Running a marketing agency today means constantly adapting to a never-ending stream of platform changes, algorithm updates, and emerging technologies. Just when you’ve mastered Facebook Ads, suddenly TikTok demands attention. You finally understand SEO, then Google changes everything overnight. Each pivot requires new skills, different team members, and updated client strategies—all while maintaining profitability and keeping clients happy with results. The pressure to stay relevant while delivering ROI has never been more intense.
Today, we’re fortunate to have Darwin Liu, Founder and CEO of X Agency, a digital marketing firm that has not just survived but thrived through 15+ years of industry transformation. Darwin leads a powerful team of growth hackers, engineers, writers, and storytellers dedicated to driving profitability for their clients. His agency culture emphasizes excellence, curiosity, and integrity—qualities that have helped X Agency navigate the evolution from early social platforms to today’s AI revolution while consistently delivering results.
Social Pulse Podcast host Mike Allton asked Darwin Liu about:
⭐ A framework for evaluating industry changes and determining which require immediate adaptation versus which can be monitored or ignored.
⭐ Practical strategies for managing the financial aspects of agency transformation while maintaining profitability through transitional periods.
⭐ Insights on building a flexible, adaptable team structure that can pivot quickly as digital marketing continues to evolve.
Learn more about Darwin Liu
Resources & Brands mentioned in this episode
Full Transcript
(lightly edited)
Take us back to the beginning. What was the landscape looking like when you first started compared to today?
Darwin Liu: Yeah, so a quick funny story for you is I graduated from college in 2008, right? And when I graduated, I googled how to get rich online, and that was where internet marketing was the wave.
So when I hopped on Warrior Forums, it was big, and that’s where I learned my craft. Back in the day, it was like the wil,d wild west. Google CPCs were somewhere around 8 cents to 30 cents. I was able to promote an affiliate, but I’m going to call them scams. They’re not scams; they’re sort of free rebuild offers, but they didn’t have all of that, all of these security protections nowadays, where we do something like that, and Google automatically flags it.
Facebook just came out, and that’s actually where I made a lot of money in black hat when Facebook first came out. They didn’t have all of their tech team now cleaning everything up and figuring out what people are doing on their platform. But back in the day, I would do free email offers for iPads, where if they entered their email, they would get something.
But it’s actually a rebuild that the key will do these offers. They used to have these back in the day, but not anymore. I would just create a quick script on Facebook, which they allowed back then. Nowadays, they don’t, which first forced them to add all their friends right before they could enter their email.
The second step was to enter their email. I would get paid a dollar 50 per email. And, I was making hundreds a day just from doing something like that. Nowadays, you cannot do stuff like that. Nowadays, they have cleaned all this stuff up dramatically, and we’re white hat is the sort of way to go now. And a lot of these black hat techniques have been removed.
Mike Allton: Yeah, I blame Cambridge Analytica for ruining it for all of us. Yes. Oh my gosh. It’s funny ’cause like I said, I’ve been doing this for a long time as well. One of my big moments from a negative perspective was when Google+ shut down. I dunno if you remember. Google Plus, some of the folks listening might remember it. Most of you don’t. It’s been gone for over a decade, but when it shut down as a major social network owned by Google, I had a quarter million followers. And that was a major source of not just interest in engagement, but actual revenue for me.
What’s been one of the most challenging shifts that you’ve had to face in your industry, and how did you overcome it?
Darwin Liu: Yeah. And then, I wanted to backtrack to your Google Plus piece, right? Yeah. This is one recommendation, and there’s a, there’s a question I get a lot from folks, which is, there’s the Amazons, the Facebooks, and the Pinterests.
Why aren’t we just focusing all of our dollars and marketing all that on, say, Facebook? Because that’s where we get all of our money. And my response always is. Because at the end of the day, you don’t know what’s going to happen with the social media platform. Whether they change their algorithm, whether they remove you from ever showing anything again, or whether they go defunct, you do not control any of that, which is why we always say email, even though it is a dinosaur, is still something that our clients should a hundred percent focus on.
