Jack of All Trades, Master of AI

For my entire career, I’ve been told the same thing in slightly different words: pick a lane.

Be the SEO guy. Be the PPC guy. Be the web guy. Niche down. Specialize. Build a moat around one skill and charge for that one skill forever. Every agency I ever looked at was built that way too: an SEO team, a paid team, a dev team, a design team, a content team, all sitting in their own Slack channels, with hand-offs taped together by project managers and a Monday.com board no one updates.

That model is dead. Or at least, it’s dying, and I don’t think enough people in this industry are saying it out loud.

The biggest holding companies in the world are already saying it with their org charts. Omnicom just bought IPG for $13.5 billion, then announced 4,000 layoffs and sunset more than 20 of their own agency brands, including names like DDB, FCB, and MullenLowe. WPP dropped from 108,044 employees to 98,655 in a single year while pivoting to what its new CEO calls a “simpler, lower-cost, AI-enabled business”.

This isn’t really a story about agencies getting smaller. It’s a story about the shape of an agency changing. WPP’s CFO said it plainly: “AI integration requires different skills and different talent”. The roles being cut are not the roles that will get hired back. The duplicate account managers, the layered creative directors, the siloed media planners. Those seats are gone. The AI engineers, the data leads, the product people building Omni and WPP Open as central operating systems. Those seats are filling up. Headcount will probably climb back. The org chart underneath it won’t look anything like the one we have today.

The old model is being rewired from the top down. I think it’s also being undercut from the bottom up.

The old setup never actually worked that well

I run an agency. I know how this is supposed to go. The client comes in, you scope the engagement, you assign an SEO lead, you assign a paid lead, maybe a web person, maybe a designer. They all do their part. They report up to a project manager. The project manager reports to the client.

What actually happens is this. The SEO person writes a recommendation that the dev never builds. The paid person sends traffic to a landing page the designer made that doesn’t have a clear offer. The content writer ships an article that’s good but ignores the keyword the SEO team picked. Nobody fully owns the outcome because everyone only owns their slice.

The “jack of all trades, master of none” was an insult for a reason. The assumption was that depth always beat breadth, and breadth meant you were mediocre at everything. But that math only holds if depth is hard to access. If depth is sitting on the other side of a Google search, a Loom video, a YouTube tutorial, or a person who knows their lane, you have to be a specialist to deliver good work.

Depth is no longer hard to access. That’s the whole thing.

I’ve always been a generalist. It used to be a weakness.

I run Google Ads. I do SEO. I look at GA4. I review websites and tell clients what to fix. I write copy. I help build landing pages. I think about offer, positioning, conversion, and the actual business behind the marketing. I always have, because I was never willing to do just one of those things. It felt like only seeing a quarter of the picture.

For years, that was a problem. Specialists could go deeper than me on any single channel. The PPC nerd knew bid strategies cold. The SEO nerd had every algorithm update memorized. I could hold my own across all of it, but on any given day there was someone who knew their lane better than I did.

What changed is that I can now go as deep as I want, on demand. If I need to understand a niche bidding strategy, I have it explained, debated, and stress-tested in 15 minutes. If I need to audit a site’s technical SEO, there’s a skill that walks me through it like a senior consultant would. If I need to write 30 RSA variations for a campaign, I can produce them, evaluate them against best practices, and ship them before the specialist agency has even scoped the work.

The generalist used to be capped at “pretty good across the board.” Now the generalist has a ceiling that keeps moving up.

AI didn’t make me a specialist. It made specialists optional.

I want to be careful about how I say this, because I don’t think AI replaced expertise. It just made expertise something I can rent by the hour instead of something I have to hire full-time.

At Ruskin, I use Claude every day. I have skills for keyword research, search term review, ad writing, monthly reporting, SEO check-ins, and account audits. Each one of those is essentially a senior practitioner I can pull off the bench when I need them. I’ve open-sourced the ones I use in my own work at marketing-boost-skills on GitHub if you want to see what they look like. I don’t have to keep an SEO lead on payroll to do an SEO audit. I don’t have to brief a content team to draft a blog. I don’t have to schedule a meeting with a designer to review a landing page.

I do the thinking. I make the calls. The AI executes, double-checks, and fills in the parts I’d otherwise need to outsource. Because I’m the one connecting the channels (paid traffic to landing page to email to SEO), nothing falls through the cracks the way it always did in the old model.

The output is better, not worse. Not because the work is more impressive in isolation, but because it’s coherent. One person can hold the whole client in their head and ship across every channel without losing the thread.

This isn’t just my experience. There are now solo operators running 12 retainers at $750 to $1,000 each, using under $500 a month in AI tools to do the work that used to require a team. Growth Method profiled a “Gen Marketer” who replaced three specialist contractors for a single client, increased output 40%, and cut cost in half. Digiday found AI is now used at 99% of agencies surveyed in their 2026 research. The shift is already in the numbers.

What this means for agencies

I think we’re going to see two kinds of agencies thrive in the next few years, and a whole middle layer get squeezed out.

The first kind is the deep specialist shop. Three or four people who do one thing better than anyone else on earth. Enterprise clients with enterprise budgets will still pay for that.

The second kind is the small, AI-native generalist shop. One operator, or a few of them, each running a full book of clients across every channel, with AI doing the work that used to require a team of 10. Faster, cheaper, more accountable, and (this is the part the old guard keeps missing) better at the actual marketing, because the strategy isn’t getting filtered through five hand-offs before it ships.

The dying middle is the 30-person agency with a head of SEO, a head of paid, a head of content, a creative director, an account team, and an ops team. The cost structure doesn’t make sense anymore. The output isn’t differentiated. The client doesn’t feel a single owner. AI eats that org chart from the bottom up.

You can see the squeeze happening already. The big holding companies are merging and slashing because scale is the only way to afford the AI investment, and they need to amortize it across thousands of clients. The small operators are growing because their cost base is a Claude subscription. The middle is paying for the overhead of being big without the leverage of being big.

Pick the lane that doesn’t get narrower

If I were starting today, I wouldn’t specialize. I’d get good at thinking, get good at writing, get good at the business behind the marketing, and use AI to be a specialist on any channel the moment a client needs me to be.

Jack of all trades, master of none. Better than a master of one. Especially when “all trades” now includes a tireless team of experts that costs me a subscription instead of a salary.

The cliché was right. The world just had to catch up to it.