Episode 07 · The Search Signal
About this episode
Marketing budgets have dropped to less than 8% of company revenue, roughly 18% below where they sat four years ago, and the AI line item inside those smaller budgets keeps climbing. Put those two numbers next to each other and one move looks obvious: hand the search work to AI, stop paying for the people, and bank the difference.
This episode tests that move. Michael Transon works through what AI genuinely does well, why a tool every competitor can buy this afternoon can't become your advantage, and what search actually costs to resource at $1 million, $10 million, $20 million, and $100 million in revenue and shares how to make the case to a CFO who wants to know why AI can't just do your SEO.
Michael's POV in 60 seconds
One thing
Gartner puts marketing budgets at 7.8% of company revenue in 2026, about 18% below where they sat four years ago, and CMOs are now putting just over 15% of what is left into AI. Budgets shrank and the AI line item grew at the same time.
So what
That math makes running search in-house on AI look like found money, and it's not. Every competitor can sign up for the same models and the same SEO skills this afternoon at the same price. In a channel with a limited number of rankings, citations, and mentions to win, a tool everybody has raises the floor for the whole category without moving anyone up relative to anyone else.
Now what
Fund search at roughly 1 percent of revenue, and spend it on judgment rather than production. AI should carry the data pulls, the first drafts, and the keyword legwork at every tier. What changes as the budget grows is how much human expertise you can afford on top of it: a few hours of a senior freelancer, an agency bench of specialists, an in-house hire, or some combination.
Questions this episode answers
Can AI replace an SEO agency or an in-house search team?
AI replaces a layer of the work, not the accountability. AI handles the production layer well: data pulls, first drafts, keyword research, and reporting, the work a junior analyst or a freelancer used to do. What it can't do is take responsibility for a result.
How much should I budget for SEO and AI search?
Roughly 1% of company revenue if you want search working as a growth channel. That number comes from two figures you can check yourself: marketing budgets are running at about 8% of revenue, and companies treating search as a real growth channel put somewhere between 10 and 15% of the marketing budget into it. One eighth of 8% lands at about 1% of revenue. The math breaks down above roughly $100 million in revenue, so treat it as a floor and a starting point for the budget conversation, not a formula.
Should I hire an agency, go in-house, or use freelancers?
It depends on your revenue tier and on whether you already have someone senior to direct the work. At around $1 million in revenue, AI carries most of the function and a few hours of a freelancer catches what it gets wrong. At $10 million, one in-house hire consumes the whole budget once you load in benefits and tools, so a senior freelancer or a limited agency retainer buys more judgment. At $20 million, go in-house if daily control and institutional knowledge matter most, and go agency if you need a bench across technical SEO, content, link building, and answer engine optimization (AEO). At $100 million and up, run both.
Chapters
This Week in Search: Alphabet's Earnings Make Cloud the Cash Cow
Information Agents, AI Mode Integrations, and the SerpAPI Ruling
The Question: Should You Run Search In-House on AI?
Budgets Are Down, the AI Line Item Is Up
Why Search Should Keep Its Funding at All
What AI Does Well, and Why No Tool Is Your Advantage
Two Cost Assumptions Worth Challenging
Resourcing Search: The 1% Rule at Four Revenue Tiers
The Case to Make to Your CFO
Three Takeaways
Resources mentioned
Take this further
Don't miss an episode
New episodes every Tuesday. Get them in your podcast app of choice, or watch on YouTube.
Full transcript
Transcript lightly edited from Riverside's AI-generated draft. Any errors are ours.
