In this episode of the Six Five Podcast Enterprising Insights, host Keith Kirkpatrick discusses the current and future landscape for enterprise application pricing models, which are being impacted by the use of AI agents, generative AI, and the growing demand for linking software expenditures to ROI. He then moves to the Rant or Rave segment, where he picks one item in the market and either champions or criticizes it.
You can grab the video here and subscribe to our YouTube channel if you’ve not yet done so.
Listen to the audio below:
Disclaimer: The Enterprising Insights podcast is for information and entertainment purposes only. Over the course of this podcast, we may talk about companies that are publicly traded and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors and we do not ask that you treat us as such.
Transcript:
Keith Kirkpatrick: Hello everyone. I’m Keith Kirkpatrick, Research Director with The Futurum Group, and I’d like to welcome you to Enterprising Insights. It’s our weekly podcast that explores the latest developments in the enterprise software market and the technologies that underpin these platforms, applications and tools. This week, I’d like to discuss the current and future landscape for enterprise application pricing models, which are currently being impacted by the use of AI agents, generative AI, and this growing demand for linking software expenditures to return on investment. Then I’m going to move to my rant or rave segment where I pick one item in the market and I either champion it or criticize it.
So let’s get right into it. Let’s start talking a little bit about the shifting pricing models for SaaS vendors. So if we think about the market over the past several years, typically, large vendors, are going to be offering their software on a software as a service basis based on the number of seats or a number of individual licenses that are assigned to an actual human user. So for example, if you are a user of a large CRM and you have a headcount of XYZ, you’re going to have a certain number of people who actually have that one-to-one license model going on. Whereas Keith as a user, I take up one seat license, it can’t be used by anyone else at the same time as me because that is how they actually control and generate revenue. But hopefully, as the business scales up, grows, more employees, then they have to buy more seat licenses in order to give everybody access. Now, that model actually was working really well for organizations, for the vendors, that is, because if you think about it, you could make the case that as a business is more successful, there’s a greater need to invest more into that software because that’s one of the reasons why you’re doing better is because that software is helping you achieve your business goals.
Now, we’re entering a different era now where if we think about what’s going on with two converging factors, you have artificial intelligence, namely generative AI. We’ve all heard about it, I’ve talked about it a number of times obviously on this channel, but also, automation. And the combination of those two things, which is really giving rise to what they’re calling AI agents. Now, these are not just bots which are almost pre-programmed or very limited in function. AI agents are actually designed to carry out full tasks and actually use logic to understand what the customer’s intent is and then go out and actually carry out the various steps internally in a digital realm to achieve a specific process or carry out a certain work order or respond to a request, what have you. Now, there’s a few things that are going on there. Number one, by doing that, you’re freeing up human workers to focus on higher value tasks, doing things that humans do really well, and what is it that humans do well that machines aren’t quite up to the task yet?
Well, that’s really displaying empathy for a customer situation. If I’m having a problem with the company and I call up and I’m steaming mad about something, well, I appreciate when the human agent at the other end of the computer screen says, “Hey, Mr. Kirkpatrick, I’m really sorry about what you’re going through. It’s really frustrating. I know when you’re trying to do X, Y, and Z and it’s just not working.” Now, whether or not they really mean it, it doesn’t really matter. The point is we are all human and we like that acknowledgement that someone is listening to us and capturing or understanding and empathizing with our frustration. So for the vendors, there’s really a couple of things that are going on here. Vendors have definitely realizing that companies are not going to be buying as many seat licenses as we start to really implement these AI agents because realistically, these agents are able to accomplish tasks where there’s no need for a human to really interact. And what I’m talking about are things like if I want to check my balance with the bank or if I want to initiate a product return, or if I am a B2B business and I want to place an order for something that I always place and just need to change the quantity. Essentially, tasks that don’t require a lot of heavy lifting in terms of decision-making.
