:28:34] [SPEAKER_02]: it took Amazon more than ten years to arrive at profitability and there were oodles of

[00:28:40] [SPEAKER_02]: articles written during that period that questioned fundamentally whether they ever would.

[00:28:46] [SPEAKER_02]: I think it worked out okay for those guys over there. Same with Uber, same with every other new

[00:28:53] [SPEAKER_02]: technology that has eventually achieved maturity. Now, let's be grownups. Not every technology

[00:29:01] [SPEAKER_02]: that is launched makes it through the trough of disillusionment and back up to the plateau of

[00:29:06] [SPEAKER_02]: profitability. Some things are fake. Some things are just hype and not anything meaningful beyond

[00:29:13] [SPEAKER_02]: that. But every technology that has ever gotten to maturity had its period of disillusionment where

[00:29:20] [SPEAKER_02]: people said, you know, I'm just not quite sure about that. I'm not seeing the value here.

[00:29:26] [SPEAKER_02]: Well, that's not just a technology seller's problem. That's a user skill problem.

[00:29:32] [SPEAKER_02]: If you give me a circular saw, I know how to turn it on but I don't know how to do much

[00:29:37] [SPEAKER_02]: more than that with it. But that doesn't mean it's not a valuable tool because there

[00:29:41] [SPEAKER_02]: are plenty of people with skills and training who use that same tool to accomplish remarkable

[00:29:48] [SPEAKER_02]: outcomes. What are we actually trying to accomplish with AI and what are the skills that are

[00:29:54] [SPEAKER_02]: necessary to actually empower the users to accomplish those outcomes? I'll say it again

[00:30:01] [SPEAKER_02]: here like we've said it before. I don't believe AI is anywhere near the concept of

[00:30:07] [SPEAKER_02]: replacing humans but it will enhance the capability of humans who spend the time and

[00:30:13] [SPEAKER_02]: energy to get good at it. It is in its natural hype cycle. We get to sound like geniuses when

[00:30:20] [SPEAKER_02]: we say, oh no, things might be a little bit bleak for a while. Will any of these vendors

[00:30:25] [SPEAKER_02]: ever achieve profitability? Not tomorrow. Not in any of the next couple of quarters because

[00:30:32] [SPEAKER_02]: that's just not how markets of transformative technology are created. It goes up the hype

[00:30:39] [SPEAKER_02]: cycle, it comes down the disillusionment cycle and eventually we grow up and we learn how to

[00:30:45] [SPEAKER_02]: use it effectively. Case study, Amazon. I think that there are enough smart people working

[00:30:52] [SPEAKER_02]: on AI that we will find valuable uses and outcomes to be achieved. It's just not there

[00:31:00] [SPEAKER_02]: yet. It doesn't mean it's broken, just means it's new. I love that both of your answers are couched

[00:31:05] [SPEAKER_01]: in value to the end user and to the providers doing it. I'm going to observe it. I'm going

[00:31:10] [SPEAKER_01]: to cheat a little bit with my answer and say the answer is yes. It is both a hype, a bubble

[00:31:17] [SPEAKER_01]: to be burst because the lens of that is generally investors who are looking for you.

[00:31:22] [SPEAKER_01]: They love hype or they love drop because it's a buy or a sell opportunity. From that

[00:31:27] [SPEAKER_01]: perspective, yes. It is also the natural hype cycle. The technologies go through exactly

[00:31:33] [SPEAKER_01]: the cycles that you talk about and what is the trick for us as solution provider advisors,

[00:31:39] [SPEAKER_01]: the three of us and such, is to provide insights for the providers. That's why we

[00:31:45] [SPEAKER_01]: want to look at use cases as heavily as we do. By the way, preview for those of you

[00:31:49] [SPEAKER_01]: that are listening live, there's a ton of use cases on today's episode of The Business of

[00:31:53] [SPEAKER_01]: Tech as I dive into a couple of specific ones, but it is also a bubble when we think about

[00:31:58] [SPEAKER_01]: it from an investor perspective. Guys, this has been super fun and I appreciate you joining me

[00:32:04] [SPEAKER_01]: to mix it up and I look forward to having you both on for some conversations on these

[00:32:09] [SPEAKER_01]: live shows upcoming. Let's do it. Thanks for having us. This has been great guys.

[00:32:16] [SPEAKER_01]: Thank you both for joining and we will have both Seth and Ryan on again on future ones.

[00:32:21] [SPEAKER_01]: Now I want to give you a preview of an upcoming interview that's dropping this weekend.

[00:32:26] [SPEAKER_01]: I spoke with Jonathan Romley, who's an expert on hiring and he's written a book on

[00:32:31] [SPEAKER_01]: international hiring. It's called Winning the Global Talent War, How to Find the World's Best

[00:32:36] [SPEAKER_01]: Employees and Thrive in a Competitive World. We talked about the incoming deluge of

[00:32:43] [SPEAKER_01]: resume applications among other things. I will make sure that you get access to that

[00:32:47] [SPEAKER_01]: and know that that video is available right now. My Patreon supporters already have this video

[00:32:53] [SPEAKER_01]: if you want to listen now. Jonathan dove into how the deluge may actually be a new way for

[00:33:00] [SPEAKER_01]: users to be hacked and to get the organizations with a flood of those uncertain and unknown

[00:33:06] [SPEAKER_01]: resumes and he talks in the interview a little bit about techniques and changes he

[00:33:12] [SPEAKER_01]: makes a big difference in recruiting. That interview will drop on YouTube and the podcast

[00:33:17] [SPEAKER_01]: feed this weekend and if you're interested, I encourage you to listen. You can get access

[00:33:21] [SPEAKER_01]: to it early now at patreon.com slash MSP radio. Now I want to thank Sales Builder,

[00:33:27] [SPEAKER_01]: our Patreon sponsor, whose support makes this show possible. Focus on your IT sales

[00:33:32] [SPEAKER_01]: workflow with the power of automation and visit them at salesbuilder.com. That's buildr.com

[00:33:40] [SPEAKER_01]: and vendors. You too can get your name mentioned on the live show. It's a simple monthly

[00:33:44] [SPEAKER_01]: subscription visit patreon.com slash MSP radio to learn more and listeners. You can support the

[00:33:51] [SPEAKER_01]: show. The best way to do that is like share and follow on your favorite platforms. If you

[00:33:57] [SPEAKER_01]: let someone know about the show, it's the number one thing you can do to help and you

[00:34:02] [SPEAKER_01]: can support directly with our give what you want model. You set what you think the content

[00:34:07] [SPEAKER_01]: is worth and you get access to interviews early access to one on ones with me and special

[00:34:13] [SPEAKER_01]: content drops periodically. If you have a question and are listening to the recording,

[00:34:17] [SPEAKER_01]: send it in at question at MSP radio.com. Thanks for joining me for the business of

[00:34:23] [SPEAKER_01]: tech lounge and I will see you next time.