6 Signs Your Agency is About to Fire You As A Client

By 2027, 70% of software interactions won't have a UI at all. The button is dead. So this is the whole concept here, Neil, is around designing for agents, okay? So there's this great post from this guy, Teddy Riker. And so if you think about this, there's more and more agents that are popping up. You think about Supabase, the number of deployments that are popping up, right? That's going parabolic right now. So Salesforce did an announcement. They're like, what are they saying? Oh, we're basically going to get— they're going to get rid of the UI. Like, it's— they're not going to get rid of it more so, but they're talking about Headless 360. It's a sweeping initiative that exposes every capability in its platform as an API, MCP tool, or CLI command, so AI agents can operate the entire system without ever operating the browser, right? Even what I'm doing right now with the API for SEO thing, I think even what you're doing with your software too. So I think humans are still going to need to operate UI, like you and I are going to need to look at the UI.

Yes.

We still want to point and click, but I think it's going to be 80/20. So 80% is going to be more focused on the agent. 20% is still humans looking at UIs that are still— maybe these UIs are managing the agents and it's reporting on how things are going over time. So what the interaction pattern was before, Neil, over the last 20 years was you have a user, you have an interface, then you have a database, okay? Now it's you have a user and then you have a user's agent. So it could be using Claude or Hermes or OpenCLAW, then you have a database, right? That's where it's all going. And the key thing is, I'm not going to go through this entire article, but you have to think about whatever it is that you're building, you have to design for agents. So when we're building new websites, what are the .md files across the entire thing? Cloudflare now has a thing where it's, is your site agent ready or is it agentready.com? It'll grade your site. Their site is not even agent ready if you put their site in there.

Okay. So, but with this, I think there's a few things that you guys should keep in mind. One, the status for software interactions. It's not necessarily for your website. 2, it doesn't mean that you can get rid of these buttons altogether because there are still some human use cases. It's more so the interactions will mainly be happening by agents, so the agents won't have to click on the buttons at all, but you still need the buttons there for when a human interacts. Now, that may change by 2030 or sometime by then because most things may be done by agents, but this is talking about by 2027. And remember, a lot of companies and humans don't adapt as fast as technology is.

Yep. So I think, by the way, so Ramp, they launched their MCP, and we'll just talk about one of the things you can do here in terms of designing for agents. Neil's opening his water, so that's the sound. So you want to make sure that you build feedback loops. So when Ramp launched their MCP, they had a problem, observability. So they can see all the tool call volume, but they couldn't see the surrounding chat context, right? So when you're building these, when you're designing for agents, you want to make sure that there's some type of context or rationale on all these tool calls, because then you can figure out what the problems are, right? Your agents are going to run into a lot of these problems. Humans are going to call a lot of these things. You're going to need some type of feedback tool. And that way, it's going to help your agents figure out what the problems are. And so you can build on top of that, right? And then you can have this kind of self-improving loop. And so I would highly recommend reading this. It's called Designing for Agents. It's on X. Honestly, a lot of the best AI content right now is on X. People are sharing the most cutting-edge things. So I'd highly recommend that you go there. That's what I would say it is. And I don't think by 2027 all this stuff is— the UI is going to be dead. I think it's going to, again, I think it's going to be an 80/20 thing. It's going to flip.

And for everyone who's listening, I know a lot of you guys work with agencies, so we thought it'd be fun to do an episode called There Are 6 Signs Your Agency Is About to Fire You as a Client. Most agencies hide them.

Here they are. Okay. So Number one, when they start to ask about getting all their data, right? Hey, can we get all our data? Can we have all the reports?

No, this is about your agency firing you.

Your a— oh, your agency about—

okay.

Okay.

All right.

You start first. I need to think about this one.

Uh, number one, if you start yelling and being mean to people in an organization and they're not as responsive to a hostile quote unquote tone of voice. We've seen that and eventually we fire, we then go look at the contracts and stuff and we terminate people who are rude.

Number 2, I would say is if you are non-communicative, meaning that you expect the world, but then you're not communicating because it's a 2-way street. It's a relationship. And if you want the agency to help you grow, you have to help the agency help you.

Yes. Number 3. You're missing deadlines. So an agency needs you to succeed, not just on the communication end, but sometimes you are assigned tasks and you have to complete those tasks. If you're not helping the agency help you, you will struggle. An agency, no matter how good they are, will require some information and stuff from your staff and your team. It could be minimal, but if you never provide them and nothing ever happens, it's a lose-lose scenario. And here's why. A, you're not going to get the results and you're going to be pissed off. B, it's not going to be fun for them to work on your project because internally they're going to be like, well, the client's not going to ever give us what we need. So we're not going to be able to execute on anything. There's going to be no results. What's the point of even doing this? People want to enjoy what they're working on. And more importantly, they want to see progress. When they see progress, individuals feel more fulfilled, both on the company level and on an agency level. So you need to make sure that you're completing your task.