But now, if we are talking about the charging industry shifts, when I started, we used to run paid ads on manual bidding. Automation was not a thing, and that’s where we built our bread and butter, right? So what we do here is we do scientific slash systematic testing on platform-specific settings. And when we do something like that, where we build our systematic marketing approach, we have this thing called the X system. We were able to both scale and increase revenue at the same time. And our whole playbook, when I first started doing this for our agency, was built on manual bidding.
Now, automated bidding, keyword targeting, road, and all that jazz on these platforms came somewhere around maybe 2019. And when that came, we had this whole playbook that we’re generating amazing results on using manual bidding. We had to shift our focus, say, “Wow, okay, this is changing, yet our playbook got wiped out.” So we basically just had to build a new playbook; we still have the same ideology in terms of how we do things. But overall, we have this stuff planned out, right? AI is the new thing, obviously, and that’s emerging tech.
At the end of the day, we just have to be early adopters and utilize the tool scientifically tested. See what works, see what doesn’t.
Mike Allton: I asked the question, and I love your answer because I think it’s instructive. That ship has passed. We now have automatic bidding, so that’s no longer an issue for everyone, but many of the agencies listening are helping their clients today. With SEO and the SEO landscape is changing overnight. Google is now shifting more and more towards AI overviews and AI answers instead of this traditional list of responses. And to your point, that means we just need to create a new playbook, and we need to figure it out on behalf of our clients.
Now, one of the things that I think a lot of agency owners are struggling with, but we haven’t really talked about on this show, is the financial aspects of transformation.
Whether it’s investing in a new technology or figuring out new talent, or taking the time to write a new playbook, how do you approach these kinds of transitional periods?
Darwin Liu: Yeah, and no. The funny thing is, this is an iterative approach, right? When I first started, we would see this cool new tech, let’s just say, let’s say task tracking. A perfect example might be using something like a ClickUp or something like a Zoom info for sales. When we first started, we got a sales pitch. Great, this looks great. Let’s go purchase all of the licenses annually. Let’s roll out to the whole team, which never really works, to be quite honest. So after messing up a few times, right?
Nowadays, say we’re using a new platform, a new tag, what we’ll try to do is we’ll try to pilot it, whether it is getting a dummy account to use, or whether it is, if we’re using Clickup, we’re going to say, Hey, we’re going to roll it out, the beta testers within X Agency to see if it works for us. Build the soft stand operating procedures. Off that sort of beta testing team first. Then, if it works and everything’s great, we have the process built up, then we roll to the team. We try not to sign into crazy contracts and try to at least get some trials at the beginning, right?
Because nowadays, what doesn’t matter is what tools you have; you definitely have a lot of companies competing with each other. So one, I would suggest that never hurts to ask for a free trial or a dummy account just so you can poke in. Also, if it’s a new platform, try to get references, right?
So, for big clients that have used it, we try to talk to them if possible. In terms of the financial side, when I first started, it was just buying anything for the budget. Nowadays, we do have a bottom, flat margin that we try to always run on. So if we are trying to bring on this crazy expensive tool, we need to figure out where we need to pull that money from. And if we can’t, then we can’t sign on board yet.
Mike Allton: I love it. So be smart, be patient, be methodical about how you’re doing those things now. Earlier, I talked about your agency, and I mentioned the fact that you’ve got growth hackers and engineers, writers and storytellers, which is unique. I don’t recall having ever spoken to an agency owner who specifically listed Growth Hackers and engineers on their roster.
How has that come about? How has your team evolved? I imagine you didn’t start 15 years ago with Growth Hackers and engineers.
Darwin Liu: Yeah, it evolved in a sense because we needed folks who truly understand the tech side. And what we try to do, we call this sort of we call it algorithm hacking.
There is no real phrase for it, right? And to truly understand the algorithms, doesn’t matter what platform, we need someone with a sort of engineering mindset. Now, are these truly engineers? No, not necessarily. These are coding folks who understand code marketing and how it applies across the board.
And we need the growth hacking mindset for the marketing side, right? And then, we do a lot of ads, and words do matter nowadays; attention spans are getting shorter. So we do need folks who can figure out how to capture attention. Immediately within the first second, which is how all this stuff feels together.
