PartnerUp #116 - The Future of AI, Agents, and Agencies with Barrett King

What is up PartnerUp!

Barrett King, Sr. Manager - Global GTM Strategy, Partner Ecosystem at Hubspot, discusses how he thinks AI technology will affect agencies. Barrett, Jared, and Isaac talk about the current labor constraints agencies face.

It’s hard to have enough labor when projects come in without too much overhead to survive.

The crew talks about new business models that may arise because of AI. They discuss how this may affect individuals, and the value agencies are able to provide.

Never miss an episode of the world’s number 1 podcast on partnerships by subscribing to PartnerUp on Apple Podcasts, Spotify, or wherever you get your podcasts. If you’re a visual person, sub to our YouTube and see the full recording of us learning out loud.

Share the episode with your commentary on LinkedIn or Twitter, and we’ll highlight your commentary. We love to hear your thoughts on each episode and always comment back or respond to emails/dms.

Hey! We’re real people!

Subscribe & Listen On:

  • Apple | YouTube | Spotify | Amazon | Google | RSS
  • Or literally, anywhere you get your podcasts. Seriously. Ask Alexa: “Alexa, play "PartnerUp the Partnerships Podcast” and magic…

Episode highlights:

  • Intro 0:05
    • Isaac takes on the mantle of the backward hat.
    • The disruptive nature of AI over the next several years.
    • How Hubspot innovates with agency partners.
  • AI's impact on the agency space. 4:18
    • The foundational statement of go-to-market language.
    • AI as an additive component.
    • The first step is to deliver value back to agencies.
    • AI throws a wrench into the agency marketing
  • How to drive a double on your margin? 7:29
    • How to drive a double on margin.
    • How Jasper is helping agencies improve margins.
    • Using AI to get 75% of the work done.
    • The agency's problem.
  • Opportunity cost is different for everyone. 12:21
    • Opportunity cost is different in every economy.
    • What is going to make an agency more successful?
  • How to add more services to AI. 14:47
    • AI as a service or prompt writing.
    • Reduced gross margin and more customer value.
  • AI’s impact on marketing and advertising. 15:21
    • No immediate examples of agencies offering AI services.
    • Targeted ads for AI.
    • The challenge of being an agency right now.
    • The shift from project-based to recurring billing.
  • Does ai as a service become the next frontier? 18:53
    • AI as a service is the next frontier.
    • The first interesting thought about agency space.
  • The problem with monthly recurring payments. 21:03
    • Monthly recurring payments vs project-by-project payments.
    • An agency of one, networked and individual contractor.
  • The power of decentralization of work. 22:30
    • How to master being a specialist again.
    • Decentralize and decentralize.
  • From project to retainer to service agreement at scale. 23:34
    • The first two years of the agency.
    • The baseline of service agreements at scale.
    • Reallocating resources from project to retainer to consumption.
    • Twilio is a great example of consumption PLG.
  • Why didn’t you take the risks of AI? 27:02
    • Why AI is not going to replace a person today.
    • How AI can replace functional tasks.
  • Agencies are like the mortar for software. 29:00
    • Agencies are like the mortar of the company.
    • AI tools might shrink the ICP.
  • Specialization in the agency space. 30:23
    • Ai specialization in the agency space.
    • Hiring experts in different industries.
  • The differentiation between brands and agencies. 31:27
    • How to differentiate yourself in a crowded market.
    • Why agencies are great to partner with.
  • How to build a more resilient business. 37:40
    • Uniqueness and resiliency of the data set.
    • The future of marketing and sales.
  • Getting comfortable with machine learning in a closed environment. 39:50
    • Getting comfortable with the idea of machine learning.
    • AI as a solution to articulate value.
    • How AI can influence human behavior.
    • The danger of AI taking over the world.

Full transcript:

Jared Fuller 0:05
All right, what is up, partner up? We're back. Isaac's in the chair. I'm liking that we're back on a schedule. Isaac consistently rolling things out every Tuesday, amazing guests amazing content. Partner UPS back people. How are you feeling that?

Isaac Morehouse 0:19
Man I, I am taking on the mantle of the backwards hat. Now that you are all cleaned up and adult like so. I'm gonna go ahead and just degenerate a little bit while you continue to move in the other direction. By the way, I just noticed I have on the ecosystem week t shirt. I believe that was exactly one year ago today. We're recording this on June 13. Wasn't ecosystem week June 13 When we kicked it off.

Jared Fuller 0:44
I think so I think that was that was the timeline.

Isaac Morehouse 0:47
Yeah, shout out to partner stag. That was an epic event. That was that was the moment when I was like, Oh, crap, we have to do events, don't we?

Jared Fuller 0:54
Right. Now that was totally the moment of actually we're syncing up with Tyler here soon. So stay tuned, maybe maybe something around that week is coming back? Maybe not. But we are talking. So keep, keep listening out. But I'm excited today to talk about something we haven't really dove into. There's a great a bunch of great content creators out there right now that talk about the intersect of agencies and partnerships. And while this topic is like I certainly relate to as much as any in partnerships being why I own my own agency and I could have been a HubSpot partner in 2011 2000 12,013 and I wasn't so I was an idiot. I learned most of what I know about partnerships from pika puja and HubSpot partner program everyone knows here that I think that they really are the gold standard for how you build a company around this. So I brought longtime connection and friend onto the show and I've been on his podcast Barrett king over here to partner up to talk about what the heck's happening in the world of agencies Barrett, welcome the heck to partner up my man.