Number 4, you're unreasonable with your expectations. So I know someone, the company he works at right now, they are working, let's just say it's a SaaS company. Their CEO all the time is talking about how fast Anthropic is growing, okay? Cream of the crop, totally different businesses, but unreasonable expectations, right? That happens for agencies too.

So he wants similar type of growth to Anthropic?

Yeah, it's like, how's Anthropic growing so quickly? Why are we not growing as fast as Anthropic?

Right? Is this a startup or a scaled business?

It's a scaled business, couple hundred million a year.

Yeah, yeah.

So unreasonable expectations will, to Neil's point, they'll cause the agency to burn out and they don't want their people to burn out because their people are their business. Honestly, it's, people are becoming even more important now, so.

Yeah, dude, people are becoming really, really important and if you just have unrealistic expectations, it's just, Not going to end up working. Number 5, and this is one of the most comical ones that I've dealt with, and I deal with this multiple times a year, is when you're asking an agency to do something that is against your data. So if your data says this is what's working and they're showing you, and a lot of times customers argue with agencies saying, no, I want you to do it this way, and you believe it's a better way. Then why are you really paying an agency to help you out if you already know what's better? And I'm not saying you're wrong, but you hire the agency because of their expertise, or hopefully you did. And typically they should be specializing in whatever area that you're asking for help. So if you're telling them how to do the job they specialize in because you feel you know it better, they're probably going to end up terminating you. It's just a question of when.

Number 6, You ask for more work all the time, but you don't want to pay more. So you're constantly asking for more. This is known as scope creep. And so what happens is it's, you expect more communication, okay? You want the world, but at the same time when the agency's like, hey, this is going to cost us more in terms of times and materials, but then you don't feel like paying it, that's going to cause an issue as well because they, the agency does have other options for the most part. And if they do have other clients that are willing to pay it for longer terms as well, and it's a better deal for them, They're going to take that.

Totally agree.

All right, so next one. So basically, Anthropic is paying $400 grand a year for an events manager, not an engineer, an events manager. So this is my good buddy Marc Andreessen retweeted this. So look at— think about it this way, okay? So you have an events lead brand, I'm pulling this up right now, okay? So Anthropic's mission is to create safe, reliable AI, annual compensation for this role is listed below. Basically, they're looking for someone to own the execution of brand experiences that translate Anthropic's values into physical moments. So Marc Andreessen says, when one thing becomes abundant and cheap, another thing becomes scarce and valuable. So what is becoming more abundant and cheap now? Code. Okay. I would say intelligence is becoming more abundant.

Engineering.

Engineering.

Yeah. Like actual engineers that you're hiring within an organization.

Even though we need more engineers now.

No, no, we need more good engineers.

More good engineers. Everything is we need more, but we need good.

Yes.

We need more good. That's what we need. That's good. But another thing becomes scarce and valuable. So what's more scarce and valuable now? Events, in-person stuff.

Yes.

All the stuff that like, because all the online stuff is just, by the way, I do tell my team now, the only thing I'm concerned about with the people who are really good now is actually the burnout piece. So to your point, they do definitely feel momentum, but even for me, like sometimes I just, I gotta slow it down a little bit, right?

Yeah.

So anyway, all that to say is I think this makes sense. If they find a really good events team and they're able to put on amazing events, That's good for them because those are all marketable moments as well. And those are also good for to, to be posted to social too.

Dude, I totally agree with you.

Quick break for agency owners. You know the content bottleneck. Clients want more content, but writers are more expensive. Freelancers are inconsistent and you can't scale. ClickFlow fixes that. It creates production-grade content your clients will actually approve, not generic AI slop. It's content that actually matches their voice. One agency using it reported Saving 90 hours a month per person. That's margin back in your pocket. Plus it handles internal linking and content audits, which most agencies charge extra for. There's a 14-day free trial at clickflow.com. Your ops team will thank you. Back to the show.

Speaking of events, you want to do an old Marketing School classic?

Yeah.

