Sam Wood.
We spent a decade teaching people to go deep in one channel, and are now surprised that we cannot find enough people who understand how a business works. The skills gap in agencies is one the industry built for itself. The more interesting question is whether AI can stop us repeating it with the next generation.
Being narrow is completely fine when you are starting out. Early on, depth in a channel is exactly what you want from someone. Learn the craft, get good at it, become genuinely useful. The problem shows up later. As people get more senior, the expectation changes. You are meant to understand how the pieces fit together: how business objectives relate to marketing objectives, what corporate strategy actually looks like, how to read a P&L well enough to understand the real cost of the advertising you are running and the return it is delivering for a client. That is the job at a senior level, and fewer people seem to be able to do it.
As an industry, particularly on the digital side, this is largely our own doing. Think about what the last decade has looked like. You need to learn Google Ads. Or Meta ads, content, SEO, or programmatic. We pushed people to be deep in one thing and we rewarded them for it. Now we are surprised to have a workforce of highly specialised operators rather than generalists who understand how that campaign ladders up to the client's business.
This is not me being diminutive about specialists. It reflects what we prioritised, and those people did exactly what the industry asked of them. That is on us, not on them. But those same people are becoming more senior, and they want bigger titles and bigger paycheques as they go - as you’d expect. Meanwhile the owners hiring them need someone who can connect things: talk to clients about strategy, spot upsell opportunities, read that a client's competitor is entering a new market and say what it means and what the client should do about it. That is a genuinely hard person to find right now, which is a tough position to be in.
If you are chasing for that senior role right now, my honest read is that it will be hard, and you will probably have to do some training or make a trade-off on something you wanted from the role. That is the market as it stands. The more interesting question is what happens with the current crop of juniors coming through, and how AI might start to matter.
The legal industry is seeing the negative side of junior AI usage at the moment. A growing body of comment about big US firms describes junior lawyers effectively becoming AI wrappers, running work through legal AI tools and passing on the output without engaging with the substance. The concern is not that the tools are inaccurate. It is that they collapse judgement into answers too early in the learning curve, producing lawyers who are just supervising technology rather than mastering the work themselves. That is about the worst version of this you can imagine for juniors.
But marketing is a simpler and lower-stakes use case than legal work, and the same tools can point slightly sunnier. Instead of using AI to do the thinking and skip the learning, we can get people spending less time buried in the platform doing the doing, and more time asking how the pieces fit together. If setting up a campaign now comes with a follow-up question to an AI tool, "I have built this campaign, these objectives, these channels, this spend, help me understand what that does to overall customer acquisition cost", or "how does that show up in a P&L", then we might start bridging some of the gaps I talked about earlier. That is a junior building the exact business context we keep complaining our seniors lack.
We currently have a real problem where mid and senior people are missing the connective, commercial skills, and we will have to train them up regardless. The question is how we stop it getting worse with the next generation. A lot of the answer is prompting, in both senses. Some of it is literal AI prompting. But a lot of it is the human kind: getting mids and juniors working in-platform to ask one more, two more, three more questions about the impact of their work on the business as a whole. As my two-year-old would put it, why, why, whyyyyy.
Plenty of agencies have told me they are not hiring juniors anymore, not because of AI but because of the time and energy it takes to train them. I have some sympathy for that, but it is an unfortunate place to land. Flip it, and treat juniors as an investment again while using AI to help build the broader business skills our industry is short on, and you start to solve a couple of problems at once. Your team is already using these tools whether or not you have sanctioned it, so the real question is not whether to allow it. It is whether you use it to broaden your people or just to speed up their output.
I will declare the conflict of interest: former agency CEO and GM, now a generalist consultant to agencies, so of course I would argue for generalists. But breadth has been sidelined for the better part of a decade, and we are now seeing the cost in how hard it has become to find mid and senior people with the range agencies need. There is still plenty of specialist work to be done, and a lot of it is not going anywhere. The opportunity is to reset how we build people around that work: ask the extra why, and get teams using AI to understand how the pieces fit rather than only to finish the task faster. If I were prioritising one thing in an agency right now, that would be near the top of the list.