Barrett King 1:53
I'm pumped. What's up guys and the backwards hat Isaac absolutely working for you Jared cloth the hair the comb over dudes crushing it. You guys are nailing it. The look is great.

Jared Fuller 2:02
Oh my gosh. I do not like that. This is becoming a topic of recent conversations. We were at an event last week there were people in the literally in chat people were like, Isaac, you need to tell Jared the new looks okay, but he's wore that shirt before and I'm like, oh, that's like no audience. You're not allowed to comment on this. You're this No, no, no. Let's not bring that into question. I am style illiterate. I have to have my my wife dress me. So Barrett, speaking of style. I think in terms of trends right now. This all kind of like kicked off for me and made me realize how serious this moment was, but also give me a lot of like, calmed me down as I listened to kip Bodnar from HubSpot talk at the Pavilion cmo Summit. And Kip didn't mince any words about the disruptive nature of AI over the next several years, right, like right, the second right now, if you're using AI for generative content, I think you're an idiot. In my opinion, AI is not a generative Content tool. But it does a lot of things. And of course, we'll get better over the coming months. I mean, I think we can say months at this point I here's nuance for other use cases that facilitate and disrupt a lot of go to market and I have to imagine that the team at HubSpot kicked me gave me a lot of like, okay, HubSpot is gonna figure this out. But what do you do with the agency partners? If, you know the big innovation that Pete told me about in the beginning was like, we help them sell a better product. And I don't mean HubSpot. I mean, what I used to sell one time, big package, you know, a website redesign 50 grand, right? And then three months later, my p&l goes to zero. Cash flow is nothing right. And I'm like, I gotta sell more and then deliver more. So there's this very lumpy, terrible business. HubSpot innovated the true innovation to me was helping that business move to a recurring retainer based model. Obviously, creating content for the CMS mutual benefit. Like everything in partnership land connects entire industry that's doing things the wrong way. HubSpot launches a CMS builds a partner program around it. They write content and help you use the product better month over month. But now, everyone's kind of like, wait a second, what the heck is going on? What are we going to do with this shit? What the heck can you see? And man?

Barrett King 4:18
I think it's interesting, because it's not it's what you described, like we were robbed spotted really well, was the foundational statement of we help you move from project to retainer, like that was the fundamental go to market language and just straight deliverable that we gave to our partners. Earlier partners were marketing agencies, SMB marketing agencies, folks that were, you know, 510 15 employees, 20 3040, maybe employees, but they were smaller in their growth cycle and they were trying to build to go to market beyond the four corners that they had established before that. I think fundamentally, what worked was that we were able to give them a solution that allowed them to deliver tangible ROI more effectively improve their team's efficiency, all this stuff that you would think directly correlates to what AI I mean, to your point, in an aspirational way should do. Like I think AI is interesting. And if AI if you're listening this Sunday, like, we're cool, we're homies we're friends don't like destroy my house or whatever is gonna happen. I've always jokes like that Skynet in the future. I think AI. And I agree with you, to some extent, becomes an additive component to augment the work that truly good exceptional employees are doing today. And maybe somewhere down the road that changes. Who knows, I think from the agency perspective, it's a close margin, human capital centric business model that delivers exceptional value to the end customer, but is wholly reliant on people and ability to get to that outcome, how fast we do the work, excuse me, how effective Can we do it? What is the value of that in terms of market? Does it get commoditized? Or not? All of those questions start to become a little blurry when you bring in tools that can be additive to that that go to market. So when I think about the impact on agency in particular, and I think about it in context of like certainly companies like HubSpot, other organizations that have an ecosystem around the agency's smaller si space, how much you bring tools like that into your own stack. Step one, like how far can you bring something like AI as a enablement component of your solution, to then deliver value back to those agencies. That's sort of like the first step of it. And the second piece, obviously, is determining how to enable them to build services around it. Again, to your point, Jerry, like I think we and companies like HubSpot have been very intentional around sending, saying it's not just about buying our software, or reselling our software, it's about building a collective Better Together story in which you improve your margins deliver better value to the customer on all the things I just described. Ai throws a wrench to that.

Jared Fuller 6:41
Right. So the couple of things that I want to unpack from this, because you can already tell folks, if you don't follow Barrett on LinkedIn, you definitely should pay attention to his show. The way that you're approaching this, I think he's very intelligent. And I'm, I'm like, out of the loop a little bit in terms of since the, let's say, advent of GPT, and the mass proliferation of this concept, but when I've brought technologies are solutions to agencies, and been in a position, let's just use, I can talk about this, like drift in live chat, right. So like, I was working with a bunch of agencies, and the agencies where it worked, is attended to have to go a little bit more on market tied into an ABM strategy, you know, really get more deeply integrated with marketing automation gets sticky, larger sales team, he got really complicated. And that's how I built service hours. But something that I thought would work really well was something like this. Tell me if AI, if this analogy would work for AI, or wouldn't work for AI? Is that there are agencies that would want run? What? Live Chat centers, right, and they would build on what, like a per agent, kind of per hour type model. And I'm like, Look, the business that you sell right now, is what I mean, it's 30 40%, gross margin, maybe 50%, gross margin if you're kicking butt, right. So it's a low margin business. And what I brought into these organizations was like, look at how we can do these exact same things. Keep your buildings the same, but drive a double on your margin. Right. So for example, instead of selling 10 agents into this account, you could do with our drift AI, what you need with five agents or four agents, but still have a much bigger margin. And no matter what I did, didn't matter who I talked to all but one agency really leaned into that, and just didn't make that didn't come it didn't work. Like when I was like, hey, I can help you improve your margins. And it was objective, like we had a lot of proof drift had a great brand. And we were successful with agencies just not in that way. Is that because that was good times. And now it's bad times and agencies are like, shoot, we need to change, or like it would that same pitch be a relevant thing, hey, utilize AI to drive better margin in terms of production of content, but then also adapt it for social, for instance, or, Hey, you better go down the stack and also integrate it into the other tech tools that they use, like, where are you seeing the offering change from retainer, and selling hours to packaging and stuff like that? Where are we at in this transformation? Hopefully, that's helpful setup.