Hey, welcome back to The Search Signal. I'm Michael Transon. I'm the founder and CEO of a search marketing agency called Victorious. And The Search Signal is how we help marketers understand and take action on what is happening in the world of search marketing. So as a team that works on, you know, hundreds of websites annually, we bring what we learn from running search campaigns for some of the biggest and some of the fastest
growing brands today to this podcast for you to apply to your own work. So before we get into today's topic, I do want to spend a little bit of time bringing you up to speed on is happening this week in the world of search marketing. So let's start with the big one, which is Alphabet reported its earnings this week. Search revenue grew about 17% to $63.3 billion, which sounds you know very healthy on its own, but
that growth was slower than last quarter. So search lost about three points of its share of the company's overall revenue. Network advertising, which is the ads that Google places on other publishers' websites, also declined one percent, which lines up with a lot of the traffic losses that publishers have been reporting on pretty much all year. surprisingly, the business line with the most
or maybe unsurprising was Google Cloud. That was up in an astonishing 82% to nearly $25 billion. So since Google's earliest days, search ads have been the cash cow of the business. So Google is very careful about changing the search results page. Cloud is now big enough and it's also growing fast enough to carry the company. And so that's going to free up Google to experiment, in my opinion, pretty aggressively
with new search surfaces and not having to worry about protecting every dollar of you know their search ad revenue on a quarterly basis. So I just want you to keep that in mind as you hear the rest of today's and this week's news because everything else we're going to be talking about and covering is an example of Google doing exactly that. And then also one more thing from that earnings call Google CEO said that Google is sending "billions" of clicks to websites
every week. And I want to point out that that is a static number and it does not tell us anything about whether those clicks are growing or whether they are shrinking. And Google did not share any trend data. So when a company like Google hands us a big number but doesn't give us a comparison point, the comparison is usually the part that wouldn't look good. So this is something that we need to be thinking about.
Moving on, Google's VP of search, Liz Reid, introduced something called information agents at Google IO. These are background agents, and they keep watching the web for a user and they send you an answer, typically unprompted, when something relevant shows up, rather than waiting for you to just run a search yourself. So Search Engine Journal tied that announcement to a patent
that Google filed that describes how this would work. And the way it would work is when the web can't answer a question yet, Google stores that question, keeps checking it as new pages get indexed on the web. And then once a source shows up that it's, you know, quality systems consider credible enough to answer it, it then pushes that answer to a person as a notification. It launches
first, for the paid AI Pro and the Ultra Tiers, before we expect it to roll out more broadly. Google also opened up AI mode this past week to connect directly with outside apps. So you're thinking Instacart, Canva, and YouTube music are the first. So for example, someone can add groceries to their cart and pull
things like design templates or build a playlist, and they don't ever have to actually leave Google search. Google said that more partners for this program are in the works. And, you know, as search absorbs the actions that used to happen on your site, um, more of those transactions start and sometimes end inside of Google. And that is traffic and conversions that used
land on properties and websites that you controlled. So there are two things that you can do about this right now. The first thing, we need to figure out whether an integration partner in your category has already been named by Google because being there early is going to matter. And then second, I would I would recommend you go into your GA4, your Google Analytics, and pull your organic search conversions
from the last 90 days, so that you have got a baseline that you can compare against later after this goes live. And if those conversions fall after these integrations roll out, we have an idea and can understand why that might be happening. The other news I want to share today is Google confirmed that the top stories carousel, and that is the row of recent news articles, is now going to go live inside of AI overviews for
specifically US mobile users. What it does is it links out publishers on developing news stories and it is going to give extra prominence to sources that a reader has marked as preferred. So this extends the push that Google started earlier this year to put fresher perspectives and source links into AI overviews. Google has not shared any engagement data yet. So we don't know whether people
click these links or they're just seeing them. So for any brand that is publishing timely or newsworthy content, this is a new placement that you will have to compete for. And it is going to be rewarding freshness and it rewards authority. So your, you know, classic keyword rankings are not going to get you into this. But if you publish that kind of content, there might be two questions you might want to ask your search team. Number one,
Are we set up in order to get placement on the topics that we own? And when we do show up there, are we getting clicks or are we just getting impressions? Because you know, a placement that is getting seen but never clicked, that is a brand awareness play. And I would say that's something that we should plan for and measure like a brand awareness play and not a traffic play. And then lastly, a federal judge in
California this week threw out one of the or threw out the core of Google's copyright case against a company called SERP API. So Google had sued SERP API. They had argued that they bypassed its protections to scrape and resell search results to AI and to rank tracking companies. So that judge ruled that scraping public search results that contained no copyrighted material doesn't break
Copyright law. So those claims are essentially gone for good. So Google does have the option to refile based off of its licensed knowledge panel images getting scraped. But why this matters for you is, you know, a lot of the rank tracking and a lot of the AI visibility tools that we are all relying on today to see where we show up across search, those run largely on this kind of scraped data. And this ruling makes it a lot harder
for Google to shut down that pipeline through copyright claims. So that is the news for this week. Let's jump into today's episode.
