Let AI take the busywork.

There’s been plenty of talk over the past couple of years about what AI means for the agency model.

Depending on who you ask, we’re either being automated out of existence or at the dawn of a new age of creativity and efficiency. Personally, I think reality sits somewhere in the middle, but the opportunities far outweigh the risks. 

First of all, AI is not about making good agencies irrelevant. It is, however, going to make a lot of traditional agency work harder to justify. The critical but repetitive low-level tasks that have filled plenty of retainers for years are being stripped away. Reporting, campaign build, bid management, account hygiene or commentary are all versions of the same analysis, and much of that work needed to be improved anyway.

So, if you can accept that change is happening, the conversation becomes centred on what agencies like ours choose to become.

At Vervaunt, and across IDHL, we see that as a good conversation to have. Not because it makes our life easier, it doesn’t, but because it forces us to focus on providing value with even more determination.

We've been through this before

When I first started in paid search, a huge amount of the job was manual optimisation. We built teams to manage granular keyword structures and copy requirements; SKAGs as one example whilst assessing and applying granular time of day, day of week and location bid adjustments. Whilst assessing keyword level bids.

Today, machine bidding has long outperformed manual bidding and Google’s systems can make optimisation decisions at a scale no human team can match. The same is true across social. A few years ago, we would spend days working with a brand to define audiences, map interests, and build detailed targeting structures. Now, in many cases, platforms like Meta are just as good at finding the right customer as we ever were.

We’re clearly going through a new environment of automation. This was simply the first wave of automation becoming the new normal.

We are now moving into the second wave: campaign builds, reporting, account maintenance, performance commentary and routine analysis are all moving in that direction. Ultimately, AI and modern technology will be the drivers behind the lower layers of agency work being automated.

the lower layer of agency work is being quietly rewritten by machines. And if we are honest, it's unlikely any of that work will exist in the same way in the future. So, what replaces it?

Automation isn't the threat. Crap value is.

The answer is not “doing the same, just more of it or faster”. That is useful, but speed only matters when it improves the quality of the decision, the efficiency of the spend, or the commercial result, and that isn't guaranteed. 

The opportunity that removing low-value effort activities provides is to reinvest that time in the areas where agency expertise is harder to replicate, Judgement, strategy, creativity, commercial understanding, and joined-up thinking is where the real value of an agency has always come from. 

If your model depends on revenue from manual optimisation, you probably do have a problem. But if your value comes from understanding how growth actually happens across media, product, customer experience, data and margin, then I think you'll have a much more interesting future.

Creative becomes a bigger performance lever

A good example is creative. As platform automation takes on more of the account management load, ad build and optimisation work, creative becomes the key differentiator for performance.

We've always championed the impact of great creative but we're now spending more time working in partnership with internal brand and content teams to drill into the data and insights and understand which creative ideas are really driving results, where fatigue is setting in, what messages are moving customers, and how performance data can shape the next round of content. I am adamant that's a better use of agency time than endlessly tweaking inside an ad account.

Basically, the platforms can optimise delivery but, for the moment at least, they can't decide what a brand should say, why a customer should care, or how an idea can evolve once the campaign starts revealing its impact.

Paid media cannot stop at the click

Another positive change we're seeing is a much closer relationship between ecommerce, CRO and paid media.

Historically, there was a fairly neat handover. One team drove traffic and another worried about what happened when the user got there. That separation makes less and less sense today.

More than ever before, growth comes from understanding the entirety of the journey. That means the media strategy, the landing experience, the product detail page, the checkout, the offer, the margin, the return rate, the LTV, and much more. None of those things live neatly inside one channel so siloed working and thinking can only hold you back.

This is where Vervaunt’s role, and the wider IDHL ecosystem, becomes more valuable. Paid specialists cannot only be media buyers. Today, they need to understand the onsite experience, challenge the user journey and connect performance back to the commercial model.