Barrett King 9:19
Now it is I think it's too early to decide whether this is something that becomes a service or something that it becomes an internal tool. And the reason I say that is I think there's this this expectation that AI is going to boil the ocean. Like instead of saying make a cup of tea, we're saying I want you to boil the entire ocean. The reason I describe it that way is like Jasper is a good product example of that, right? Jasper does something cool. I don't know. It's super intimate later. Some former HubSpot is there so I've got like anecdotal feedback, but I talked to somebody recently. She was a marketer at a tech firm that I advise she specifically uses Jasper to write baseline copy doesn't publish that copy like your point earlier. You just give the tool this thing and say like go use my audience voice for me, but Jasper is Smart, it's learned her voice over the last couple months she's used it. So now when she says write me a prompt on x, y mu PR, you know, press release on X or Y, it does enough in terms of like the foundational work to get her to a position where she can then go and refine it, and then publish it. And so I translate that to agency and I say, okay, cool. Most of the agency, frontline service delivery role is done via copyright, SEO, optimization, etc. So let's go first from the enablement in the margin component we're talking about, what if all of your copywriters everyone that writes a blog, anything long form, those are developed email, copy, etc? What if they could have 75% of that work done in the voice of your customer, because it's already this AI is already consumed your you know, your prospect your customers content, it gets a new baseline. And so now, it's not about the I don't mean to minimize but like the meaning of the menial task of like, let me go and just start from scratch. What if it gives you enough baseline that you can then go and really provide extreme value on top of that? What if your agency becomes not just a production house in terms of volume, but this really curated, you know, audience specific customer specific language with vertical experience, etc, because AI feeds so much of that mechanism into your already talented writers. So there's like, sort of component one, component two becomes maybe the other end of the argument, which is, you know, when this stuff starts to be prevalent in will use SEO, right? When you think about like optimization as a solution? It's there's a bunch of companies popping up. What if now, you can say, let me SEO optimize, you know, an entire web presence of a customer with the push of a button? And then it's not just going to do the work for you? Because I'm not sure we were at that point. But what if it could just then go back and audit all of your teammates, you know, potential impact in terms of mistake or possible potential good outcomes, and say, look like, here's the stuff that, you know, you you spend two hours doing this work? The agency can build on those hours, obviously, we run this machine learning that takes 15 seconds and gives you that?

Jared Fuller 11:59
Bear, I want to I want to pause you right there, because I think I get what you're saying I think the audience follows immediately the question that I would have, as it relates to that, because it's a compelling pitch, is that helping them sell any more or save any more customers? Like? So that's the agency's problem, right? Their margins are hurting their ability to sell, but like, I'm not sure that that helps an agency win any more deals,

Isaac Morehouse 12:21
it's always shocking, to be reminded of people's willingness to pay for things that they could do themselves, that's always going to be why that's always going to be a part of every economy, because opportunity cost is different. Right? So the difference in, you know, like, you can make a self serve tool for me where it's like, look, now, it's really easy. Changing your oil is as easy as however easy you make it, there's still going to be a bunch of people that will say, Great, I'm still going to pay somebody to do it. And now the people that do it, they just have an easier time doing it than they used to. Because how 10 minutes of my time might be worth more than the cost of paying somebody else based on whatever else I'm doing. Right? That's always the case. So no matter how many self serve tools, I mean, what are all the SAS tools, they're all theoretically self serve tools, guess what we all do? We hire third parties to run them for us, hey, I'll walk you through the setup wizard, the do it yourself setup, I'm gonna hire somebody to do my DIY setup, right. I mean, that's basically what this is. So I, I have not quite seen evidence that AI is fundamentally different. It's all of these new things. They make certain things easier. And they create, they create new opportunities for service providers in some areas, and they replace service providers and other areas. But I still think you're gonna have Jared, we just had a call with your wife's company, notably the other day, and it was incredible walking through some cool stuff that they're building there. But one of the things I immediately noticed was she was showing us some things that these AI tools could do. She wrote these incredibly long, detailed prompts that got us this end result.