Okay, so if you are a marketing leader right now, your budget is probably tighter than it was a couple of years ago. And when you're being pressured to fund the same work with less money, you've got to figure out how to do more with less, right? It's natural at that point to question whether you should cut funding to search marketing now, especially because
there is an option that you have not had before, which is AI can do a real chunk of the search work. And, you know, right now it looks like a big money-saving opportunity. So you might ask yourself, like, why not pull search in-house? You know, run it with AI and save you would spend on a freelancer or with an agency. And that is the question that we're gonna explore today. So we're gonna
get into truly AI can take off of your plate and why having the same tools as everybody else doesn't actually make you more competitive. So we're going to break down a couple of cost assumptions that I would respectfully challenge before you choose to cut anything. And then we'll get to a very practical part in today's conversation, which is how do you resource your search channel today? And kind of budget do you actually need? So
by the end of the show, I want to be able to help you right size your search budget and also know how to think about spending it now that AI is a very real and meaningful part of the mix.
Okay, so marketing budgets, they have shrunk a lot over the last couple of years, and they honestly are not showing much signs of recovery. So Gartner surveys CEO spending every year, and this year's numbers just came out last month in June. And they said was in 2026, marketing budgets are sitting at about 7.8% of company revenue.
Now, four years ago, that number was closer to 9.5%. So call it roughly 18% or so lower than it was just four years back. And there was a second number in that survey that caught my attention when we were digging into it. So inside of the smaller budgets, CMOs are now putting, you know, just over 15% of their spend into AI.
And you think about that for a second. Like budgets have gone down. And then the AI line item has gone up. So from that data, I'd extrapolate that, you know, marketing leaders are leaning on AI a lot to stretch budgets that have essentially stagnated. And that is exactly what is making running search in-house with AI so tempting right now. And would not be remiss to say.
All of this is happening while the way people find and choose companies is going through one of the biggest shifts we have seen in years. So leaders are hunting for, you know, cheaper ways to keep search going at exactly the same moment search is changing its absolute fastest. So I see this personally looks like on the ground every week when I sit on some of our sales calls at Victorious. There are companies with real revenue, right? Businesses
you know, way past the startup stage. And they are coming to these calls with budgets, a fraction of they should actually be spending. And underneath, you know, all of these conversations is, in my opinion, probably the same unspoken question, which is can't we just do this ourselves with AI? That question really rests on the perception that AI has commoditized search.
Right. Anthropic ships SEO skills for Claude now. And the other platforms are honestly looking to be heading in that same direction. So the work must be cheap to bring in house, right? Well, let's dig into that because the answer, while seeming simple on the surface, is not as simple as it might seem.
Okay, before we get into the how, let's deal with the most basic question first, which is why should search keep its funding at all? So, like when a budget shrinks, you know, every line item has to earn its place in the strategy. And if I was a CFO personally, I would probably be putting search under the same, if not a even deeper scrutiny. And now keep in mind, I run a search agency. Of course, I'm going to tell you to keep funding search. It's literally my
business, right? Which is exactly why I would prefer in this conversation to instead lean on the data here so we can just rely on the evidence itself. So let's start doing that by talking about how buying itself has changed. So we'll start here. Forrester surveyed business buyers this year and found that 94% of them used AI somewhere in their buying process. 94%. That's basically everybody. So
you know, for years, the first step of a purchase ran through Google, the classic SERPs, 10 blue links, and you optimized to show up there. Now, a huge part of the customer journey happens inside of an AI tool, right? Somebody is asking ChatGPT or Claude, or they get an AI overview at the top of Google's search page. They'll typically still land on you know a vendor's website eventually, but by the time that they do
they've really already created their shortlist which they want to move forward with. So, you know, we have to think of search as the front door for our customers. And whether it is a traditional Google ranking or an AI answer that is naming your company, search is how you show up in the places your customers are looking. And right now, that front door, I would say, is under major, major reconstruction. And then
because it's under major reconstruction, if you choose to defund search right now, you're going dark at the exact moment your buyers are forming new habits about where to look. And then once those habits start to settle around your competitors, you're going to have to fight very hard to get your way back in.