Better questions beat bigger dashboards

More spend than ever is flowing through platforms like Meta, but chasing surface-level efficiency is the wrong race. A campaign can look healthy on paper and still be sending budget in the wrong direction once you factor in all the other variables. That means agencies need to get closer to the numbers behind the numbers and look beyond traditional dashboard metrics to become more like brand, marketing and business analysts.

Which products are driving higher-value customers? Which categories convert well but hurt your margin? Where are returns distorting the view of performance? Which regions offer lower acquisition costs but are weaker commercially once the full journey is understood?

AI can help us get to the answer faster, but only if we know what to ask and what the answer should be.

Tech is part of the agency product

In the future, I think every serious agency will be using their owned data combined with AI to create unique client value. In the world of big network agencies and their eight-figure tech budgets, developing proprietary tools sounds daunting, but we see it as an opportunity to build products around our specialisms.

For IDHL, one of the clearest examples is Vantage, our in-house performance intelligence platform. It's built on a semantic layer across a huge dataset of analytics accounts, allowing us to interrogate multiple datasets in ways designed around our clients and their specific needs.

Developing our own tech also means that rather than looking at a client account in isolation, we can see relevant macro patterns, shifts and behaviours across a larger and more relevant dataset and in much less time.

AI-based tools and technologies like this change the nature of the conversation. Instead of spending weeks pulling data together, we’re using the technology to find the right questions.

The point is not to build another shiny dashboard (the world has enough dashboards). The point is to give our teams better context and evidence, so they can make sharper recommendations, and our clients see the direct benefits.

Starting an agency today

Probably one of the questions my cofounder, Paul, and I get asked the most is "If we were building an agency from scratch today, what would it look like?"  
 
My answer is I would build what we're building today; a multidisciplinary growth specialist who understands how media, creative, data, commerce and profitability connect.

I would also prioritise technology from day one. The tools, workflows and systems an agency uses are no longer just internal infrastructure. They are increasingly part of the client experience. Direct platform integrations, clean data structures, useful automation and intelligent agents will only become more important to how agencies deliver consistently and at scale.

At Vervaunt, we’re a decent way down this road, including direct integrations across social channels and the use of Google agents to identify optimisation opportunities across campaigns. That includes flagging issues like overbidding, underrepresented high-performing products, and areas where the account needs human attention.

Governance isn't optional

Of course, more automation creates more responsibility. So, if agencies are going to automate more of the work, clients need to understand what's being automated, why it's being automated, and where human oversight sits.

All of that means governance can't be treated as an afterthought. We need clear standards for how AI is being used, what data is involved, what gets checked, what gets signed off, and where the limits are. Clients need to know the whos and whats, and not be left guessing whether a recommendation came from a person, a model, or a mix of both.

Across IDHL Labs, this is a big part of how we think about applied AI. Our goal isn't to automate everything we can, just because we can. The aim is to define the right challenges, understand the opportunity and the risks, and make informed decisions that benefit the agency and our clients, with the visibility every stakeholder needs. 

We think that's a more useful ambition than chasing novelty or making claims that fall apart the minute someone asks to look under the hood.

Making ourselves the most useful agency

AI will automate a fair chunk of agency work, and there's no point pretending otherwise. But the work being automated is not the work our clients value most. In many cases, it's the work that got in the way of us spending more time on strategy, creative, journeys, insight and commercial accountability.

The changes AI brings about won’t be felt equally. Every agency will use AI, so adoption won’t be a differentiator. It’ll be a question of whether they use AI to protect the old model, and do the same work faster and cheaper, or to get better at what clients care about: strategy, creative, and commercial outcomes. The automation bit will quickly become normal. The useful bit is where the real value will be.

So yes, AI is taking the busywork. Good riddance. At the end of the day, the future of agency value was never going to be found in doing more of the same. Like every industry shift we've seen, success will come from being commercially useful and responsible for the outcomes clients expect.

If that’s the agency model of the future, we’re good with that.