AI for Go-To-Market: The New Revenue Team Playbook
Episode Summary
Revenue teams are being rebuilt from the inside out, and most of the change is happening below the org chart. Jonathan Kvarfordt, VP of GTM Strategy and Marketing at Momentum, walks through what AI is actually doing inside sales prep, agent workflows, and buyer journeys. He is blunt about what has not worked, direct about how AI tools are really being priced despite the outcome-based pitch, and specific about the GEO and PR programme he ran and how he measured it. The line he draws is between scattered experimentation and operational lift.
Key takeaways
Throwing Copilot licences at a team and expecting transformation did not work. It assumes people are good prompters, and the ceiling on time saved by summarising email is too low to move pipeline or revenue
The outcome-based pricing story is not what the market is actually doing. Every AI-native tool Kvarfordt names, Claude, ChatGPT, Cursor, Lovable, ElevenLabs, prices per seat with credits, because enterprises want financial accountability rather than an open chequebook
Enterprise AI is harder than startups make it look. Governance, security, and infrastructure that can carry the data and human load of a 10,000-person company are what an enterprise platform brings and a startup usually cannot
Prepare for three buying paths, not one: AI-only, AI plus human, and human-only. Some buyers will refuse AI entirely and others will prefer it, and nobody knows their own preference until they have tried it
PR is not dead for AI search, it is load-bearing. Kvarfordt ran PR, social across Reddit, LinkedIn, TikTok and X, and volume content simultaneously, because the engines draw on all of them
He measured it with Peek and Searchable, tracking which URLs were being referenced. Blog content moved fast, PR took longer, and he chose PR subjects that were both prompted often and newsworthy
About Jonathan Kvarfordt
Jonathan Kvarfordt is a go-to-market strategist and AI enablement leader, VP of GTM Strategy and Marketing at Momentum, acquired by Salesforce in 2026. He is also the founder of GTM AI Academy and cofounder of the AI Business Network, with more than 15 years across revenue operations, enablement, and GTM acceleration. His work focuses on turning AI into practical revenue outcomes through orchestration, workflow design, and adoption programmes that stick, and he publishes what he is learning on LinkedIn and the GTM AI Podcast.
In this episode
| 00:00 | Welcome and guest introduction |
| 00:55 | From revenue operations to GTM AI |
| 02:50 | Where the early adopters are stuck |
| 03:40 | What enterprise platforms bring that startups cannot |
| 05:19 | Why the system of record is not going away |
| 06:30 | Bidirectional data, and where the artifacts live |
| 07:59 | Notion as the knowledge base beside Salesforce |
| 09:47 | Claude Cowork, explained slowly |
| 11:50 | Skills, agents, and what sits at the centre |
| 12:34 | Individual and team productivity without engineering |
| 13:42 | Why Copilot licences did not transform anything |
| 16:16 | Why there is no settled answer yet |
| 17:13 | How AI tools are actually priced |
| 18:12 | Consumption pricing, and back to the future |
| 19:06 | Three buying paths: AI-only, AI plus human, human-only |
| 20:48 | Agentic commerce and the e-commerce series |
| 21:39 | Shopify, ChatGPT payments, and buying a house |
| 23:21 | Whether brand still matters |
| 26:21 | A friendly disagreement about what brand means |
| 27:26 | Making sure the AI knows enough about you |
| 28:34 | Building for the old world and the new one at once |
| 31:45 | Measuring PR with Peek and Searchable |
| 33:07 | Resources |
| 35:28 | Leadership skill: encouraging the heart |
| 36:57 | The one thing to remember |
| 37:30 | Wrap-up |
In Johnathan’s words
“People over this last year threw Copilot licences at their team and expected transformation, and it didn’t happen.”
— Jonathan Kvarfordt (13:42)
“Enterprises don’t want to go to a place where they open up a chequebook and say, "Go spend whatever you want and have fun."”
— Jonathan Kvarfordt (17:13)
“It’s the things we said we never would do, but yet we’re doing them. We don’t know how we’re gonna interact with AI until we have an opportunity to interact with it.”
— Jonathan Kvarfordt (19:06)
“Right now humans have a hard time trusting anything. So I knew I had to lean on other humans to make other humans feel good about buying technology.”
— Jonathan Kvarfordt (21:39)
“The best question you can ask yourself with AI is, "I wonder if I can do this with AI?" And be willing to try and fail, because you will.”
— Jonathan Kvarfordt (36:57)
Resources
Jonathan Kvarfordt
• Jonathan Kvarfordt on LinkedIn: linkedin.com/in/jmkmba. He calls it his GTM AI journal, and it is where he publishes first
• GTM AI Podcast: gtmaipodcast.com. Weekly, free, deep dives with other practitioners
• GTM AI Academy: gtmaiacademy.com. The training practice he founded
• Momentum: momentum.io. Confirm the current destination, since he says on air it redirects to Salesforce after the acquisition
People he recommends
• Liza Adams: On LinkedIn. Named as one of the clearest voices on AI strategy
• Toni Perry: On LinkedIn. Liza Adams’ partner, also recommended
• Emilia Moeller: On LinkedIn. Based in Europe, recommended specifically for AI search
Tools and technologies discussed
• Peek: peek.ai. One of the two tools he uses to see which URLs the engines are referencing
• Searchable: searchable.com. The other half of his AI search measurement
• Claude Cowork: The tool he walks through at 09:47, and the one he introduced Christina to
• Notion: notion.so. Momentum’s team knowledge base, sitting alongside Salesforce
• Salesforce: salesforce.com. Discussed both as the system of record and as Momentum’s acquirer
• Shopify and ChatGPT payments: The agentic commerce integration he raises at 21:39
Related AI Realized episodes and events
• AI Search Visibility: When AI Says Your Company Is Dead: Curtis Sparrer of Bospar on brand visibility in AI answer engines, and what to do when one gets you wrong.