Now, a lot of these agents are starting to be equipped more intelligence to handle that. But again, the goal is to take these use cases where ultimately, they’re very transactional in nature or informational in nature, and you’re essentially harnessing the power of AI to quickly identify what is the intent of that request, and then using AI to fulfill that very quickly, and again, saving these human agents for these other more soft, touchy-feely type issues. So what this means again, is if I’m an organization that maybe let’s say I used to have 100 customer service reps who would be using or utilizing the CRM or a contact center platform, well, maybe that head count is going to go down, so maybe I only need 50 or 40 or whatnot. Well, management realized that well, that’s going to impact their bottom line. And what they needed to do is figure out a way to keep that revenue coming in, even amid a drop in the number of seat licenses.
So really, there’s a couple of different approaches, but one of them that’s really gaining a lot of steam right now is this consumption or usage based pricing model. So what is that? Well, that basically means that instead of just charging organizations based on the number of seat licenses, it’s about how much they’re actually using the software. And how do you do that? Well, it’s really about tying it to specific actions, whether it’s something like every time the agent is employed to handle a specific task or there’s a certain amount of API calls to a certain system you can charge based on that. It can also be used to look at something like with generative AI, like, okay, how much… Let’s assign a certain value to a certain amount of compute. So let’s say handling a really basic lookup task where you’re looking at customer information and maybe some product information, maybe that would be charged a value of two tokens. Whereas doing something more complex might be charged five. So again, you’re tying the value of this technology to a specific amount of resource that’s being consumed.
And what’s really great about that is, again, as a customer uses more, they’re going to pay more. And of course, on the flip side, theoretically, they’re getting more value because they’re not having to hire human workers who are, generally speaking, going to be more expensive on a per head basis than an AI agent. You also don’t have to pay them benefits. You don’t have to worry about them getting tired or cranky or deviating from best practices. You don’t have to worry about compliance issues, all of that kind of stuff. We know all of that, but mostly, it’s about when you think about it from the perspective of the vendor, it’s about figuring out a way to deliver value to the customer while it’s continuing to generate revenue in a way that’s very transparent and it’s directly tying the actual expenditure on software to a real outcome or real benefit, actually.
Now, I think that’s really important because we’re in a different mindset now where organizations are really looking at AI and saying, “Well,” or at least some organizations are looking at it and going, “Well, I hear about all this AI talk. I need to know how is it really going to benefit my organization?” And let’s not just talk about it in generalities. I want to actually see real metrics. Are we seeing improvements in productivity? Are we seeing improvements in efficiency? Are we seeing a decrease in risk because of, you’re shifting that risk away from humans, which you are naturally going to introduce more risk in certain scenarios than computers. So in that sense, there are a lot of real benefits there.
Now, of course, there are a lot of concerns too in terms of making sure that these bots and agents do what they’re supposed to do, making sure that they do it efficiently and making sure that they’re deployed in the right scenarios. Humans ultimately, whether we’re talking B2B or B2C, they’re still human. They still want to make sure that yes, if it works as it says it does, then that’s great. We can accomplish a task very quickly. But the second you run into real friction, there has to be a way to then quickly flip back over to a human to address the problem. I think that’s where, if you look at a great example of this, is if you think back to about a year, year and a half ago, Frontier Airlines, which is that very, very low-cost airline carrier, they decided they were going to shift all of their customer service and support to a digital model, meaning essentially, you can’t ever pick up the phone and talk to a human.
Now, in that scenario, they took the calculated risk of saying, “You know what? Our customers do not care about anything other than the lowest cost. And the way to really make sure that you keep that cost low is by eliminating the cost of employing human agents.” These particular customers or customer types, they are not valuing picking up the phone and talking to your human. They would rather save the money and have to deal with the digital agent, even if it doesn’t do everything that a human does, even it doesn’t say, “Oh, I’m really sorry that your flight was delayed and your bags are off in Tulsa and you’re in San Francisco.” So Frontier took a very, very calculated gamble on that, and to date, I believe it’s paid off for them. I don’t think that is going to be the case for many other types of organizations where price is not the overriding factor when it comes to customer loyalty. So it’s going to be interesting to see how this is done, which organizations look to deploy this approach with pricing.