When I started, it was just okay, let’s start a marketing agency, because that’s the field I’m in. And this all came after learning what we needed. Nowadays, what we found if you bring in a web developer, they don’t necessarily understand the tag side and if you bring in a sort of a marketing person who understands the tagging stuff, they don’t understand the coding side, which is why we needed to fuse a lot of these folks, put them in a room together, have them, share the knowledge and just put everything together so we can have a good core products in terms of servicing.
Now, when you started the agency, was it just you, or did you have a team, and how many folks are on your team today?
Darwin Liu: Yeah, a quick story would be, when I did my Black Hat marketing, it went great and all for about two years.
What’s wrong with Black Hat is that you make a lot of money fast, but Google or Facebook comes in, and they find out what you’re doing, and they kill it. And I was sick of that model of finding the next hole that I can fill before it gets plugged. So what ended up happening is I had to work at an agency, and while working at the agency, I started at an entry level, right? Like and what? Every single year, I was promoted. I did build a name for myself in this industry because, at the end of the day, it is a very small industry. So while I was working in the industry I made, I made a lot of folks. They loved what I was doing. When I immediately left, we were at seven figures within the first six months.
Because folks did reach out. They tried to come with us. So yeah, within six months, we were about seven figures. We currently, right now, in terms of full-time employees, I believe we’re about 17. We have some part-time and contracting flow somewhere around 10 to 15. So not too large.
We are trying to grow to the next phase right now. But yeah, we’re still growing. The Terras have hit us hard because primarily a lot of our clients are e-commerce based, but we’re still, as you already know, marketing changes with the drop of a dime. And we are just trying to move with the waves.
I’m wondering if you could share an example of maybe of one of the clients that you’ve worked with, how you’ve helped them respond to a major industry shift, and what kind of results they achieved as a result.
Darwin Liu: Yeah, more of a more recent example would be Connected TV. There has been a big shift into video, obviously video content, and Connected TV. For the folks who don’t, might not know what that is, running ads on the Netflix‘s the Hulus, and the HBO Maxes of the world. A lot of folks aren’t doing that nowadays, and a lot of advertisers don’t realize that you do not have to have a hundred K budget to spend money on this.
So we’ve been doing more CTV for a lot of our clients. It is a big thing, and we do bring it up to our clients in terms of what else they can do. What ended up happening was we have been running more CTV through Mountain, and we are Mountain’s partner, and we have been able to achieve a 4x ROAS on top of branding.
So our first client that we brought on board, they were afraid of using it just because they didn’t know the returns, and we didn’t know the returns as well. What we did was get some case studies tested on some smaller brands, and then really just try to get more case studies through all of the clients that we’re testing. And what we end up doing is using these case studies to show our future clients. But yes, CTV, we have been doing amazing on it. Branding. One client spent somewhere around a hundred, 50 grand-ish, and they made somewhere close to, let me, I have the notes down, but 4x return on investment.
Mike Allton: Wow, that’s amazing. And I’m familiar with that kind of platform. It’s something I’ve been wanting to get into and test. I think I’m still waiting for Hulu to approve my manager account. So I want to test some podcast ads and that sort of thing there. But that’s really cool that you were seeing those kinds of results. What, any particular industry with all those e-commerce or something else?
Darwin Liu: We are e-commerce and nonprofit, so same thing. And non-profit isn’t a ROAS number, it’s a CPL number, right? Because we try to get donations for them. But these are all amazing results. And when I say four x return, I’m talking about GA4 revenue. I’m not talking about platform revenue. And that is a different ballgame. So we’ve run YouTube ads, YouTube CTV, YouTube brand, and YouTube mass heads at a hundred K a pop.
And it never returns that kind of money on a branding campaign. And the fact that these are paying for themselves does prove value. We also use statistical analysis. So I don’t know if anyone knows about Pearson Correlation. Nowadays, it is very hard to sell paid social or anything social, or say CTV, because the higher-level managers, the C-Suite folks.