• Agentic AI and Revenue Work: What Actually Pays Off: Christopher Penn of Trust Insights on where agentic AI has produced revenue, and how to measure it.
• The AI Discoverability Shift: GEO and Search: The executive roundtable on generative engine optimization and search.
• GEO Strategies webinar: A working session on generative engine optimization for enterprise brands.
Frequently Asked Questions
-
It failed for two reasons: it assumed people were good prompters who understood the technology, and it targeted work with too low a ceiling. Jonathan Kvarfordt points out that most people were not good prompters, and that there is only so much time to be saved summarizing and writing emails, none of which lifts pipeline or revenue. His reframing is to ask what has to change internally so that AI produces revenue rather than saved minutes.
-
Per seat, with credits, despite the outcome-based pricing narrative. Kvarfordt points out that every AI-native product he can name prices this way, including Claude, ChatGPT, Cursor, Lovable, and ElevenLabs. His explanation is that enterprises want financial accountability rather than an open chequebook. He does not rule out outcome-based pricing eventually, he says the market is not there yet.
-
Governance, security, and infrastructure built for load. Kvarfordt, speaking from inside Momentum after its acquisition by Salesforce, argues people oversimplify how hard an enterprise motion is: a 10,000-person company brings integration surface, data volume, and human concurrency that most startups have not built for. His example is Zscaler, where tens of thousands of users could not have been served by writing custom code against their Salesforce logic.
-
By supporting three paths rather than betting on one. Kvarfordt expects AI to eventually run an entire sales process with no human involved, but says the real question is what the buyer wants, so companies need an AI-only path, an AI plus human path, and a human-only path. His analogy is Uber: people who were taught never to get into a car with a stranger now do it routinely, and he took his first Waymo last summer. Nobody knows their own preference until they have tried it.
-
Kvarfordt and Christina end up agreeing it matters more, while defining it differently. His practical point is that trust is currently scarce, so at Momentum he leaned on customers and influencers rather than company messaging, because humans needed other humans to feel comfortable buying. Christina’s framing is that brand is rising in relevance but the mechanics of building it are changing.
-
Kvarfordt built for the old world and the new one simultaneously, because neither has gone away. The parts: a running PR program, which he argues brings authority to the brand rather than being obsolete; social presence across Reddit, LinkedIn, TikTok and X, because Perplexity and ChatGPT draw on social data; and volume content beyond social, including YouTube. Core SEO principles still apply underneath all of it.
-
Kvarfordt uses Peek and Searchable together to see which URLs the engines reference. Blog content shows impact noticeably faster than PR, which takes time and varies with placement. His sharpest tactic is subject selection: he chose three or four PR topics that were both frequently prompted and newsworthy, so a single placement worked for the engines and for human readers at once.
-
Kvarfordt calls it encouraging the heart. His reasoning is that when everyone has access to the same tools, the differentiator is the individual wielding them: two people with identical technology and knowledge will produce genuinely different work. He sees the leader’s job as stoking someone’s individual genius and helping them think more critically and creatively, and he is explicit that the people who compound are the ones geeky enough about a subject to keep experimenting.
-
[00:00] Christina Ellwood: Welcome to AI Realized, the podcast for enterprise executives leading AI deployments. From tackling security, data, and operational challenges to navigating organizational transformation, AI deployment offers a unique opportunity to redesign organizations from the inside out. I'm Christina Elwood, your host for today's episode, and we're talking today with Jonathan Carfort, the VP of Go-to-Market and Strategy for Momentum. Welcome to the show, Jonathan.
[00:29] Jonathan Kvarfordt: Christina, it's so good to see you. How are you?
[00:31] Christina Ellwood: I'm doing great, and I'm so excited to have a chance to talk with you. I'm pleased to say that you were the person who first introduced me to Claude Cowork. So you've, you're a- Yes ... quite a prolific poster on LinkedIn and have been doing a lot of work re-related to AI agents. So maybe we could start with how did you get involved with AI, and how did you get involved with starting to help other people adopt it?
[00:55] Jonathan Kvarfordt: I'll try to give the short version of this. My background is in rev ops and enablement, so before AI was a thing in GTM, I was doing enablement of technology and processes and skills and all that stuff before all this craziness started to happen. But about five years ago, the company I was with bought an AI technology from this guy who built his own model in India, and he became our chief AI officer, and my job was to help communicate what the heck AI was to both the internal salespeople and customer success, as well as the market. So I helped train what the heck AI was and how to use it. So I learned a ton from him around AI, and then obviously I did enablement for a career. And then as ChatGPT launched, I had this aha moment where I saw that this is gonna shift everything. So ChatGPT launched in November of 2022. I launched the GTM AI Academy in December of '22, because I realized very quickly that a lot of people were posting things on LinkedIn or other places that I didn't think were very effective. And it's not their fault, because at the time, no one taught us how to talk to AI. We had no idea what we were doing. And not that I'm saying I knew, I just had some, a leg up because of the AI pro that I worked with, and just from my training background, because I'm, I've realized more and more that enabling people and training people have a leg up because they train humans, and a lot of the same training you do for a human is the same thing you do in a AI agent. So anyways, I dove in. I've had 10,000 people plus go through the academy. I've been in some sort of company that's had AI, and the most recent one is Momentum, which is a data orchestration platform, where I do a lot of the AI agent training inside of Momentum itself. And then I work with and advise both go-to-market and AI companies on both go-to-market and AI, and then talk to a lot of CROs and CMOs and other people around strategy of, like, how do you think about AI inside your company and get it operational for either a customer experience point of view or internal operations. So I'm in it, sell it, teach it, market it, all the things AI all day long.