So, generally speaking, that's the case for funding the search channel at all. Now, let's also talk about AI. And let's talk about AI as the tool your team can use to do search work because I also want to be very fair about AI does well in the world of search. And it's a lot. And you know, my honest estimate, I've run this agency for 14 years, is that AI can handle most of I would call the production work, right? This is the high volume and repeatable stuff. So
think of things like data pulls and first drafts of things, keyword research, legwork, reporting, right? The work that a I would say a junior analyst or a freelancer, right, used to spend their time doing. AI is very, very good at that layer and it is only going to keep getting better. You should be using it there. Now, if Claude's SEO skills work for
your use case, right? or ChatGPT or Gemini or whatever your team runs on, and you can verify the output, then you should use it. I don't think that adopting AI is an optional decision. The real question though is how do you decide to use it? And AI being useful and AI being an advantage are two very different things.
So let's talk about how we turn it and make it into an actual advantage.
Okay, so here's the question I would ask any marketing leader who is about to rebuild their search program using totally AI. Who else in your competitive space can buy and build exactly you just bought and build? The answer is most likely everybody, right? Every competitor you have can sign up for the same tools, can sign up for the same models, and the same SEO skills this afternoon.
For the same price or better that you paid. In some parts of a business, that is totally fine, right? If you're building software, for example, a tool that makes your engineers faster is worth it, even if your competitors have the same tool, because the product either works or it doesn't, right? Marketing, specifically search marketing, is a different situation. In search, there is a limited number of spots.
The rankings, the citations, the mentions inside of an AI answer. And the whole point is to outperform your competitors. So whether you do that depends a lot on a lot of different factors, but there is no finish line. It is a perpetual race to the top. Now imagine everybody in the category adopts the same AI tooling that you do. Has anybody gained an advantage? No. That work might get done faster.
But the set of answers everybody is competing for or keywords, those haven't changed. What happens is the baseline has risen, but nobody has moved up in relation to anybody else. We saw a version of this in our Q2 research where we walked through, we did this in the last episode. We had seen that 96% of the brands we studied were being described accurately by AI.
Essentially saying being known turned out to be table stakes. And yet, nine out of ten of those same brands never got mentioned in a single buyer answer. So, point being is being known was the baseline. And it didn't earn anybody an actual mention in AI answers. So in search, the advantage lies above everybody else is doing. And so no tool can give you an edge. So
The edge has to come from something else. It has to come from the people that are using the tool. And I would call this the human expertise layer, right? And I mean specialists who have seen and run enough campaigns to know what to target, to skip. They know really causes a problem. And also, you know, when the data isn't giving you a full and complete picture. That layer.
That human layer costs now more than ever. And it's precisely because it is the only part of a search program your competitors cannot buy off of the shelf.
So let's step back because we have seen this sort of thing before. Like there have been several big waves of technology that people said it was going to replace expertise, right? The fact is, the right tools have always raised the floor on performance. But, you know, spreadsheets, for example, didn't suddenly turn everybody into a great finance leader, right? Google Analytics, another example, that handed every marketer
the same dashboards, but it did not make everybody a great marketing strategist. What these tools did do was up-level everybody at once, all at the same time. But the differentiating factor was people did with them. But honestly, also, you know, AI deserves, in my opinion, even more skepticism than previous technology shifts because of how these systems work, right?
We did a whole episode on this a while ago. We talked about the history of LLMs and a language model at its core is just a prediction machine. It's like it is exactly as confident when it is wrong as when it is right. And an experienced SEO strategist, a search strategist, is the exact opposite of that, right? The real value of experience is knowing you don't know, right? When to double check, when to
look at a recommendation that just doesn't feel right, right? AI gives you an answer, but an expert gives you a judgment call that they are personally on the hook for, right? You ask an AI how to approach your SEO, and it is going to come back with a big synthesis of the best practices that it has read, whether that's from like training data or from search results that it does on a web search, right? And here's the thing: general best practices, they aren't wrong, right? But are they?
They are averages, right? They work across thousands and thousands of unrelated businesses and websites. Your business is not an average. You have a very specific competitive landscape, specific margins, a specific sales motion, and buyers who ask specific questions. Somebody has to be able to translate the general advice into decisions for your specific situation, right? Which practices apply to you?
Which don't apply to you, and which tactics could ultimately, if you do them, set you way back, right? That translation takes discernment. And discernment comes from having put in the reps across a lot in a lot of different situations. And you know, the analogy, I keep coming back to this, is like a hammer or a chisel, right? You hand me, you hand Michael a chisel, and I'm gonna, you know, knock some chunks off a block of marble for you, right?
Hand the same chisel to Michelangelo and you get the statue of David, right? Nothing about the chisel in those two scenarios changed. The differentiator was never the tool. It was Michael versus Michelangelo. It was the expertise.