Jared Fuller 13:58
Yeah, they're like two pages, Barrett, two page prompts to like, spit out a blog post on a summarizing a call like this,

Isaac Morehouse 14:04
right? Like, we could have done that ourselves. Right theory, there was no barrier. Oh, look, all you need to know is the English language. Yeah, but I don't want to sit there and write that prompt for all that time. I'd rather just pay you to do it, because you already did it really well. Right. So like, there's a there's a way in which there's, of course, there's going to be new opportunities here. I don't want to set up the AI stuff myself, even if I can theoretically. So anyway, just just as a high level observation.

Jared Fuller 14:28
That's that's a fair point. So I'll I'll tie this all they get buried. The question then becomes is the price need to go down like what's gonna help agency sell? So I think Isaac's valid point is very valid, but I'm still lacking the what is going to make an agency more successful right now. And then visa vie all the people trying to partner with agencies, that's the thing you pay attention to is what's going to make the agency more successful.

Barrett King 14:46
Given I think the tool does that. I mean, I mean, it should at least I don't I don't disagree with Isaac. I think the agency couldn't add more services Jaron right, they can wrap something in that brings in AI as a service or whatever you wanna call it right prompt writing as a service. Maybe someday. So

Jared Fuller 15:00
it needs to be not just reduced gross margin, but also more customer value.

Barrett King 15:04
I think it's got to be both because the thing is like you so you can make more money off of it, that's fine. And there's probably going to be a whole suite of tools, you'll see I touch a variety of industries, just like makes you better. Right, it gives you the prompt I think we described. But it also asks you to have the other side of the same coin. I don't think industries survive when they're just single dimension.

Jared Fuller 15:20
No, definitely not.

Isaac Morehouse 15:22
Are you seeing any agencies right now examples of agencies that are like immediately throwing into their service offering, either on the front end and advertising, hey, we'll help you set up AI tools and implement them or just on the back end, using them to make their work quicker, but not advertising it as a new service?

Barrett King 15:40
Yes to using? Absolutely. Like, I've talked to countless even HubSpot partner, Salesforce partners that have all made comments of where it's frontier, though, we don't know specifically what we're going to do. We're going to go by license and try it for six months and see what happens. The other sides, I think, very interesting. There have been a few that I've seen targeted ads, say if you're exploring AI, we can help you. I don't think that that's I mean, that is tip of the spear like there's so much learning to be done. How do you monetize it? I mean, how do you and then what are the like, frankly, just legal and other concerns that come with using a tool like that. And you could do that for hours. But I think that's probably the other, there's an industry that will show up servicing both the onboarding, setup, and then enablement of AI. But I think each of those AI tools so far has been an attachment to some form of an existing industry, until it really starts to break out significantly, we won't know how far it goes,

Jared Fuller 16:30
I want to frame the the shift, so to speak, that I'm trying to get to is I feel like those folks that are trying to partner with agencies right now are really focused on their challenges as a software vendor, right. So like, whatever their go to market challenges are or their success challenges or whatever they are. And I'm really trying to paint the picture for the challenge of being an agency right now, in a lot of ways partner hacker, you know, how we started was, in some ways, like we're a media company, but also an agency like we had to, we had to, you know, add some additional value to our customers, ie partner tech companies. And still do speaking as partner tech companies, Isaac, I'll do a mid roll shout out for my matrix. So I'm excited to for our partnership with my matrix, this episode's brought to you by matrix we have a lot coming with them. So I'm just gonna give them a general shout out, go to my If you're on the enterprise, or channel side, listening to this and trying to scale your program, more coming around our partnership with my matrix and partner hacker, but the thing I want to get to is, when I told this to Peekapoo data, he told me, Barrett, I've never thought about it that way. But I think you're right, is that the shift that HubSpot captured was the same shift that let's say zero, the accounting system captured and they stole a lot of QuickBooks partners. And what was that shift? From project based one time billing, to recurring billing, right. So what industries could go through that similar transformation? marketing agencies, because there's dozens of 1000s, hundreds of 1000s, globally, right, and accountants, there's dozens of 1000s, and then hundreds of 1000s, globally. So Xero stole all of the QuickBooks, you know, quote, unquote, partners that were installing CD ROMs, right? And one time like, Hey, let me install QuickBooks for you. And I'll do your taxes at the end of the year, right to zero going, No, here's your monthly Reconciliation Report. Here's your this and those value added monthly services. The only other industry I could think of that still has that type of disruption, from one time to like monthly value add would be like legal. Yeah, right, that still hasn't really been disrupted at that level. But now my question is what's after that thing? For, let's say, the agency space, we had project base to retainer base to? Is there another shift that generates that type of value for an agency like I would have made millions of dollars? If I had a retainer based model in my agency? I just didn't know. Is there something? What's next after retainer?