[02:50] Christina Ellwood: You are a perfect person to be addressing our community because they are the early adopters of AI in the enterprise. And as you know, they have been forging the same, uh, trail that you have by adopting before there were lots of rule books and playbooks and- Right benchmarks and frameworks and everything else. You're in very good company with this audience, and of course- Yes ... are at different parts of their journey as well. So maybe we could start by talking a little bit about the fact that we're now in a build versus buy market, uh, where just a few months ago we were, uh, really still in a build market.
[03:27] Jonathan Kvarfordt: Uh-huh.
[03:27] Christina Ellwood: So maybe you would talk to us a little bit about how you are thinking about the adoption of AI in the enterprise as we move into this new era
[03:40] Jonathan Kvarfordt: It's a very good question because as we speak, AI continues to get better, and every week it does more things than I ever thought it could do a year ago. So it's a hard question to answer 'cause of where we're at. However, I will say this because Momentum was recently acquired by Salesforce, and I can tell you from being inside the organization and seeing what they're doing, they're not dumb to what's happening in the marketplace. They know what's going on and they're planning appropriately. There's levels of things that you get with an enterprise-type technology like a Salesforce that you don't necessarily get with an AI startup, such things like governance and security, which is very important to the enterprise. The other side is it's just they're built for an infrastructure that can handle certain weight and load of both data and human usage that not all startups are ready for 'cause they don't have the infrastructure for it. So I think people oversimplify how complicated it can be to be enterprise-type motion, 'cause when you're handling-- When you have an enterprise-type technology, you have a company of 10,000 people or larger, and you have to be able to mix in with different places and understand the different integrations, and all this stuff becomes very complicated very quickly. You can't vibe code a CRM anytime soon. Like, you could, but it'd be, for me as an individual, it would be very difficult to vibe code a team-based enterprise CRM. It'd be very hard to do that. Now, will that day come at some point? I'm sure it will, but I also think that by the time that day comes, Salesforce will be up to speed with where they need to be able to compete with an AI-native CRM. Not all technology's there. I just know from Salesforce they're gonna be, they're gonna be fine.
[05:13] Christina Ellwood: And I always think, too, when it comes to these systems of record like Salesforce-
[05:18] Jonathan Kvarfordt: Yeah.
[05:19] Christina Ellwood: Part of the reason that Salesforce has become a system of record isn't just that it has the feature functions of a CRM. It's that they built a database and a data architecture- Yes ... that is extensible and capable, as you rightly point out, of handling an enterprise load, complexity, security, governance, et cetera. And so it earned its place as a system of record. Now, it needs to go through some changes to be able to keep up with where things are going.
[05:48] Jonathan Kvarfordt: Yes.
[05:48] Christina Ellwood: But that place of the strength of the system of record is not going away. Now, it's a structured database, and a lot, and obviously generative AI is about language and unstructured data. So what is your vision about how those two things are coming together in the world of go-to-market, since that's really your core focus as a practitioner?
[06:15] Jonathan Kvarfordt: Of different types of unstructured and structured data?
[06:18] Christina Ellwood: In the world of AI, how do you see those, the language side with generative AI coming together with the structured side of a system of record like the CRM?
[06:30] Jonathan Kvarfordt: Well, one of the things we did with Momentum is a bidirectional technology, meaning that it's not only feeding data into the CRM and it, it can take actions based off of other data that's going in there like, like usage or ZoomInfo or Clay data or whatever. And there's a lot of things you can do as a result, and there, there's a couple of different things, and I don't think we had the ability before to see these patterns because we didn't have the data to do it with. But now when you have a technology like Momentum that can analyze every single thing that anyone's saying in your company, like the prospect to a customer, you start to find patterns of data that you can then mix with like product usage or third party signals or other things that you never had access to before. That world's gonna be pretty, pretty cool because there's things I'm able to do with Momentum that I never could have done before from a marketing point of view to help me inform of what did my ICP just say this last week in conglomerate from enterprise down to SMB? I can do that analysis now. I could never have done that three years ago. And then I could also say, "Okay. Based on this, what's the aspects of those people in the size of organization, the industries they're in, what's the titles they hold?" All this information I can get really nitty-gritty with and then use AI to create content to go after more people just like them that resonated with our message. Does that make sense? So-
[07:41] Christina Ellwood: How are you putting those, those artifacts how you-- Where are you tracking that information? Because Salesforce isn't set up to track it. There really isn't a system that is designed to track the conclusions that you just identified and the actions that were taken and how that feeds back to the action in the structured CRM data.
