We've been here before.
When desktop publishing arrived, it fundamentally changed how work was produced. It made things faster, reduced reliance on certain specialist roles, and yes, people lost their jobs. The craft of paste-up, the reliance on typesetters, the separation between design and production - all of it shifted. But the industry didn't collapse. It adapted. The agencies that were already strong carried on. The ones that weren't struggled, and no amount of new software changed that.
AI sits in that same category. It changes how things get done. It doesn't remove the need for what we actually do.
And yet the industry is slipping into a familiar pattern - treating a tool as an existential threat, rather than asking the harder question: are we confident enough in our own value?
AI won't fix what wasn't working before
If an agency was struggling before AI, this won't turn things around.
You don't become better by writing a better prompt. You become better by knowing what good looks like, having the experience to recognise when something isn't there yet, and having the judgement to fix it. That comes from years of doing the work, not from access to a tool. AI can support that process, but it can't replace it.
If anything, it will highlight the gap between average and genuinely good work more quickly. When more people can produce something that looks acceptable, the bar for what stands out gets higher - not lower. A decent-looking piece of work is no longer a differentiator. The thinking behind it, and the ability to apply it effectively for a specific client in a specific context, is what separates the good from the forgettable.
The agencies clients already trust are not suddenly going to lose that position because a new tool exists. That trust has been built over time, through consistent delivery and genuine understanding of the client's business. A tool doesn't replicate that.
This has always been a talent business. Having access to the same software has never meant people produce the same level of work - and the same applies here.
What AI does is remove friction in delivery - first drafts, repetitive tasks, the admin that fills up people's days. That should create more space for thinking. And better thinking is what improves the work.
The pressure will come through pricing
This is where it will show up most directly, and it's the conversation the industry needs to be better prepared for.
Clients are under pressure too. They're being asked to do more with less, and many of them are being encouraged to adopt the same tools internally to become more efficient. It's inevitable that they will start to question agency costs in that context - what efficiencies are being made, and how that is reflected in what they're being charged. That's not a new conversation, but it will become more pointed.
The risk is straightforward. If pricing is anchored to time, and the work gets done quicker, the assumption will be that it should cost less. That's the trap - and it's one that's easy to fall into if agencies haven't been clear about where their value actually sits.
Just because something can be done faster doesn't mean it should be done cheaper. The value has never been in the time it takes. It sits in the outcome delivered, the quality of the thinking behind it, and the experience required to get there. A senior strategist who can solve a problem in an afternoon is not less valuable than one who takes a week. The opposite is true.
Strong agencies should be confident in that position. If you're working with a well-respected agency, you expect to pay more. That's true in consulting, in accountancy, in law. The fee reflects the talent, the experience, and the likelihood of achieving the result. That logic doesn't change because the tools have changed.
This might also be the moment to move away from time-and-materials pricing altogether - agreeing upfront what the outcome is and what it costs to deliver it. Value sits in the result, not the hours. Agencies that can make that shift will be in a stronger position to have the pricing conversation on their own terms.
And the stakes are higher than they might appear. To attract and retain top talent, you need to be able to pay for it. That requires charging appropriately for your work. Once fees start being reduced purely to reflect efficiency, it puts pressure on the one thing that actually drives quality: your people. If you can't afford the best, the work suffers. And when the work suffers, so does the agency.
We've been undervaluing ourselves for years
Free pitching, discounting to win work, competing on cost - none of this started with AI. These behaviours have been embedded in the industry for a long time, and they were already doing damage before any of this became a conversation.
AI just makes it easier to fall further into them.
If agencies position themselves around execution, they will be compared to anything that can execute faster and cheaper. That's a race no one wins. The moment an agency's primary offer becomes the ability to produce work quickly, it is competing with tools that will always be faster and always be cheaper. That's not a sustainable position.
The real value has always been in the thinking, the judgement, and the ability to deliver an outcome that moves the needle for the client. That hasn't changed.
This is really about confidence
AI will not replace strong agencies. It will put pressure on those that haven't been clear on where their value sits - but that was already an issue.
If you were a poor agency before AI, this won't save you. If you're a good agency, it will make you more efficient. If you're a great agency, it gives you more time to focus on what actually matters. The thinking, the ideas, the outcomes.
The industry has spent too long being apologetic about what it charges and uncertain about what it's worth. That's the problem AI is exposing - not creating. The agencies that understand their value, communicate it clearly, and hold their position on pricing are the ones that will come through this in better shape.
That's where the value has always been. It's not going anywhere. The question is whether the industry is willing to stand behind it.