Barrett King 18:53
Is it a I know you're asking me a question? Is it AI? Like, does AI as a service become the next frontier? Because it's, if you look at the number one problem plaguing in any agency right now, yeah, we can talk about margin. And it's, it's human capital, enough skill to do enough skilled work. And so, I mean, we could speculate, does AI get to a point where it's intelligent enough and learns fast enough to say, we'll let it loose on our internal server for two months, and it comes back and it creates, you know, actual agents that produce at the level of a human being probably at some point, the rate that we've been going, like, I speculate that there's a world in which you could say, you know, agency is injured, like, what if your agency was seven to 10 people that could produce at the level of an agency of 30 to 40. Today, yes, your margins are better. But it's like it really interesting probably, because you still have to have that first shift always taking place which is moved from project to retainer, right now, what if you don't need more, like needed disruption is actually that you can be agnostic will take on your ongoing services will do retainer work. We can do your projects, too, though, because we can redesign a website in six weeks instead of six months and it gives you that same Polish An output that gets interesting. That is the

Jared Fuller 20:02
first interesting thought I've heard to be perfectly honest about the agency space in the past two years is that project to retainer could go retainer to project. And look at the total LTV of that account. And obviously, that retainer is what pays the bills, and you keep some margin. But then you're that go to have like, Hey, you're the agency that can execute the project. And this is where I think AI is interesting AI as the project manager, right, that helps you go out and pull resources together integrates. And that's where things like, oh, gosh, what was the GPT? Auto GPT, for example, could like, Hey, you have something like that. And they're helping execute a project. So you hire this firm, and you're like, hey, every marketing campaign, I know, they can just help execute 10 times faster. But they're also just doing this retainer of work. That is very interesting, because my needs is a business change on the fly. Isaac, you're a CMO now, like, how many things in the last 30 days could have you used a really like, Hey, here's my, here's my SWAT team that I'm paying.

Isaac Morehouse 21:03
No, I'll tell you I hate I hate. monthly recurring, hey, you pay this agency this much per month. And then they're going to do that. I'm like, I don't want that. I want to come at you project by project. I like paying by part because I don't know, sometimes I might not have anything for months. And then all of a sudden, I got a whole bunch of stuff. Okay, I'm getting a little crazy with this thought here. So this could be way far afield? Let's go No, do it I'm in. But you know, the way that like, and it's sad that Enron got associated with this. And it became such a horrible scandal, but the way the energy grids, okay, where are we going with this one? And then and AWS, right, same model, where it's like, we're gonna we're gonna put resources wherever they're demanded the most at any given time, whether it's electricity computing power over, imagine it. Imagine a model you typed, when you said this, what got me there, you said, Imagine you have an agency and some of your agents are like AI agents, right? And then those are, those are human resources. And one month, you only have the demand for two people to work in your agency, you don't have enough work. But another agency has all the sudden a huge spike in demand. Can you can add those resources to be allocated and the almost the agency, it's like

Jared Fuller 22:14
that this is very interesting now. Yeah. Oh, yeah. It's like an

Isaac Morehouse 22:17
agency of one. That's all networked. And individual contractor, it's almost like the company itself is broken up into like, a flow of individuals where they need to be,

Barrett King 22:30
I think it lets you master being a specialist again, I mean, so much of the work is distributed, to your description of of being able to decentralize it. What if you've got five, specific, you know, like, you've got a head of sales, head of CES, Head of Content, you know, how to social we can keep going. And each of them has their, their agent army, they can spin up as many heads as they need up to a certain limit over a period of time. I mean, it gets really interesting when you talk about the ability to produce without a human being at the steering wheel. Like, I remember being excited the first time I got in a car that had cruise control, because I'm dating myself, but like, I'm 38, right? Like, my first car didn't have cruise. I was probably, I don't know, 8090, my second or third car had it. Oh, shit, this is cool. Like, all of a sudden, the car can drive itself, not really all it did was keep the gas pedal down. So think about that. Like what if all of a sudden your agency in five years, three years, whatever his timeline horizon is, you could not have to struggle with human capacity. What if you could focus on quality work or whatever your specialty was, leverage an outcome and lean into this solution as a way to do that, it gets interesting,

Jared Fuller 23:34
that is such a piercing insight. So from project to retainer to being able to go back from retainer to projects, right. So like, maybe there's a baseline of scale or service agreement at scale, because that's what's so terrifying about being it, like having that latent demand in your because like, as an agency owner. I mean, the only thing that made me sleep well, at night after the first two years, were the clients that I had closed that came back for more work, right, because it wasn't a net new sales cycle, you know, from the beginning, but it was also still stressful, because it's like, oh, shoot, I just had three projects pop off, like, I'm going to be working, you know, a solid 7480 hours this week, because I don't really have that demand. But then that allows me as a business owner, to not think with those constraints anymore, which, which is what the market wants, like, we want to be able to execute projects that are ad hoc and are more responsive.

Barrett King 24:31
No does it because all of a sudden, now you think about it being so we'll use the existing terms. Your retainer is an ongoing services contract of some kind that says you have access with a

Jared Fuller 24:41
defined scope is the other issue. I think it's the benefit. And the issue is you have to have a defined scope. So that way you can staff it and deliver the service and the customer knows what they're buying. But then three months, half of it could be out of date.

Barrett King 24:55
Well, and then the access point like I said, Well, what if I don't need that person this month? What if I don't have that I only want to write one blog, not three, like all of the flexing that takes place within the service set is a loss margin. Now you have to worry about that you can reallocate me, I mean, I can speculate all day and say maybe it's still hourly based, all of a sudden, you get back and say you can buy a certain set of hours, you can distribute across different AI solutions. But it gets super interesting, because it's such a super, super valuable part of our economy, the way that agencies deliver value across software and services and everything else. But it's such a just, again, human capital strapped and just super reduced service offering that now you get flexibility. You get verticalization, you get specialization, it gets interesting.