[07:59] Jonathan Kvarfordt: Yeah. For us, it's Notion 'cause that was our knowledge base as a team, and we have some integrations to where like- Salesforce becomes like the, I don't wanna say dumping ground 'cause it sounds bad, but like it's the place where everything feeds through. So then I'm able to extract that information, and then I have some automations and AI agents that help me through the analysis part of it. And then I literally go through physically and I read things like old school way, 'cause I wanna know, I wanna know, make sure the AI's not being crazy. And then all my analysis is done in, in an enclosed AI that Momentum has the enterprise deal with, so I know the data's protected. And then I use that to share. I use Notion as a way to dump that information as far as an analysis so that anyone, the founders or anyone else can see, like, why am I doing what I'm doing, is because it's there. My team can see it. It becomes this open source content hub of some sorts. And there's, the world I'm working with now, like Momentum can do a lot of cool things. One of the things we do is we use things like a Make or a Zapier. Now I'm using things with Cloud code to where it's becoming the bridge in between Salesforce and Notion so that I'm no longer the one copy pasting. It does it by itself and can look inside of it. The agent can look, grab information from a Salesforce, feed it into the AI itself, and then the AI can look at the Notion and say, "Okay, this is what happened last week. Here's what's happening this week. This is when we need to shift," and then create content as a result. A lot of things happening without my involvement, but it's definitely, it's pinging me saying, "Hey, we saw this pattern change from last month to this month. Should, you wanna create content like this?" And I sometimes take the re- recommendations of AI, but a lot of times it's me going, "No, I think we need to do this," because of my own experience and instinct and all that kind of stuff. But I love having the AI as another feedback mechanism to help me know what's going on in the space real time, so I don't have to worry about it anymore. I don't have to go listen to 15 calls to figure out what's happening in the last week. We speak faster.
[09:47] Christina Ellwood: So for people who aren't as familiar with Claude Cowork, since it is still relatively new, let's slow that down a little bit. So you're using Cowork, you've connected Cowork to Salesforce, you've connected it to Notion, and you have written an agent that's doing the analysis, and the Cowork agent runs the analysis and compares it to the previous week's data in Notion and then publishes it back to Notion. Is that correct?
[10:17] Jonathan Kvarfordt: It's a very simplified version, but it's essentially, yeah. Now let me get a little more nitty-gritty to make sure, hopefully it makes sense for this. Right now, my Claude Cowork does not have an integration into Salesforce. I have to use Momentum to feed it into another place that Cowork can access. So it's the same thing, I'm just connecting the dots a little bit differently. Inside of Cowork, Claude just released a concept called Skills. Skills is essentially a markdown prompt with a YM- uh, ML, which is a summary of what's inside the skill itself. They did that because it gives the agent the ability to... I'm gonna try to make sense of this. If I prompt Cowork, I have a list of like 40 skills, and each one of those skills at the very top of it has a line saying, "This skill is about branding," or, "This skill is about competitive analysis," or whatever, or products. It then accesses each skill as it needs to based on that one-line summary. So if I have a task and I say, if the AI gets fed some information, it's gonna take that and go, "Okay, we just got new data from Salesforce or Momentum about what someone just said last week." It then goes into its skill set and say, "Okay, I need to pick these five skills," which is like content creation, analysis, and whatever else is going on. I have deep dove, deep dive, deep dove into, into the skills itself, so I know the quality coming out of it is really good, and then the AI agent will call on those skills based on what's needed, and then it pumps out all the content as a result. Does that make sense? So yeah, there's a lot of building- Cowork
[11:45] Christina Ellwood: skills. So when you're saying skills and, and it and so forth, you're referring to Cowork in those cases, right?
[11:50] Jonathan Kvarfordt: Yes. So Cowork like- Is at the center ... in Cowork, I'm interacting with an a, with a, with an LLM chat. That's who I'm interacting with. But it's now become more agentic because that chat has access to all of the content, skills, data, all this stuff. And based on my chatting and my schedule that I put it on, I can say, "Every day at 7 o'clock, go analyze this data, come back to me and give me a report. And based on the report, then go into Notion, update the information." Does that make sense?
[12:17] Christina Ellwood: Yes, it does. So- And I appreciate you, you breaking that down because I think for a lot of people that are listening, this is a relatively new set of capabilities for them, and it's illustrative of this era that we're in of going from build to buy and bridging the old and the new and so forth.
[12:34] Jonathan Kvarfordt: Yes.
[12:34] Christina Ellwood: Because what you've just described is for an individual or team productivity, you're able to build these agents relatively easily using the more, the Cowork type of model or the OpenClau or whatever the equivalent is of your LLM of choice. You're able to build these to automate a number of different tasks 'cause at the individual or team level. That is not the way you would do it for an enterprise-wide deployment of the technology, nor is it a solution that you're buying off the shelf. So it's a little bit in between, and it's down at more the individual level. As you zoom up and you think about tools that you can buy at the very top, like a Salesforce that may be enabled with AI or other tools that are AI native but are designed to be complete systems of their own, the equivalent of a SaaS product that is built with AI and is a natively AI, what are you seeing there that is changing how you're thinking about the use of AI in the enterprise as a whole?