Don’t understand it, and they say, Okay, show me the numbers. And you cannot, because GA4 does not show the numbers, right? So one way we figure out how to tackle that is to use Pearson Correlation is a statistical model that shows correlation between increased spending from one channel in terms of increased revenue from another.
So we’ll use, say, Facebook ads, or we’ll use CTV ads, and we’ll correlate it with, okay, this is actually boosting your revenue on your direct revenue, on your referral revenue, on your organic social revenue.
Mike Allton: Love it. This is fascinating. Now, I joked earlier that social media we thought changed rapidly, and that AI came along, and it’s changing even more rapidly. The pace of change certainly seems to be getting faster and faster every single year.
Has your approach to adapting to new tech changed, and if so, how?
Darwin Liu: No, not necessarily. Not necessarily. And I think there is that law that they talk about, I just forgot the name of it, but technology doubles, right?
Mike Allton: Moore’s Law.
Darwin Liu: Yeah. Okay. There we go. Yes. And at the end of the day, we try to be early adopters, right? AI just came out, and everyone in the company, we, so we prioritize learning. Everyone in the company is forced to use AI. Use different AI tools to understand it, and we share our knowledge within the group.
So, in terms of changing our approach, no, we try not to focus on shiny objects. We do jump on trains depending on how big it is. And AI is a big train. So we do focus on that. And what we end up doing is it’s the same, quick, maybe three-step process. One is one, early adopters; two is everyone is required to use the tool. Three, we share our knowledge, right? And then four, what we do is once we understand what’s happening, we build our playbook based on it, using scientific tests.
Mike Allton: Love that. As an AI evangelist, that’s music to my ears. ‘Cause you’re right.
This isn’t even a revolution. The fourth industrial revolution is changing how all businesses operate, so it’s not optional. Unlike past social platforms like Clubhouse, yeah. Businesses could skip Clubhouse, and they wouldn’t be losing out on anything.
I’d love it if you could clarify a little bit more specifically how you’re working with clients to help them adapt to new approaches, new industry, and new technology.
Darwin Liu: Yeah, and I’d say depending on the client type, right? We have publicly traded companies that are not just going to jump on the next wave of, say, AI tools you want to use. So we generally try to go to our smaller clients who are ready and willing to do something like this. So one is that we will always test ourselves.
Whatever we are testing, whether it is AI or a sort of new optimization practices, we will run it on our site, our own marketing campaigns first. If it works out, we will then move on to our early adopting clients, who are generally the smaller clients who are not huge corporations that have to go through seven chains of leadership. They are not publicly traded. They’re willing and easy to, they will say yes quickly off the bat, right? Once we get metrics and numbers to prove a case, that is when we bring all this and sell to our larger clients. Because the larger the clients are, the more change that they have to go through, the more change they have to go through.
They’re focused on these top-level numbers, and if you can’t provide that to them, they are never going to make a decision on, say, trying the next new AI tool or trying AI auto blogging or cortisol content, stuff like that.
Mike Allton: I like that your approach to emerging technology is the same as your approach to a technology that just happens to be new to you. Because your approach to OTI is the same as AI, same as anything else, right? You’re going to learn it yourself, and then you’re going to work with smaller clients first and establish not only success rates, but obviously, you’re going to create your workflows and frameworks and build up your knowledge and expertise while you are hopefully achieving success for clients in creating case studies.
Such a smart approach. I hope everybody listening is writing this down and thinking about, yeah, this is what I should do, whether it’s adding a new social network or a new service. Start small, start internally, and then grow from there.
I’m wondering, Darwin, just as a last question, looking ahead, are there any specific emerging technologies or trends that you think the agency owners should be preparing for now to try and stay competitive over the next three to five years?
Darwin Liu: Yeah, and I’ve thought extensively on this, to be quite honest. Whenever I have a chance to reflect, I tell my team this as well, which is that we do not want to be the agency. That is, back in the day, the horse and buggy, right? When Ford came by, people thought he wouldn’t do a single thing, and they kept their horse and buggies, and then what ended up happening is you’re just holding the horse and buggy when Ford comes by with the engines, and you’re left in the dust.
