Florence versus the machine: Why agencies must dive into change

Florence versus the machine: Why agencies must dive into change
Like Florence Welch, we need to think of ourselves as and, not versus, the machine

Learn from history: resisting innovation invites obsolescence and eventual ridicule, writes Croud UK’s MD.

Florence & the Machine was one of several high-profile acts to serenade advertising execs at this year’s Cannes Lions, but perhaps the biggest debate raging — at Cannes and beyond — was whether the machine (artificial intelligence) really needed Florence (us) any more.

The Luddites, who tried to halt the industrial revolution by destroying spinning jennies, have become a metaphor for futility. Meanwhile, Blockbuster’s stubborn attachment to their business model and eventual refusal to buy Netflix has become a case study on what not to do in business. Resisting innovation invites obsolescence and eventual historical ridicule; agencies need to take heed of the lessons of the past.

Throughout history specialisation and subsequently automation have catalysed diversity through fostering a heightened focus on creative and intellectual pursuits. Yet, whether in the press or the office, we currently have a cacophony of negative noise around artificial intelligence (AI). On the one hand we have “the baptists” who have a biblical belief in our own AI-driven demise, and on the other the “bootleggers” who want to slow progress because it benefits them. Yet, I believe we must rise above the noise and seize this opportunity.

Even by a conservative estimate, AI enhances the workflow speed in areas like media buying and creative production by around 25% annually. This means that in three years we will see similar tasks taking less than half of the time that they require now. With up to 60% of people’s time freed up — we have a huge potential opportunity. As agency leaders, we have two options.

Option 1: Unpack efficiency

The simplest thing is to lower your service prices due to needing fewer human resources, cheap prices wil indeed attract some new clients. However this approach will invariably lead to a race to the bottom, negatively impacting the role of agencies in the client value chain.

It provides a technical challenge too — as AI platforms receive less input over time, their efficacy may decrease — a phenomenon known as “model collapse”. This results in AI becoming self-referential and increasingly derivative. Used incorrectly, AI creative output will be no better than stock art; plastic people sitting around a board table, engaging in a clearly fake conversation, smiling with dead eyes.

Option 2: Prioritise creating new value

This approach still heavily utilises technology and automation, but only when it enhances output quality, not merely to cut costs. I apologise for the 1980s Lethal Weapon reference, but the first step is to view AI as Danny Glover (Murtaugh) to your Mel Gibson (Riggs), or vice versa — a complementary buddy who makes what you do that much better. As an example, I wrote this article and then used ChatGPT to make some refinements, most of which I actually rejected (but it cut down my editing time significantly).

It is gains like this that you need to meticulously track across your agency. By being clear on time saved you can liberate your teams to think beyond conventional boundaries, be more proactive, learn more about their craft and how to better use AI. By doing this, I guarantee you will build ever stronger client relationships, and ascend up the value chain. To achieve this, we need to excel in selling ideas, not just time, in this industry.

A clear framework for evaluating opportunities

In a challenging environment, the luxury of attempting everything simultaneously is one that few brands and agencies can afford. I find it useful to evaluate potential advancements across three parameters, in order to select which opportunities to best pursue.

  1. Reach — How many people can the advancement connect with?
  2. Impact — How engaging is the innovation or format? How much time can be saved?
  3. Difficulty — What are the barriers to entry? How much development is required?


By setting a minimum score and only investing time once the score reaches a satisfactory level, one can drive innovation strategically.

Looking ahead, I am thrilled by the potential of mixed reality, especially with Apple’s (much pilloried) foray into this domain. Yet, their proven knack for defining categories, even when they aren’t pioneers, could uplift the industry. Remember all the doubt around the iPad, I bet some of you thought it was a bit pointless!

Not only do we have the Apple headset; at the other end of the scale, we have the Oculus Quest 3 releasing at a price point of $500, packed with advanced technology such as hand tracking and super high resolution. A few more rounds of enhancements, and we could be on the cusp of a truly disruptive innovation that ends up on the heads of millions. This will then create a whole new canvas for our industry to utilise, with the potential for limitless creativity and a whole new sector likely to emerge.

Maybe Florence was right all along. It can never be versus, it should always be and when it comes to the future of media and machines.

Emil Bielski is managing director at Croud UK

Media Jobs