[13:42] Jonathan Kvarfordt: Again, it's a delicate balance because security and governance is such a huge part of this that you can't skip over that. There's two sides of this. The one side is what's happening internally, and I think people over this last year threw Copilot licenses at their team and expected transformation, and it didn't happen. Because number one, it requires them to be good prompters and understand the technology, which they don't, and number two, you can only get so much time savings and a- amplification of a human with so many emails being saved. Like, I don't think you can save a lot of time with summarizing and writing different emails. It's not gonna lift up your pipeline or revenue. So the question is, how do you analyze or how do you look at what's happening in your company internally so it brings more revenue and pipeline? Like, how do you shift that gear? Which at some points, which is why Momentum was so successful, we were that technology. Like, people couldn't vibe code at Momentum. People in enterprise like Zscaler, who had tens of thousands of users, had to use a Momentum to m- 'cause otherwise, to vibe code for tens of thousands of users based on their Salesforce logic would've taken them forever to do and make sure it was secure and safe. So it is this delicate balance of if peop- someone's listening, I would say not to go against technology like Salesforce. It's more looking at what's the future you're going to and who's building for that future, which again, Salesforce is. They're slower 'cause they're bigger obviously, but they are going that way. At the same time, using new technology like Momentum and figuring out what is the real friction points we're looking to do? Can we build for it? And then you absolutely should. And if you can't, don't tr- don't think you're gonna figure it out, 'cause you won't. Like, I can't tell you how many people who came to Momentum and said, "We're gonna go try and do this ourselves," and then came back three months later and they said, "We can't do it." I'm like, "I know. This is hard." Like, it's, it's not as easy as you think. And the other part of it is do you want to have your internal resources, the people who are internally are building you products to get more revenue, do you want them focused on internal operations or do you want to focus on more revenue externally with products? Like, personally speaking, you need to be externally focused with your internal resources because technology's moving so fast, you need as much as, as you can to make sure you're keeping up with the game. 'Cause otherwise you're gonna be eradicated. If Salesforce didn't have a fleet of engineers making sure their team was up and running, no matter how well they got their internal operations going, they'd be eradicated. So you have to have this balance of how are we facing the marketplace and making sure we're keeping up, and how are we making sure our internal teams are working? It's always a continual back and forth of figuring out, do we hire for this, do we buy for this, or do we build for this? Which again, depends on-
[16:13] Christina Ellwood: You don't have... There's not an obvious answer at the moment
[16:16] Jonathan Kvarfordt: No, and I don't think there's gonna be for a while. And the only reason why is because Cowork just came out two months ago. It's dramatically changing everything. That is gonna have an impact on this whole conversation in the next six months. So theoretically speaking, someone could get, like right now in Cowork, you can get an enterprise license and you can have skills or agents that are for the entire organization, kinda like a custom GPT was. That can have major influence and impact on what technology people buy or build, but it's not gonna be for every single use case and not for every single problem. So th- that's why I think people need to understand what is the larger outcomes you're going for? Where are the friction points? What can you build for and what can't you build for? And then go buy what you can't and then run and make sure it's aligned to hitting revenue or pipeline targets.
[17:02] Christina Ellwood: So a lot of these tools are being sold now priced based on output or consumption.
[17:10] Jonathan Kvarfordt: That's actually inaccurate.
[17:12] Christina Ellwood: Go ahead. Say more.
[17:13] Jonathan Kvarfordt: Yeah, there's-- When you look at all the tools coming out, like Cloud Cowork and others, they're not based on that. They're based on seats and credits per seat. So you have a certain amount of licenses or credits or tokens per price per seat. Enterprises don't want to go to a place where they open up a checkbook and say, "Go spend whatever you want and have fun." They want to have a secure accountability of what's happening financially. Like, you look at any of the AI native, Eleven Labs, Cursor, Lovable, Claude, ChatGPT, all of them have a fixed price per seat, all of them, and they're all an AI native. So the whole outcome base-- I'm not saying outcome could not happen. I think it can. We're just not there yet.
[17:54] Christina Ellwood: I was actually thinking of the vertical offerings that are out there when I said that they are all trying to go to output base. But you're bringing up exactly the concern that I think everyone has and why I think it's unlikely to be anytime soon that we can go to that kind of a model. But- Even if
[18:11] Jonathan Kvarfordt: you took it from a- Salesforce was one of
[18:12] Christina Ellwood: the first to claim they were gonna go to an all consumption-based model, so it- Yeah ... really what provoked the question is, are we back to the future here or not?
[18:21] Jonathan Kvarfordt: I don't know. I know obviously Salesforce has had success. They've had some failures. They just had the last earnings call last week where they-- Mark was very honest or open about it, and I don't know if they're gonna switch the models or not. But when you think about, for example, you're a human, if we're gonna use the human payment model as a reference, I, I don't get paid by my outcomes. I get paid a bonus on a certain percentage, but at some point my pay is capped. If I don't want to have an uncapped way, I go entrepreneur, so I don't have a cap on my pay. So I, I just don't see a world where it's just gonna be an open checkbook. I just don't see that happening based on saying, "You have these bajillion outcomes, and this is all things." I just don't think that's gonna be a good financial outcome.
[19:00] Christina Ellwood: How do you think people are gonna buy their, their AI tools in the long term?