Okay, let's now get to the budget side of this because whether you decide to use AI or how much you decide to use it, that is ultimately going to be a financial decision, right? And when companies decide to hand most of their search work to AI, the plan tends to rest on two very big assumptions. Assumption one is that AI frees up your team's time.
And you can do the same work with fewer people. And it sounds totally, completely reasonable, but there is very strong evidence that says actually otherwise. So HBR, Harvard Business Review, they published a field study on this exact thing in February 2026. And those researchers spent eight months inside of a US tech company of
about 200 people, and they watched how the work with AI changed in practice. And they found is that AI doesn't reduce work, it actually intensifies it. Individual tasks did get faster, of course, but the time that those tasks saved filled right back up. And people took on additional projects, they worked more hours.
and absorbed work that would have been previously justified with another hire. And a big piece of where that time went was adding a layer of discernment to AI generated output. So, like the engineers in their study spent way more of their day reviewing and correcting AI generated work and finishing the code that the AI had left half done, right? So the tasks moved faster.
But the burden of the expert oversight grew right along with them. And people felt more productive, but that doesn't mean that they were less busy. So if quality matters to you, and I'll say it should matter to you, right? The headcount savings that you're penciling in on the next year's budget might not ultimately pan out. And there was a second assumption that's also baked into this, which is that AI will stay as cheap as it is forever, as it is right now. And if maybe we'll even get cheaper.
And I'm here to tell you, I don't think that will happen. And the reason for this is it's cheap because the companies behind it right now are operating at enormous, enormous losses, billions and billions of dollars a year to get the world using their tools. That cannot go forever. And if you've been paying attention, the prices are already climbing on some of the newer AI models. So if your plan today assumes
today's AI pricing is going to apply three years from now. I think you might be setting yourself up for a very expensive surprise. And now it's totally taking a step back. It's absolutely understandable. And I would say it is expected that you should be testing the limits of AI right now, right? Find the corners of the room, right? That is the right instinct. And the companies that are doing it thoughtfully are going to, in my opinion, come out ahead eventually.
But on the other side, some companies are going to be cutting too deep as well. And when they need to rebuild the human expertise we've been talking about that they had decided to outsource to AI, they're going to be rehiring in a market where every other company that made similar cuts is chasing the same people. And you might very well be finding yourself in a bidding war for talent. We've already seen this play out because
the AI experiment actually started with content, right? We had a couple of years of AI-generated copy that just flooded all of the channels. And you saw memes and they're still out there, just about total AI slop, right? But this year, the Wall Street Journal is reporting that companies like Google, Microsoft, Notion, even Anthropic, the company that makes Claude.
They are building dedicated storytelling and narrative teams. There's job postings for storytellers that have been surging all through last year. And they are also still accelerating even now in 2026. So the earliest and most aggressive adopters of AI content are now also paying a premium to stick that human expertise back in the loop.
Okay, so with all of that on the table, let's pivot and let's get real practical and let's talk about ways that you can resource your search channel and how AI tools can figure into that mix. So there have traditionally been three ways to staff search. You can build an in house team, you can hire an agency,
or you can manage freelancers yourself. It's natural to want to add AI to that list as you know option number four. But AI really isn't a fourth option sitting alongside those other three. The way I look at it is it is a layer that runs underneath all of them. Your agency is using AI. An in-house team uses AI. A good freelancer uses AI, right? Whichever way you
AI will most likely be doing a lot of the production work, the data pulls the first drafts, all the repeatable tasks. So the real decision is which one of these three human models you want to build on top of it, right? AI changes the calculus differently for each of the three, right? It might reduce what you need from freelancers and junior in-house roles the most because
those roles were mostly executing on production. The agency math is a little bit more complicated because you cannot and you don't want to pay an agency to just complete tasks. You pay an agency to be responsible for a result. AI can complete a task. It can't take responsibility for your search performance. So, whatever mix you do choose, you need to keep an expert inside of the process.
So let's break down the budget considerations behind these decisions. And I'll give you a rule of thumb and I'll also walk through the math itself so you can apply this to your own situation. So start with the Gartner number we talked about earlier. So marketing budgets right now are running at about 8% of company revenue. Then ask what share of the marketing budget goes to search. So for companies that are treating search as a real growth channel.