Jared Fuller 25:40
What this made me think about is I don't know that the phrase is necessarily right. And I know that there were a lot of models where this is it's working very well. And then other places, it's not working as well. But if I think about it, it's from project to retainer to consumption. Right? And what what I meant what I mean about consumption, consumption is typically thought of in like a very narrow context. I think Twilio is a great example of consumption, like just nailing consumption plg out the gate, like, Hey, it's 10, you know, it's a penny, you know, if text at 10 cents per text, right? Like, very easy to understand that model and overtime, right? Oh, we're paying Twilio? 20 grand a month, like, okay, maybe we should probably finally talk to someone there. Right? Like, that's how you would that consumption would work there is it's pretty linear, or it at least is corresponds to your product growth. But I think from the service world, you've never had that ability to have that kind of scale. But I think that ability is the thing that an agency would be really well, I think position to differentiate in the market is that, hey, we have a baseline plan, so to speak, or, you know, service or retainer. And then here's all the ways that you can spin up your marketing projects, your Comm, your campaigns, your resources, your whatever, as agents, and then you're paying for the thing that you want.

Barrett King 27:02
Why didn't the risks, so that's the thing, right? So there are models like that I've talked to agencies that do what you're describing, what they try and do is bring enough customer demand at the door that they can move resources accordingly. And then they try and subcontract, and all these different things. It doesn't work, like you always get bitten, the end doesn't work you do. The math doesn't work.

Jared Fuller 27:18
It's so hard, so hard to execute. He can't know that the math is terrible. It's terrible on that, because the client demand is always is always in the in the, you know, the the, the front windshield, that's the client demand, you're driving right into it. And then everything else is like often in the rearview mirror. Like, how long does it take to get a new resource, spin them up, get them into your tooling, get them you know, how you want to work, how you want to run your business, like that client demand, boom, all of a sudden, you're looking in your windshield, and the client demands behind you, you know, you're like I wasn't responsive fast enough. And then it screws up your business.

Barrett King 27:57
Exactly. What if you can be nimble. But what if you're nimble, what if you go to this, this platform and you say, I need to turn on a new SEO specialist to content production specialists. And somebody you can help us do frontline prospecting and like, you know, demand gen. Great done costs $100 ahead per month for you to run those on the server. And then you can go and charge $1,000 per head to go and optimize that service for your customer. Like, there's something to be said around the expertise of managing these systems. And that's why I think, you know, to kind of full circle this for a second that the AI component of it to your earlier statement. And I think it's this is valuable that you said it in different contexts, like AI is not going to replace a person today. But I think there is a world in which it can replace the functional tasks that people have to do, especially in agency where it's just so button clicks specific, like just that mouse click moment, what if you can pull so much of that out of it, reduce the overhead cost, improve your utilization, drive a better margin, and then get the flexibility that we're talking about on top agencies become a serious player and component of any go to market now, across SAS and other industries,

Isaac Morehouse 29:00
agencies serve this really nice, it's like you have a bunch of bricks, and they don't perfectly fit together. And the agencies are like the mortar that you can take stones stones is a better example, you can take stones that aren't exactly the same fit, but you get the mortar in there. And what I mean is like you have all the software tools that are built for, we need to be able to serve a really big market. So that means we need to have sort of least common denominator, however you want to put it type of use cases. But every individual company has a use case that is there is no such thing as the ICP, right? It's ideal. It's not a real person. So every single person user has something slightly different than they want to do than what the software was built for. And so that in between, that's where the humans come in. That's where the agencies or you know, whatever come in, to help that and I think the AI tools, they might shrink that in some ways by letting there be an additional layer of software specialization, but it's still going to have to be the software tools still are going to have to serve more generalized buckets than you will ever get With just human to human, so it's, it might like, make that layer thinner in some ways, but you're still gonna always have to have that in between, like, who's gonna play? Who's gonna play Ambassador arbitrage go between between me and these AI tools? Absolutely, I'm gonna want somebody to do even if I learn one of them, I want somebody to be like, Oh, I know six of these different tools, and I will be the orchestrator of all of them for you.

Barrett King 30:23
Yeah, I'll tie it back to remember like early web development, right? When people were learning different languages. And then you had like, Flash show up, and everyone said, Oh, Flash, man, it's gonna change the game. It's drag and drop, and plug and play. And like it didn't last very long. I think that there's a world in which AI specialization which everyone obviously talks about, but specific to the agency space, and the ability to say, We are experts, Chief ninjas, at this, this very verticalized, hyper specific, whatever it is like industry or segment of an industry. And we use our tools that we are masters of maybe we write our own language at some point. And we do our own learning model development. And so this is probably the like the SI version of the hyper technical agency that sits in that space. But the creatives now can take advantage of hiring individuals who are experts in these different languages and doing bunny ears, but like, these specific types of bots, and use cases, and they can master driving it. So it's almost like the, the tour operator that knows the history of the space. They're not driving the vehicle, but they know how to tell that story. That's interesting.