[19:06] Jonathan Kvarfordt: It's a good question. I think as AI gets smarter, it's gonna have the ability to take someone through the entire sales process without any humans involved ever. And so the question will be: What will the consumer want to go through? What's their choice? And I think companies need to be prepared for having either an AI-only play or an AI plus human play or a human-only play, because some people, for whatever reason, will be like, "I don't want to have anything to do with AI. Just let me talk to a human." And other people will be like, geek out on having an AI experience and go through the motions. It's-- And I think some point we'll be weirded out when we have things that we couldn't do. It's like Uber. When Uber first came out, when we grew up, when I grew up, they told me never to get in the car with strangers, and you gotta get in the car with strangers all the time because of Uber. And then I thought, "Would it be crazy if I got in a car with no driver?" And I just took a Waymo for the first time last summer. So it's the things we said we never would do, but yet we're doing them, so it's like we don't know- How we're gonna interact with AI until we have an opportunity to interact with it and see, do I like this? 'Cause like personally, I took a Waymo. It was cool. I didn't like it. Like, Uber's way faster because the human knows how to interact and do things, and the robots right now are very much to the letter of the law and follow the rules and guides, and Uber doesn't do that . So it's like it gets you faster. And it's my-- And I do that versus renting a car. I have all these options of transportation, yet I pick an Uber. I think the same thing's gonna happen with AI. We have all these options of how I want to buy products, services, houses, insurance, whatever, and we'll be able to pick which way you wanna go down. And I do think that when you do pick a human-led sales process, it will become with a specialist who is like really freaking good at the technology and like expert in the field, and will come with a premium. It's not gonna be cheap to do that
[20:48] Christina Ellwood: Interesting view. We're doing an e-commerce series on this podcast which will be hosted by Anna Luo of DaVinci Commerce. Yeah. And they will be talking about this very issue of the, for the consumer side of the world, the c- what she calls the collapse of the buyer journey. Yeah. That it'll be in the LLM, they'll be asking about a product, "I wanna buy an under desk treadmill." It'll give them options, they'll click a button, they'll buy. So that opens a new opportunity for brands to have a direct relationship with consumers that they've never had before. It also changes the level of exposure that a consumer has with any given brand. They no longer expo- they no longer engage brand first, buy second. They're buying first and engaging brand second. What's your thought about how that's gonna change the way the world works?
[21:39] Jonathan Kvarfordt: Yeah. This is a very good question. Shopify just announced that they're doing the official integration of the ChatGPT for payment processing and the payment infrastructure, which is fascinating. And I think about this a lot with like sales-led companies where, I mean, is the world exists where it's all product or AI-led to where there's no salespeople involved? Like, I find that world fascinating. You think about something as complex as buying a house. There's an existence where you could buy a house with no human involved. So I think that the new world of, like you just said, buying first, brand second, there will be an aspect of it, but I also think it's on a spectrum. I don't think every single product that everyone's gonna make is gonna be product first. It depends on if you're buying shoes versus if I'm buying a million dollar technology, it might be a little bit different. I can't tell you now how many people, like with the marketing with Momentum, I leaned on customers and influencers and other people to help talk about Momentum, because right now humans have a hard time trusting anything. So I knew I had to lean on other humans to make other humans feel good about buying technology. And in the future, will someone buy a million dollars with technology with AI? I'm sure they will. But I just don't know... Yeah, I just don't know if it'll be because they just chatted on Perplexity and all of a sudden said, "Oh, cool, I'm gonna buy a million dollars of technology." Bam, and there's nothing else going on. I just don't think that's gonna happen anytime soon.
[22:55] Christina Ellwood: But there's a, like you say, that's on a continuum. There are very few consumer purchases that are a million dollars. You named one.
[23:01] Jonathan Kvarfordt: Yeah.
[23:01] Christina Ellwood: So for many of the consumer purchases, it seems that will happen more quickly, of course, than it will happen with B2B sales, but I think B2C always leads with the technology transformations- Really? on tr- transactions. It doesn't lead on, uh, in other areas, but it leads on transactions, and a purchase is just a transaction.
[23:21] Jonathan Kvarfordt: Yeah, but to be fair though, from a marketing branding point of view, I still think that the, I'll use Apple even though they're pre-AI- We talk about this a lot in marketing, that products have to do with your identity, you know, and a lot of people buy products because of identity. And I've been thinking about... I don't have an answer for this, but I've been thinking a lot to your point of will that matter as much when someone can just buy something? Will the phone or-- 'cause I'm an avid Mac guy. I love Macs, and just because I search for another technology and it brought up, I might have it considered, but I'm not gonna move away from my Mac. So it's like I, I'm loyal to technology and the brand that I have, so it's just interesting to see, like, how will that identity branding point of view of any product influence all of us? Well, and there's-- I always say too, distribution is the new moat. So if you have a distribution mechanism that's through an influencer and they push some random product, like that's another way for someone to go, "Oh, I saw this product on Bob that he talked about on YouTube, I'm gonna go buy it." I still think there is gonna be a brand mechanism or identity mechanism. It's just gonna be a different way to access it beyond just a retail store.
[24:25] Christina Ellwood: So I actually think there's another issue that you sidestepped there. So you had the chance to build your brand loyalty to Apple before anybody brought you an alternative to Apple. Yep. So you have to think about places where you don't have a brand, uh, affiliation already, or think about being a person who grew up in an era that where the only way to buy was by clicking on a link in a chatbot, okay? Their relationship to the brand has no foundation, and so that foundation has to be built in some other way. And you mentioned influencers in your own work with Momentum, that you leaned on influencers because they effectively were building the credibility for the brand and, and passing their affiliation with the brand on to the TikTok viewer. So they created a relationship with the brand. Perhaps it was commercial, perhaps it was genuine, perhaps it was both. But they were then transferring that to their viewer, and their viewer was saying, "Yes, I also want to have that same thing because I love Jonathan and he's, he says this is great, so I'm gonna buy it. If Jonathan thinks it's great, then I think it's great, then I may tell someone it's great." So that relationship is person to person, not brand to person. The brand is in there, but it's not brand first. You make that even harder when you take away the need for Jonathan at all. I don't even need TikTok to find out about it. I can find out about it passively through this chat interface. So now how do we build the brand? So I think that's a very important question for us to be asking ourselves. If we're not a consumer brand, you have the same problem, but in a different, in a some-somewhat different way on the B2B side. Because on the B2B side, you not only have to have someone be aware of what it is that you sell and who the candidates are to buy from, but you have to have some way for them to trust you on behalf of their company, not just on the behalf of their wallet.