Let's do 6 ways Eric and I are leveraging in-person events in 2026. All right. So number one, when I go to conferences, I'm setting up in-person events now at conferences. So a lot of times where I'm going to speak, let's say it's in Brazil or Hong Kong or UK or whatever it may be, I'll work with the organizer to actually have a private meeting room within their event venue. And they invite the attendees, specific ones that we select, and we'll do a free lunch or— and during that lunch we'll do presentation and networking. And that's actually been working out really well for us.

That's good. What I've always done with events is I like to throw dinners around them. I think Neil's is— I'll say Neil's actually talked about this on this podcast. So use the YouTube API, use it for marketing school. He's given out his entire flow for this, but He throws events that are tied to people who actually are people that are part of his ICP. They have some power in the organization as well. And it's pretty like, if it's a larger company, they're likely to move around, they'll probably think about him in the future. So even having that one touchpoint, it goes very far. I would just say that what's always worked well for me for the events that I typically am speaking at or the ones I'm doing, I always try to throw a dinner around them. Now the thing is, You want to make sure that the dinner you're throwing ideally is people that are in your ICP. Don't just throw it with a bunch of speakers, which is what I used to do in the past. That's fun, but it's not going to drive the best ROI for you.

Dude, totally. When you mentioned how I look at people who have power, Eric's speaking about like decision authority and he really nailed it when he talked about how we focus on executives because They're likely to also get other high-paying jobs at other companies. So for example, there was this lady who worked at Oracle and one of their Latin America regions. She wanted to work with us. They had an agency of record. They gave the work to the agency of record, so we didn't get it. She moved to another company eventually. This is over a many period, year period, like probably like 5, 6 years. She went to another company eventually, wasn't able to get us in due to budgets. And the third company she went to was Thomson Reuters. She was able to get us in.

That's good.

Right? All because she heard us ages ago and she was working at a company that she wanted to work with us but wasn't able to.

Was that your number 3?

No, no, no, that was still number— oh, that was just reinforcing number 1. Like people really forget the long game of building that brand with your ideal customer because as they move around, if you do good work, they'll keep bringing you with them. Or if they like what you have to say, eventually they'll be able to test you out, just maybe not at the current company they're with.

Actually, that's very true. Steve Young's private equity firm reached out and then apparently the guy had been following both of us for many years. And he had actually invited both of us to speak at a private equity conference. I was like, I didn't hear about this. Apparently you didn't either. But the whole idea is like they want to do something now. And then he's— so the whole idea is like, You want to make sure that you're nurturing, nurturing, nurturing. And the good thing about creating content is it gets people to remember you over time. It's getting these impressions, right? So you're actually on number 3 now.

So go ahead. Number 3. So another thing that we're doing when we're doing these in-person events is setting up side events with platforms. So if I'm going to go speak in Brazil, there's a Google Brazil, right? There's a LinkedIn Brazil. We'll then say, hey, I'm in town, we would love to do an in-session event at your own corporation, call it LinkedIn or Google. Do you want to invite some of your big advertisers? We'll be there, we're going to show them how they can even get more benefit from their LinkedIn advertising or their Google advertising so it benefits the platform. They bring in some of their top advertisers within their region. We network and it gives us access to more potential customers.

So for us typically also what we, when Neil and I, we every year we speak at the HubSpot conference and so what happens is we'll do a session with them and there's Q&A as well. What I would say is one year I remember we did a separate session too. So it's almost like we're throwing side events for these things, right? And then HubSpot actually helped us with that. That side event that we did, I think we had 200 people show up to the room. That rated way higher than the you and I doing the Marketing School one. So that session rated a 4.97 and HubSpot's crowd is typically higher.

Wait, wait, which session?

So we remember this was in Boston, the last time we're in Boston. So last year we're in San Francisco, right? Remember the year before we're in Boston. And so we did our Marketing School session, but then after that we did a session just for Marketing School fans to come in and 200 people came in and it was very interactive and it's like, hey guys, get into groups itself, I was basically directing people, hey, get into groups, talk about the latest thing. And we picked a champion at the end and they won like an iPad or something, right? Remember there's the Taiwanese person that mocked the thing?

Did they even win anything?

No, they didn't. We just said, you get bragging rights, you're the champion. Yeah, we're like, yeah, right? But that scored 4.97 and we didn't have to like that. When you foster a community like that, it connects people together. That's something we should continue to do. But our talk that year, you know what it rated?

Much lower than 4.97.

Yeah.

But it was funny on the San Francisco event, my talk was one of the highest rated.

Isn't that funny how it works?

Yeah.

So Neil and I have been speaking together for so long where sometimes he'll rate the highest, sometimes he'll rate low, sometimes I'll rate the highest, sometimes I'll rate low. Like it goes back and forth.