AI is changing the ball game. And what I really do think is the this agency model, while it will still work for larger accounts and say enterprise level or enterprise level clients for the small business folks,the folks who are serving local, smaller, say mom and pop shops who are serving clients, and these agencies are serving clients under $1 million.
The AI bots are coming, right? We are creating one internally for our use based on our sort of marketing strategies and learning from that. But there are multiple companies out there that are right now creating marketing bots, productivity bots, sales bots, all of which will change.
The whole game for say these small business owners, they don’t need us anymore, right? They can just go to a bot for a hundred bucks a month and, at the end of the day, have all their stuff done without paying thousands of dollars a month.
That’s something you have to look into, right? I would say try to add AI technology stacks to your source service models because, at the end of the day, every other agency is doing it. So if you are not, you are going to be left in the dust. And clients are, their interests do peak when you talk about AI. So the more things you can do to stay in the game on top of standing out, obviously ideally by bringing in ai one to make your team more efficient, want to make your marketing more, better and efficient you are not going to be competitive moving into, say, I’d say you give this two more years and the whole marketing agency world is going to change.
Mike Allton: Yeah, sadly, I couldn’t agree more. Yeah. Folks listening might know I’ve got my side hustle. I’ve got an AI Podcast and that sort of thing. And thanks to AI, I’ve got I’m maintaining three websites, blog posts. An email subscription with 45,000 subscribers. A weekly podcast, sometimes twice a week. Social presence on every platform, and I’m doing that in just two to four hours a week. Thanks to AI. I am not paying anybody else. Yeah, to run my marketing, I didn’t know that I had the budget to do that before, but I’m doing even more than
I never used to do thanks to AI, and small boutique agency owners who are used to working with small businesses. Need to rethink how they can best provide value to those businesses because of repetitive routine stuff. And AI is just knocking it out of the ballpark.
Darwin Liu: Yeah. One, and then, last quick example for you is we’ve started really testing AI and blogging, right? Nowadays, I also have a sort of side hustle where I have a dog clothing company, and I tested it on here first, which really was AI content, AI auto-blogging based on your existing content, right? So if AI is a skill, the more you query, you can understand what it does, and you can give a sample content, and you can give the AI bot roles.
And what we are doing is we’re bringing AI into Sheets. And what we do is we use AI for almost everything in terms of SEO research. We are building thousands of long-tail blog terms that have search volumes of 30 to 200 that no one is really ranking for. And what ends up happening is we take it and then we point all of that up to our site.
My dog clothing company, the traffic for organic was minimal, and now it’s at 10k a month, and we spend zero time on the content at all. It is just a straight AI API upload from the sheet so we don’t have to manually upload it, right? The AI can pull the images and everything. Everything is pushed out with just a click of a button.
Mike Allton: Yeah. Like you mentioned, images, I think stock photography, they’re dead. They’re dead. When you can go to Gemini and ask for any image you want, even an eight-second video, now with audio and background music. What do I need us, what do I need Broll for? What do I need a US stock photography site for?
Darwin Liu: I think, and I would say, video and image creation is almost there. It’s not fine-tuned yet. They’re getting better at it. Video creation, I think you guys are still solid for a little while, but eventually, man, they’re taking that over, too.
Mike Allton: Yeah. Darwin, this has been incredible. Very insightful interview.
I know folks are just buzzing with questions. If they want to reach out to you and ask you something, where should they go?
Darwin Liu: Yeah, you can email me at darwin@xagency.com. I always answer questions; it just takes me a while. So feel free to email me. I’ll answer them a hundred percent, answer them, or connect with me on LinkedIn is www.linkedin.com/in/darwinliu/, I believe, or visit our website, xagency.com. I do love professing and helping out, so if you have any sort of questions at all, feel free to reach out, and obviously, thanks for having me on the show, Mike.
Mike Allton: Awesome. Thank you, Darwin. Thank you for listening. Of course, we’ll have all of Darwin’s links in the show notes for you below, but that’s all the time we have for today.
Friends, don’t forget to find the Social Pulse Podcast Agency Edition on Apple and drop me a review. I’d love to know what you thought about this episode and what you’d like us to cover in the future. Until next time.