Sam Wood is the founder of OTTESU and advises independent agencies across Australia and New Zealand on strategy, operations and governance. He has spent more than a decade running and growing agencies.
Being narrow is completely fine when you are starting out. Early on, depth in a channel is exactly what you want from someone. Learn the craft, get good at it, become genuinely useful. The problem shows up later. As people get more senior, the expectation changes. You are meant to understand how the pieces fit together: how business objectives relate to marketing objectives, what corporate strategy actually looks like, how to read a P&L well enough to understand the real cost of the advertising you are running and the return it is delivering for a client. That is the job at a senior level, and fewer people seem to be able to do it.
As an industry, particularly on the digital side, this is largely our own doing. Think about what the last decade has looked like. You need to learn Google Ads. Or Meta ads, content, SEO, or programmatic. We pushed people to be deep in one thing and we rewarded them for it. Now we are surprised to have a workforce of highly specialised operators rather than generalists who understand how that campaign ladders up to the client's business.
This is not me being diminutive about specialists. It reflects what we prioritised, and those people did exactly what the industry asked of them. That is on us, not on them. But those same people are becoming more senior, and they want bigger titles and bigger paycheques as they go - as you’d expect. Meanwhile the owners hiring them need someone who can connect things: talk to clients about strategy, spot upsell opportunities, read that a client's competitor is entering a new market and say what it means and what the client should do about it. That is a genuinely hard person to find right now, which is a tough position to be in.
If you are chasing for that senior role right now, my honest read is that it will be hard, and you will probably have to do some training or make a trade-off on something you wanted from the role. That is the market as it stands. The more interesting question is what happens with the current crop of juniors coming through, and how AI might start to matter.
The legal industry is seeing the negative side of junior AI usage at the moment. A growing body of comment about big US firms describes junior lawyers effectively becoming AI wrappers, running work through legal AI tools and passing on the output without engaging with the substance. The concern is not that the tools are inaccurate. It is that they collapse judgement into answers too early in the learning curve, producing lawyers who are just supervising technology rather than mastering the work themselves. That is about the worst version of this you can imagine for juniors.
But marketing is a simpler and lower-stakes use case than legal work, and the same tools can point slightly sunnier. Instead of using AI to do the thinking and skip the learning, we can get people spending less time buried in the platform doing the doing, and more time asking how the pieces fit together. If setting up a campaign now comes with a follow-up question to an AI tool, "I have built this campaign, these objectives, these channels, this spend, help me understand what that does to overall customer acquisition cost", or "how does that show up in a P&L", then we might start bridging some of the gaps I talked about earlier. That is a junior building the exact business context we keep complaining our seniors lack.
We currently have a real problem where mid and senior people are missing the connective, commercial skills, and we will have to train them up regardless. The question is how we stop it getting worse with the next generation. A lot of the answer is prompting, in both senses. Some of it is literal AI prompting. But a lot of it is the human kind: getting mids and juniors working in-platform to ask one more, two more, three more questions about the impact of their work on the business as a whole. As my two-year-old would put it, why, why, whyyyyy.
Plenty of agencies have told me they are not hiring juniors anymore, not because of AI but because of the time and energy it takes to train them. I have some sympathy for that, but it is an unfortunate place to land. Flip it, and treat juniors as an investment again while using AI to help build the broader business skills our industry is short on, and you start to solve a couple of problems at once. Your team is already using these tools whether or not you have sanctioned it, so the real question is not whether to allow it. It is whether you use it to broaden your people or just to speed up their output.
I will declare the conflict of interest: former agency CEO and GM, now a generalist consultant to agencies, so of course I would argue for generalists. But breadth has been sidelined for the better part of a decade, and we are now seeing the cost in how hard it has become to find mid and senior people with the range agencies need. There is still plenty of specialist work to be done, and a lot of it is not going anywhere. The opportunity is to reset how we build people around that work: ask the extra why, and get teams using AI to understand how the pieces fit rather than only to finish the task faster. If I were prioritising one thing in an agency right now, that would be near the top of the list.
Sam Wood is the founder of OTTESU and advises independent agencies across Australia and New Zealand on strategy, operations and governance. He has spent more than a decade running and growing agencies.