That is typically somewhere between 10 and 15% of the total marketing budget. Just call it one eighth, right? And one eighth of eight percent lands you about one percent of revenue. So that's the rule of thumb. If you want search, and I'm talking about traditional search, AI search across all of the platforms, ChatGPT, Google, Claude and everything else. If you want it to be an effective growth channel,
fund it at about 1% of your revenue. And again, this is a benchmark. So you can back into this from the numbers. And the goal is to help make some options here concrete. So let's look at that budget funds at different revenue levels.
Okay, we'll start at $1 million in revenue. You make a million bucks a year. 1% gives you about $800 to $1,000 a month for search. And I'll be straight with you, at that budget, an agency, in-house hire, they're both out of reach. So the decision here mostly is gonna be made for you. At this level, AI can probably fulfill most of your search function. And let me tell you, I think that's fine. Right. first drafts, keyword research, the technical basics. Let AI do all of that.
And then spend whatever's left, even if it's couple hundred bucks a month, on a freelancer whose job it is to maybe set the direction and maybe also catch AI got wrong before it ends up going live. But the mistake I see at this budget range is skipping that human expertise layer completely. So remember we said about AI being exactly as confident when it's wrong as when it's right, right? You skip the expert review, and there is nobody that is standing between that confident wrong answer
and your search performance, right? AI repeats a mistake at the same speed it does everything else and across all of the pages that it touches. And that is how a small business could end up with a ton of scaling errors that could take way more money and way more time to roll back.
Okay, at about $10 million in revenue, that benchmark is going to give you about $8,000 a month. Now, that is typically not enough to hire in-house talent. So once you load in things like benefits and tools, you know, one in-house search hire is going to consume probably the entire budget. And you would be capped at the ceiling of that one person's expertise. And it is exactly at the time where AI is
already capable of doing most of they can do, but with no senior person for them to learn from and nobody to catch what they miss. So what $8,000 a month funds well is really one of two things. So either AI on the production layer with the rest of the budget going to senior judgment, which would be like a senior freelancer who is deciding what to target and what
technical fixes are ultimately gonna be worth doing? And their job is to make sense of the results are telling you or a limited agency partnership. So both give you what this budget should buy, which is experienced judgment directing the production.
Okay, at around $20 million in annual revenue, we'll call it about $15,000 to $17,000 a month for budget. You have a real choice because this budget is either going to fund one of two models. So model one is a senior in-house generalist with AI underneath them and a freelancer for overflow. And then model two is a mid-tier agency retainer. So I would say go in-house if
daily control, institutional knowledge, those matter most to you. And this would be so that you have, you know, someone in your stand-ups every morning and deepening the knowledge about your business year after year. Go the agency route if you don't have someone senior to direct the work, because you know, for the price of one good generalist, an agency is going to give you a team of specialists, and they're spanning strategy, technical SEO, content, link building, AEO.
no, to be honest, no single person can execute very well across all of those disciplines all at the same time. And then there's also a second thing that you get with an agency that I think is undervalued, which is something I would call pattern exposure. So an in-house person is you know very steeped into your business and into your website for years, and that does have a lot of real value. But the specialists at an agency work across hundreds of websites over their tenure. So
when an algorithm update hits or a new AI search surface shows up out of nowhere, they have seen things like it before and they've seen it across a whole portfolio of clients, and you benefit from that experience. So you get someone who stays at the cutting edge of what is changing because their daily work honestly just demands it from them. So the trap at this size is
in my opinion, attempting to split the difference, right? You hire a mid-level generalist and then you stretch them across the technical work, content marketing, link building, and AI search. And what ends up happening is you're under-resourced in all of those disciplines all at the same time. and you know, while we are comparing agencies to in-house builds, one other thing also applies at all of these levels, right? The commitment profile is a lot
So an in-house hire is it's a long term commitment and it also comes with a lot of overhead. You have to recruit, you've got to people manage, and you've got to deal with unwinding it and exiting a person if it doesn't work out. An agency you can bring in without any long term commitment, and you can also typically part ways on say 90 days' notice if it's not working.
Right. So when budgets and if your budget is uncertain, that flexibility can tend to be pretty valuable.
At a hundred million dollars or more in revenue, this puts your budget at $80,000 a month that you can put to search and you can stop choosing options. And if you want, and I recommend, you can run both.