Jared Fuller 31:27
And the differentiation here is what what also starts to get very interesting, right? So the differentiation becomes, I mean, we're, we're quickly getting to a world where, you know, buying a software service is no different than walking down the laundry detergent aisle, right? I mean, it's like tide versus gain versus, you know, for like, whatever like brand is what's differentiating in a lot of ways, but at the same time, it also feels like, how are people buy? How are people buying right now is they're turning less and less to Google. They're turning more and more to people who've done the thing that they're trying to do well, the people who are more willing to, let's say, align to your performance, the things that you care about, I've always felt that agencies are great to partner with, even if you don't nail them, because they understand your customer as much as anyone. What do I mean by that is that agencies are typically beholden to customer outcomes, much more so than, you know, SAS companies, SAS companies are held to utilization. I bought 20 seats, who were those 20 seats activated, like, oh, it was activated, that counts, annual subscription, boom, recognize the revenue off all access, is how I recognize my revenue for the year, like, lucky you, you know, as an agency owner, you're like, lucky you, you know, my clients don't want to pay me for work I already did, you know, much less like they're paying you for work you've delivered no value on. And it seems to me that that consumption concept, like from project to retainer to consumption could also correspond to from project to retainer, to consumption, or maybe performance. I think performance based things where like, for example, if you're doing that, and you have a very low margin thing, the upside, that some of the SAS programs have in a 20% commission on a 50k deal. If, for example, you're a marketing agency, and you're, you know, maybe driving 1020 3050, like that, that volume on the performance side, if you can not just monetize the retainer, or the project, but capture some of the value that you're creating. That's also seems much more possible in this world.

Barrett King 33:42
So to translate the language of value differently. I mean, if I think about what agencies do, and where they impact to your point, they, they know your customer, they know your market. And so if you're going to go to market and say I want someone to do something, it's interesting, I just talked to a business of the day there 45 people to get an $8 million round trying to grow their organization, he's like, we're gonna go hire a marketing leader. Why are you doing the TLDR? Long, short, like, we need marketing. So why not hire a marketing leader, like go hire a content producer, first, get your voice out there a little bit, do some brand awareness? Long and short of it is like, well, what if you can hire a specialist agency in five years that has mastered the AI version of that. And I think what's interesting is this race to the bottom you see, with the commoditized services, it's a differentiator, like you said, Jared as well, but also the ability to empower agencies again, to take back control a little bit, and drive the specific Pricing and Value conversations they want, but also with outcomes to your point results, and be able to say like we did it efficiently, effectively, and with the specialization you're looking for. And that's why you hired a firm like, you know, cool or whatever, we end up calling this thing. You know, it's interesting,

Isaac Morehouse 34:44
Jared, when you mentioned legal services as sort of playing around with that and thinking about any of those services and a lot of the things that agencies, marketing agencies and other agencies do is similar to this where the It's not like a It's not like a repeated thing with the same outcome every time. I mean, accounting is like this to that my brother used to work at an accounting firm, and they would do this exercise is like a joke, they would have the same tax return run by like six different people, and they would all come out with something different. It's like literally, like, if you asked five different lawyers to drop a contract, they're all going to be different. Now they all work. They're all correct enough for the market. But what matters? So you think about that? And you think, Okay, well, could you replace something like that with an AI, you couldn't do it with a general AI because there's too much context, but you could with like a really niche down. Like one law firm says, We want AI to study all the contracts that this one guy writes. And now you can have things that basically look like his contracts. But here's where here's where it falls apart. Now, there's a problem, right? Because even the human drawn up contracts, they have typos, they have errors, they have mistakes in judgment, what works is that I get to go back to the human and say, Hold on, there's a problem. And they say, Oh, let me fix that. That's where you want the machine to do that front end part. But you want to talk to a real human when it's broken, to make the judgment call, right. And that's where like, I think you could see some of these things where you like an agency has, we have our own AI that does these components in it, and it's got our own voice and all this stuff. And when it breaks, we have a human at the back of it that's managing it. So you're not literally dealing with a robot all the time. And you get to kind of like soften that friction.

Barrett King 36:23
You're already seeing that. I mean, there are people that are being hired over and over today talking about people being hired to own the inputs, essentially to the AI. There's a company I talked to recently that said, well, GPT is cool, but open AI is open. What if we built our own model, and we trained it to only exist on servers where you paid for access. So what if you had your own dedicated AI model consumed only your content had only your voice and to your point, Isaac, you could police what it didn't did not do I mean, there is a very interesting evolution that will take place where we talk about you said brand earlier, Jared like wouldn't brand because your marketable value your agency's AI system, is your identity. And so now the marketable in differentiating components of this become the not just you know, ability, but the access and the voice and the tone and the experience over time that these things will have.

Isaac Morehouse 37:16
Now it's getting interesting. It's crazy, because it's, it's a very real market, and it already exists to some degree. And there are steps that we can see existing very soon. But it doesn't take many steps beyond that, to start getting into like crazy sci fi space either, which is kind of fun. It's like, there's this grounded reality, but there's like we're closer than ever to some pretty crazy stuff that you can start to envision.