Now, one thing I wanted to talk a little bit about is the next phase of this, of course, is looking at Zendesk. They actually announced an outcome-based pricing model. So here, instead of looking at just consumption, they’re looking at, well, does the bot do what the customer has set it up to do? And basically pricing these AI agents on, okay, let’s say a customer calls in or writes in through SMS or through a web app. And if the AI agent solves that problem without requiring human intervention, and if there isn’t any issue, I think there’s a certain timeframe, maybe it was 24 or 36 hours after the interaction is taking place, well, then Zendesk will get paid based on that outcome being delivered. It’s a really interesting approach because in the end, it really does do a couple things. Number one, it certainly puts the vendor in the position of saying, “Look, we are really confident in our solution because otherwise we wouldn’t do it this way.” If you think about the whole issue of an outcome, you’re really confident. If you think about it in other areas where healthcare has been shifting to that model of trying to look at an outcomes based approach where you’re not just saying, “I’m going to take money based upon all the steps being followed in the right way.” Here, you’re actually saying, “No, I have that confidence that what we’re doing is actually going to deliver business results.” So that’s really interesting.
I do think we’re going to see more of that in the future because of the fact that the whole model with AI and automation, it’s totally shifting the way that organizations are thinking about support, is totally shifting the way these vendors are delivering their software to their customers. Now, along the same lines, one thing I definitely wanted to mention is that on their latest earnings call, Salesforce’s CEO, Marc Benioff mentioned that, he said he thinks that it is likely that they are going to move to some sort of a consumption based pricing model as they roll out more of their own AI agent solutions for service. I would not be shocked in the least to hear an official announcement from the company on that at Dreamforce in the next couple of weeks. I think that others will follow suit.
It is just the model of… Now, again, I wouldn’t be clear, I think it’s going to be one of those hybrid approaches where some companies may want to, or vendors may offer up a portion of their offering to be based around a consumption model and some may, and obviously, there’s still going to have some seats available because they still need to have as humans. But I do think there is going to be a shift to this new pricing because of the fact that we are rapidly moving to an era where companies need to figure out ways to become not only more efficient, but need to deliver better experiences. And in many cases, humans are not the best to deal with it. Now obviously, as I mentioned at the top, there are scenarios where they still are and you may see, and we’re likely to see a scenario where humans are working alongside of agents to deliver a better experience as well. But again, even with that approach, you would have perhaps one traditional seat license augmented by an AI agent that only serves that human customer service agent. Again, you would likely see that AI agent being priced much differently than a regular seat license.
So I want to make clear, we’re still in very, very early days here in terms of pricing. I think there’s going to be a bit of, hey, let’s throw this out to the market, see what kind of response do we get? And of course, you’re going to see pilot programs and customers taking this on in a limited basis, seeing whether or not it works for them, both in terms of the overall expenditure that they have, as well as seeing whether or not it’s actually delivering the end outcomes that they want in terms of more efficiency, better customer experiences, more loyalty, greater possibility for upsell, all of those metrics that most organizations are looking for. So again, I think to wrap on this topic, interesting to see two companies, one very large in Salesforce, hinting that they might take this approach and of course, a company that serves more than mid-market, going all in talking about outcome-based pricing. And I’m really excited to see how that plays out over the next several months with their initial customer set that might be using that model.
So with that, I actually want to move to my rant or rave segment. This is where of course, I take or I pick one item in the market and I either champion it or criticize it. Now, this is interesting. I just read a report this morning where essentially, I think the report was from, I’m not sure exactly who published it initially, but essentially, there was a news report that highlighted a presentation from the Cox Media Group that claims that it has what it’s called, this active listening technology that is designed to let a company hear what customers are saying around their phones, on their microphones, and then using that data to actually give targeted ads. So I think every one of us can certainly think of a time where they perhaps were mentioning a product they don’t normally talk about or mentioning a person or whoever, and then all of a sudden, they get an ad for that particular product or they are talking about a person and next they know, they come up in your social media news feed or feed as some you may know.
Now, the media reports that had talked to some of the other partners of Cox Media Group, including Facebook, Google, and Amazon, I believe all of those companies say they do not use this and this is something that they just don’t do, and they’ve denied this over the years saying that they don’t listen in to people’s conversations. Now, there’s a couple things that work here. One is, we’ve all been told over the years, in order to make sure this doesn’t happen, you should really check whatever app you have on your phone. You need to look at the permissions and see it doesn’t have access to your microphone. If it does, what data is it collecting? And then ideally, if you’re really paranoid about it, you just don’t give permissions for certain things, then of course, you lose functionality. But I think the important thing here is that there’s two things. One is, if in fact Facebook, Google and Amazon are not telling the truth and a subsequent report comes out and can prove that, that’s bad news, and I would definitely put that up as a rant because if there’s one thing that organizations need to realize is that consumer trust is paramount in this day and age.