And it's just really random. It also depends on the event because if an event has more technical people, sometimes their expectations are extremely technical. But the struggle that Eric and I have as speakers is when we go extremely technical, you lose a lot of people, a lot of the audience who's Not technical, and then your speaker ratings just plummet.

So last year I did that, that pissed off a lot of people.

Yes.

So like, because the year before I scored pretty high, last year not so much, yeah.

Yes, and a great example of this is during our Marketing School event, Eric started bringing up MCPs, and then I interjected and asked the audience who knows about MCPs, and majority of the people did not know about it.

Yep.

Some people got really mad at me for interjecting, Eric did not get upset. He knew why I was interjecting is so that that way it would give him an opportunity to explain it. And then during his main presentation, it was super technical. It was great and valuable, but people had an issue that it was, I think, super technical.

Yeah, it was, it was too technical. So by the way, it did, it did drive leads for the really technical people, but just keep in mind, you have to, you have to know who's in your audience. You have to know the level. That's why when you speak at these events, by the way, you should always ask the event manager who's in the audience and what would be a home run.

Yes.

No different than like, so Neil and I are both speaking at this, the SoCom eCommerce Conference, which is free by the way, Manhattan Beach, I think May 14th, we're both speaking on that day. So if you guys want to come hang out with us, it's free, you can apply, just search for Social Commerce Summit or something like that.

Continue. I just made my promo video for that event this morning.

You made a promo video?

Yeah, they didn't ask you to make a promo video?

My promo is talking about it here. Oh, this is twice.

So you didn't even make a promo video.

Oh, you know why? You got paid more than I did. I arranged that for you too, by the way. Oh, thank you.

Yeah.

Yeah. I said, oh, Neil can't do it for that. So yeah. You gave a big discount though.

Yes.

Yeah. So anyway.

'Cause this is in my backyard.

Yes. It's easy.

Yeah.

All right. Continue.

If I had to drive, actually I still had to drive a long time, but if I had to fly—

You'd have to still get there earlier. It's Manhattan Beach. It's not easy.

Yeah, I had to leave probably 2 hours before.

Yeah. You know what I have to do? I have to do a mastermind, so I have to go, I have to get there at like 3:00 PM, 4:00 PM. Ah, yeah. See? So you do more of those. I'll, I'll take the mastermind piece.

All right.

Yeah.

Anyway, continue. Uh, what number were we on?

I think you were on 3.

I was on, uh, we just finished 3. I think I'm on, uh, Ethan, you went or no?

I have no idea.

Well, let's go with number 5 cuz we lost track. Yeah. So I'll just go on the next one. One thing that I've been doing with more in-person events, if you're ever presenting, put a QR code at the end. It's an easier way to collect a lot of leads and follow up with them. Don't just follow up with them from an aspect of trying to close them. Follow up and actually try to build a relationship. Eric does actually something interesting and you should do this as your number 6. You get a list of the attendees and then your email to us is pretty creative. You want to break that down?

Ah, yes, yes, yes. Okay. So, oh, what the heck? Okay, I was going to show you like my cost diagram for Claude. Okay, so what I do is we typically will try to ask the conference organizers, okay, if we can get the attendee list. Usually it's a no, but when we're able to get it, we get an email list. Okay, so we get that and then we also get all the attendees. Okay, that's one piece. The second piece is to Neil's point, if you put a QR code at the end, you get all the leads. My sales team can see all the leads that come through that day and that they downloaded that thing. So what will happen is they will get an email, all the attendees will get an email saying, hey, thanks for attending the talk, or maybe thanks for registering for the talk, just so it covers everyone. And here's the link to the slides over here. And by the way, if you want to talk about marketing or if you need help in this area, this is what we do over here. It's very light. That actually does get— a lot of people are just like, By the way, when it comes to marketing services, it's more about timing than anything. And so when you hit enough people at the right time, they're just like, oh, by the way, yeah, I do need services. And they often forget the agencies that they talk to, right? So it's really about keeping yourself top of mind. And I'll say a bonus piece, Neil, is you want to take the raw recordings from your talk and ideally you're clipping them up. That's one piece. And ideally you're taking that long-form video. Because remember what, Neil, when we did a long-form video when we both spoke in Germany, That video alone, 120,000 views for long form. Not bad. It's pretty good. That's actually really good. Rest in peace SEO. That was the thing. Remember that one?

Yeah, also a controversial topic, which helps.

So guys, that is it for today. Please let us know what you think about this new format, new-ish format, and we'll see you tomorrow.