Jared Fuller 37:40
One of the things that the AI people that I kind of trust in terms of like building a more resilient business in this era has to do with the uniqueness or resiliency or whatever you want to call it of the data set. Right? So for example, any data that you can have that is unique that, you know, you can use for a model for a GPT for something like that, like you're going to be in a differentiated or more resilient position, right. So like a proprietary data set, for example. And I think there's plenty of situations where that data set somewhat emergent. I think the thing that is interesting to also think about is going kind of to like HubSpot for instance, it's like HubSpot is a CRM company, you know, not not to pigeonhole you. But like, you went from workflow and automation to down the stack to data, right? Where you have all the contacts, all the accounts, all the opties. Heck, now, all the partners, you're you're increasingly grabbing more and more data. So as an ecosystem, you're like, hey, I actually can or as an agency, for example, I feel like the Why are there you know, hundreds of 1000s of marketing agencies and how many sales agencies, I mean, hundreds, right, it's like exponentially different. And I feel like there may be this massive shift where it's like, okay, now marketing has access to all of the data through CRM, and then can automate all of these things have access to the systems of record. And all here's where I think I'd love to very get your point of view on this visionary future, all of the tech that also connects into it. Right. So all of the integrations and all how that data, you know, it's like we're using 40, SAS tools on top of HubSpot. That seems like okay, that proprietary data set, there is a service to sell into this company to help them make all of that work better together, the age of the connected customer. When is that world going to happen? Like the agency that is selling some AI stuff and facilitating projects, and then also, not just doing that for like the blog post, but the continuity across their entire data set, so to speak. I'd love your opinion.

Barrett King 39:50
People have to get comfortable with the idea of machine learning, not just being a fun phrase to use to sell a platform, but actually enabling it to change your data. Like Your point of access, I think what's interesting is imagine like you know about Salesforce dynamics, any CRM now, while it is a closed environment, you know, the company obviously has access to it. So what if, as it stands today, a company can say, well, we'll release a new feature, it's got AI, and it does some cool stuff, fine. But what if it gets to a level, which is starting to where now to your point, you could say, as an agency, we want to distribute, we want to deploy our AI with our proprietary, whatever you want, like hammer and nail, just keep it very simple. You're right. And so ours has a hammer and nail function that we can execute on, we deploy it into your CRM, it runs for three days, it brings back and extract specific tactical value. And that's our core differentiator and the agency next to us, they use a wrench, and the one next to them, they use, you know, some sort of an air compressor, whatever, right? When using tools agnostic, but the point is that, like, the differentiated value becomes about the functional ability of the AI to deliver something on top of that dataset. And so yes, cleanliness is one and yes, you know, today we think about like, what if it could help me extract my highest value customers and you go three or four layers deeper? i There's a company I advise right now I talk a lot about with them. The ability to to look at directive behavior, what if you could measure emotional feedback from a customer across all the plugged in stuff, you got Gong and you've got Clary and HubSpot and Salesforce and all these different platforms, and deliver. Now we're talking about not just insights on what happened, but what if you could tell somebody where their customer was going to be in three days, what have you could tell a customer where they were going to act before they acted. So AI, not just as a service, but AI as a solution to anticipate to articulate value before it's even in front of a customer. And that gets really interesting.

Isaac Morehouse 41:40
And that's yeah, it's incredible. As you were talking, I was just thinking of like, just little stuff, just crazy little stuff like, okay, you've got you're gone call and you're talking to a prospect or a customer in there. And they mentioned, oh, I'm going to be at Sastre next week. And that's automatically scraped pulled into your marketing and you have a live campaign that automatically targets the zip code that Sastra is being held at with a paid ad on LinkedIn relevant to what you were talking about with them in the conversation. So now you've got like your Account Based Marketing, you're it's like, you're just pulling in having things that can pull from all these data sources that are existing more and more and the combination, because it's English language based, at least a lot of these early models, the combination of the qualitative and the quantitative, like when people speak in normal language, not meeting them to speak in tags that are predefined by you in your you know, automation, but like in normal language and the AI being able to take that and map that onto your workflows. That's crazy.

Barrett King 42:38
tujhe there's, I know we're close to time, but quickly hear like, there's an interesting, there's a business that I've talked to who has their sucking up CRM and recorded call and written word data right now their AI model, trained across all of it will look for early behavioral indicators. So imagine you're in a sales process. It's not there yet, but it's close. And imagine in that sales process, you get an alert on your screen, through whatever zoom or HubSpot or whatever you're using, right. And it says the customer is indicating X, you should say why now you actually can influence human behavior, because the AI is trained across so much data. It's an interesting anecdote around when the whole idea of open AI things are to really blow up. Everyone got really nervous. And one of the comments from that early I forget his name, but the guy that was early into AI, he said specifically, the danger is not that it's Skynet, today or tomorrow that's taking over the world and destroying us all. The danger is that now one specific individual, if you will, one person, if you will, has the knowledge of 10 or 15 or 20 in the power that that indicates. So not only you influencing human behavior, but it's all the outcomes that come with that. And the directive nature of the Way agencies drive brand awareness, buyer behavior, etc. Becomes Ai powered, I mean that the power of that is infinite.

Jared Fuller 43:54
Very King. Thank you so much for coming on, partner up on a fantastic conversation about the future of artificial intelligence agents and agencies. I think there's our tag right there Isaac for the episode, the intersect of agents and agencies. All right, partner up Barrett. You're the man I love working with you. I look forward to seeing you inbound to partner day. We'll be doing some cool stuff there and definitely hanging out at inbound this year. So folks, we will see you all next time. Peace.

You've successfully subscribed to PartnerHacker
Great! Next, complete checkout to get full access to all premium content.
Error! Could not sign up. invalid link.
Welcome back! You've successfully signed in.
Error! Could not sign in. Please try again.
Success! Your account is fully activated, you now have access to all content.
Error! Stripe checkout failed.
Success! Your billing info is updated.
Error! Billing info update failed.