Yes, there is that almost implicit agreement that if you’re using technology, companies are going to take your data. But this was something that had come out in the past and said, “We absolutely do not use any type of active listening to track what’s going on and then send you ads.” If it turns out that they are not telling the truth and are doing that and it comes out, that’s certainly a bad thing. What I think really is going on here is that of course, target marketing has become so sophisticated that we don’t even realize as consumers of what we’re doing in terms of interaction with things. And it doesn’t even mean that we need to click on something. We can sit there and scroll on something on social media and they can look at the dwell time and realize that, oh, we are looking at certain things, or there are numerous ways that marketers can capture data about us through our devices, through what we do, how we interact with things. Because most campaigns are not sent in a vacuum. If I get an email ad or a solicitation, I’m also probably getting something on Facebook or perhaps I’m getting a call and you’re taking all that data in the background and really assessing, how am I reacting to that? They’re also using demographic and psychographic data to track and see what consumers are doing.
So my sense is that this is probably not a case of these companies actively listening in because I think they realize that the fallout from this would be pretty big. I don’t know what would come in that, other than the fact that it would just further erode any trust that they may have with their consumers. And depending on that particular consumer, they may just say, “Hey, I’m going to sign out and deactivate my Facebook account, or I won’t use Amazon anymore. I’ll use another service to get products that I want.” But I think that the big thing with all of this, and we’ve seen this with other companies and a great example with Adobe where they changed their terms of service or just basically rewrote their terms of service, and it was misinterpreted by many creators out there, and then it took on life of its own. So it’s that type of stuff where I think large organizations who do capture a lot of data, they need to periodically reiterate to their customers, here’s what we collect, here’s what we don’t collect. Here’s how you opt in for things. Here are the things we will never collect. Because ultimately, it’s about maintaining that customer trust that… Because obviously, there’s a lot of money at stake and there’s a lot of competition out there. If they want to retain those customers, they need to be upfront about it and to be clear. And I think that’s where you gain trust, because honestly, for me, I’m on Facebook, I’ve accepted the terms and conditions. I know I get tons of ads, some of which are relevant, some of which I don’t know how they figure it out with the algorithm, but clearly, I must be doing something to be getting the ads. So I accept that.
But on the other hand, what I will not accept, and I think a lot of customers will not accept, is when a company will come out and say at one point we don’t do something, and then it turns out they’re actually doing it. So I would say that right now, it’s a rave, but it could be a delayed rant in the future if it turns out that some things are going on that shouldn’t be.
So with that, I am going to sign off for today. I’d like to thank everyone for joining me here on Enterprising Insights, and of course, I’ll be back again next time with another episode focused in on all of the happenings within the enterprise application market. That’s all the time I have today. So I’d like to thank everyone for joining me here on Enterprising Insights. I’ll be back with another episode focused on the happenings within the enterprise application market next week. So, thanks again for tuning in and be sure to subscribe, rate and review this podcast on your preferred platform. Thanks, and we’ll see you again next time.
Author Information
Keith has over 25 years of experience in research, marketing, and consulting-based fields.
He has authored in-depth reports and market forecast studies covering artificial intelligence, biometrics, data analytics, robotics, high performance computing, and quantum computing, with a specific focus on the use of these technologies within large enterprise organizations and SMBs. He has also established strong working relationships with the international technology vendor community and is a frequent speaker at industry conferences and events.
In his career as a financial and technology journalist he has written for national and trade publications, including BusinessWeek, CNBC.com, Investment Dealers’ Digest, The Red Herring, The Communications of the ACM, and Mobile Computing & Communications, among others.
He is a member of the Association of Independent Information Professionals (AIIP).
Keith holds dual Bachelor of Arts degrees in Magazine Journalism and Sociology from Syracuse University.