[26:21] Jonathan Kvarfordt: Yeah. I'm gonna push back a little bit because I think you and I may define brand a little bit differently, just because the collection of people who push momentum is not the same as others who don't resonate with it. It is part of the brand. Like, they do become an extension of it. Now, is it directly funded around like you talk about? No. But there's a lot of things in life now pre-AI that have that same situation. It's like it's gonna be the same challenge as before. I just think that brands have to recreate themselves in how they're doing it, whether that's leveraging reputations of others who they trust. Like, again, I have customers who I lean on, they lean on us to make sure the brand was there, and all of them have a pattern. Like, I have a collection of 20 people who all have the same type of persona and feel that they become an extension and the length of the brand. So I think that to your point, will it change? Yes. Is it gonna be-- Is brand gonna be irrelevant? I don't think so.
[27:09] Christina Ellwood: Oh, I think brand is gonna be more relevant.
[27:11] Jonathan Kvarfordt: Yeah. I do too.
[27:12] Christina Ellwood: I think it actually is going up in relevance. I just think the way in which we build it is changing. The mechanics and the opportunities and the exposures that individuals and companies have to our brand is changing, and we have to change with it.
[27:26] Jonathan Kvarfordt: And marketers, to, to your point, like a part of that is making sure that the AI knows enough about you and your chat or however, like, the agents could be bringing you up via brand so it can represent you in a chat with some random user in India in the right way. So I think that's the shift that, for me anyways, that I'm making is, like, how do I make sure that whether it's a human prompting getting answers, or if it's an AI agent with OpenClaw going through and doing research, how do I make sure my brand is represented in a way that's strong? Yeah. It's fascinating.
[27:53] Christina Ellwood: Yeah. For how-- What were your strategies for momentum? I m- I mean, aside from the basic LMS text and the answer engine and the FAQs. Um, in fact, we're doing a, a executive roundtable on this and a webinar as well, and have had a couple articles written on the topic. But I'm interested for momentum, what you specifically did beyond the most basic things for being found on the LLMs and being covered correctly. How, what did you-- what were your strategies?
[28:18] Jonathan Kvarfordt: For-- Are, are you talking about strategies for AI search in general or just overall marketing?
[28:22] Christina Ellwood: Well, you can take, take the question wherever you want it to go. I was asking specifically about what you're doing for, what you did for GEO, AEO, and SEO, but you can take it- Yeah to the broader question, uh, uh, if you'd like.
[28:34] Jonathan Kvarfordt: There's a multi-level answer because a lot of the core principles of SEO still apply. They just-- There's, we're in this weird world where Like everything, it's shifting. There's still the old world that you have to be worried about and concerned about, and then it's shifting over to the new world. So it's, for me, in this last year of twenty twenty-five, I had to build for both because we're not pure AI yet. So how LLMs view different things, for example, some people think that PR in a general category is worthless now, and I totally disagree with that because it brings authority to the brand and to the product. So one of my staple things was I had a good PR program going on, getting out there with like interviews and product launches and PR press releases, blah, blah, blah, and that covered one piece of the pie. On top of that, you also have your social, because Perplexity and ChatGPT and others use social media as another way to get their data. So I made sure I had people talk about us on Reddit, on LinkedIn, on TikTok, on Twitter, on all these different places. So that was another piece of the pie. And then you have things like a lot of content, not just through social media, but just in general, either through YouTube or just u- content to what I always shared online, to making sure I was being reviewed by all the different advisors and blogs and all that other stuff. And then for our own, for our own website, I was doing things like making sure the text files and the robot files and the AI files were all in order that the human doesn't see but the AI absolutely sees on the website itself. And we pumped out in a month's time, like we used a tech to kind of see where we're ranking, and we found that we were not ranking in some key prompts in June. We wrote a hundred blogs with AI. We optimized it, like I optimized a lot of them myself personally, and then we designed it, get on the blog, and within two weeks we were ranking past higher than Salesforce in some categories that they are key in because at the time we weren't being acquired. So I was like, "We're onto something. This is awesome." And the cool thing is that's shifted because in July I had to be very much involved with the writing process. This last month I did the same process with an AI agent, and I'm barely touching the content and it's ranking four times as high as my last process in July, six months ago. So it's like it's continually shifting and changing, and the aspects are crazy. However, my point with all this is that AI is recognizing, I call it recognizing game. If you have junk, it's not gonna recognize you. If you have game, it recognizes you. So you need either good in-depth content on your website that's like valuable to your ICP and/or something insightful that's just not the AI slop that someone could do a one-sentence prompt to get some results for. It has to be something unique. And for me in my process, a lot of the AI agents are trained in how I think around the particular problem with Momentum and my founders and other people, so it's not just me, it's the four or five core people in the company that it's trained on so that every time a new topic comes up like OpenCloude or Cloud Cowork, we have this framework of thinking from the founders and myself we can then apply to it and then whip it out so we have the ability to get the traction and the wave of the new tech and still be applied to how we think about it. Does that make sense? Because it shows uniqueness and authority.
[31:36] Christina Ellwood: So- No, absolutely. And I wanna go back to your point about PR. How were you measuring the impact of PR on your AEO, SEO, and GEO results?