Your in-house team can own strategy, can own brand voice, and all that cross-functional work that really only employees can do. And then an agency can cover the specialized depth that even a good in-house team just won't be able to keep a pace with. These would be like technical SEO edge cases, complex AEO work, whichever surfaces are shifting fast and fastest on that quarter. So AI still runs underneath the whole thing on both sides.
Now the failure mode at this level is you get a large in-house team with no outside specialist input. And an internal team, it can gradually, honestly, just fall behind the market because it's working in an insular environment and nobody notices until a competitor starts to outperform you. So I would say step back. And we look across all four of these scenarios, and there is a very clear through line through all of them, which is
AI is present in every single one of them and it's doing a lot of the production work. But AI is never the whole answer, right? What changes as budget grows is how much of the human expertise layer you can afford and what form it takes, right? A couple of hours, for example, on a freelancer for their judgment or an agency team of specialists or an in-house hire, or you might pick all of the above.
So everything that we just talked about funds the human expertise layer. And that layer requires budget. And at most companies, that's gonna mean somebody above you is gonna have to sign off, your CFO or your CEO. So I can try to prepare you for that conversation. If I were you, I would be going into budget conversations for the next year with these four talking points.
The point one is going to be the number itself. So I would be asking to fund search at about 1% of revenue. Now I would say if your business is much larger, right? If you're doing well over 100 million, if you are in the billion dollar plus range, this does break down. But if you are below that, we can really focus on this 1% number.
And you can show the same math that we walked through today. So roughly 8% generally of revenue for businesses goes to marketing today. And search would take up about an eighth of that. So that framing is important because once the math is on the table, 1% reads is like the minimum for treating search as a real, viable, long-term focused growth channel for the business. Point two is
the compounding argument. And in my experience, it is the one that CFOs are likely to respond to because it's going to frame search as an investment rather than a cost. So paid media, for example, that's a cost, right? It stops delivering the moment that you stop paying it. Organic search behaves differently, very differently. The rankings, the citations, the third-party presence that you're funding.
You know, this quarter is still going to be pulling in buyers a year from now, whether or not your investment itself stays the same or it changes. And I would also, you know, be careful about this because you know, a sharp CFO is going to flip it on you. So if the asset keeps working, you know, after the spend stops, why not just stop spending? Right? The answer is that a cut freezes the asset, right? What you've been building is going to keep delivering for a while, but without the ongoing investment,
the lead that you have is going to erode because you're going to have competitors that are going to keep building, right? And the search landscape is going to keep shifting underneath it. So I would tell your CFO plainly: this budget maintains an asset that is still growing in value. Now, point three, I would also highly recommend to be much more precise about what the money buys, which is in search today,
judgment, right? Your CFO is probably gonna ask very reasonably, why can't AI just do all of this now? And the answer is everything that we've already talked about, right? The same AI that everybody can buy lifts every company in the category all at once. But what lifts you above your competitors, right? What wins you a bigger share of the voice, that is the human expertise layer that is directing the AI. That is the line item that is under pressure right now.
But it's also the one that creates the competitive advantage. And the fourth point, I would say, is timing. So I would say the cheapest ground you will ever gain is ground your competitors are walking away from. So if companies in your category are defunding the expertise layer to chase AI savings, and the Gartner numbers are suggesting that plenty of them are, then holding your funding steady buys you position at a
discount, right? Positions and search are going to compound. And whoever is holding that ground when budgets recover are going to be very expensive to displace. And if I was a CFO, I would know what side of that trade I would bet on.
Okay, so let's do a recap. Three things that I want you to take away from this conversation today. number one, fund search like the growth channel that it is, right? About one percent of revenue is the general benchmark. And you can back into that number yourself and bring it into a budget meeting. point number
this is important. The same AI, we've talked about this all day today. The same AI everyone has raises the game for everybody all at the same time. It is not your advantage. The advantage comes from the judgment layer directing it. So you need to invest in that layer. It's the human layer. It's the only part of your search program that your competitors cannot duplicate quickly, easily, and cheaply. And then the third thing.
Decide your resourcing mix and do it with a lot of intention. AI, at the end of the day, is a given at every budget level. But the decision that is going to shape your results is how you choose to fund the human judgment on top of it. A freelancer, an agency, an in-house team, or a combination.
So I hope this episode helps you walk into your next budget conversation with a real number and a real plan. And if it did, please follow or subscribe wherever you are listening or wherever you are watching. So our next episode finds you. And that is it for me. Thanks for listening to The Search Signal. I will see you next week.