[31:45] Jonathan Kvarfordt: I use a couple of technologies that shows me where things are coming from so I could track it on the back end. So I use a combination of Peek, p-e-e-k.ai, and then searchable.com, and they give you an understanding of like where, what URLs are being referenced and where, so I can understand like how is the PR working. It takes a little bit of time. Like my-- I can see it impacts in my blog way faster. I don't know why that is, but it just is. But with the PR, it takes a little time to get some traction from PR places, depending on if it's Wall Street Journal or what the placement is of the content. But I started to see a lot of the PR articles we had starting to help rank us in different terms. So it's kind of like a wheel. Like the crazy thing is like with PR, I told them like, "I wanna make sure we focus on these three or four subjects." 'Cause I knew that the prompting land was p- people were generally prompting about this particular subject. But I also knew it was a hot topic from a PR point of view, so I was hitting two birds with one stone. Does that make sense? Yes. So I knew it would help optimize for AI, but also help with the humans of whatever topic it was.
[32:43] Christina Ellwood: Yeah, that's really good advice. And of course, this is exactly the kind of thing that we're gonna be talking about at the roundtable and that the webinar will be covering. Yeah. Because this is an area for everyone who's in the go-to-market motion, how-- not only how do you get found and ranked correctly for your content and what have you- Yeah but also how do you, how do you create an engine around it, right? What are the systems and processes that are necessary? So I appreciate you sharing that with us.
[33:07] Jonathan Kvarfordt: Yep.
[33:07] Christina Ellwood: I'm gonna move to, to our closing here and talk a bit what resources you do recommend to listeners. And I want you to cover three kinds of resources. Resources that you recommend for learning more about you and your work, including your AI training work and momentum, and then what resources do you recommend for people to understand more deeply how to s- develop their strategy?
[33:39] Jonathan Kvarfordt: Okay. For me, my LinkedIn is my GTM AI journal, so everything I do is on LinkedIn. And then we also have the gtmai podcast.com is where we release podcasts every week, and we deep dive with others on all these topics. So both are free obviously, and hopefully they give a lot of value because I try my best to give value, not fluff. Momentum is obviously transitioning to Salesforce, so best place to go for right now is momentum.io, and then after two months it's gonna be salesforce.com. But either way, we'll probably redirect to the same place at some point. And then as far as how to think about strategy, it's a really good question because it changes based on a lot of different factors. However, I would say... Yeah, it's a good question. I'm not trying to be cocky, but I think I'm one of the voices in strategy that talks a lot about this, so I'm definitely one of them. Another one is na- a lady on LinkedIn named Liza Adams, L-I-Z-A A-D-A-M-S. She's awesome. Another, her partner is called Toni Perry. She's also amazing. In the AI search, there's a woman named Emilia Moeller, E-M-I-L-I-A M-O-E-L-L-E-R. She's in Europe and she's awesome with AI search stuff. So there's a lot of resources, but again, most of the resources that I'm quoting are people. Like they're, the strategy people I follow are these people who I look to for advice and results and seeing what they're doing and experimenting, and there's just so much good stuff out there. It's just hard to just decide on just one. So yeah.
[35:06] Christina Ellwood: Appreciate that. I do. It is very hard, and it's one of the reasons I think it's so valuable to have people like you share your top of mind resources, because those are the ones that are your go-to.
[35:15] Jonathan Kvarfordt: Yeah.
[35:16] Christina Ellwood: So in the AI era, what leadership skill do you find matters the most for you right now?
[35:28] Jonathan Kvarfordt: I think encouraging the heart is more important now more than ever, honestly. And I mean that on several different levels. I believe that when everyone has access to the same AI tools, that what's going to change how those tools are being used is the individual using them and wielding them. So if I made an AI agent team, if you and I, Christina, had the same technology, let's just say we had the same knowledge around AI, and we each went to the-- had the task of doing a marketing plan or whatever, you and I would come up with very different things to do, which is a good thing, because you have your genius, I have mine, and we have different approaches. My task as a leader is to help stoke the fire of someone's individual genius and their IP and help them think more critically and creatively, because that's where I think the one part of the mo- of the future are the people with the geekiness inside of a certain subject. You know what I mean? It's those who just, like, love talking about whatever topic it is, and they just can't get enough of it, and they just experiment and play. I wanna be a leader who helps stoke that fire within people, and I want them to be with me and be passionate, because there's nothing better than being with a team who is humble but working hard and also geeks on whatever topic they love to geek out on. That's, like, the best ever. So if I can create that environment where someone can do that and be successful and win and be AI native, I'm all for it.
[36:48] Christina Ellwood: Love it. So if our listeners were to remember one thing from today's conversation, what should it be, and why?
[36:57] Jonathan Kvarfordt: I would leave you with push yourself with AI, and the best question you can ask yourself with AI, with your team or individually is, "I wonder if I can do this with AI?" And be willing to try and fail, 'cause you will. But those who are willing to try things and say, "I wonder if I can do this with AI," are the ones who create these pioneering paths and do some really cool things, because we are in a pioneering world, and this requires people to be willing to be pioneers and to try and fail and try again. So that's what I would say.
[37:30] Christina Ellwood: Great advice. So thank you, Jonathan Karfort, for the VP of Go-to-Market and Strategy of ForMomentum. Thank you for joining- You're
[37:37] Jonathan Kvarfordt: welcome ...
[37:37] Christina Ellwood: me today on AI Realized.
[37:40] Jonathan Kvarfordt: Tha- thank you so much, Christina, for having me, and thank you for the listeners. Like, I appreciate everyone listening and appreciate you having me.