“Water, water everywhere, nor any drop to drink”: Coleridge’s Ancient Mariner is a fitting patron saint for our moment. I sit in the middle of constant flux surrounded by data – live feeds, hot takes, shaky videos and dashboards blinking like cockpit instruments in turbulence. As a Bayesian, I want to lean on priors, update beliefs and try to understand what a likely outcome is going to be.
However, when the rulebook is torn up and actors behave with impunity and whim, the priors don’t help much. The ocean is vast and most of it is salt. Marketers face a similar sea. We have never had more data, more models and more dashboards. We have large language models (LLMs) that can summarise, synthesise and spin endless content in seconds.
Yet, like the Mariner, we often find ourselves thirsty. The problem isn’t scarcity; it’s potability. The internet – the very fuel that powers today’s models – has given us abundance and pollution: bias, bots, rage-bait and hallucinations masquerading as confidence. The albatross we killed, if we’re honest, wasn’t curiosity or creativity; it was the discipline to separate signal from noise and to remember that the customer is human, not a row in a data table. Our curse isn’t LLMs per se; it’s forgetting what they are for. So, where do we find safe harbour?
Start with what hasn’t changed. The great Bill Bernbach reminded us that it is “the unchanging man” that matters. Jeff Bezos urged leaders to anchor on what won’t change. In marketing, the constants are mercifully simple: people want to save time and money; they want ease, recognition and belonging; they seek beauty, status and reassurance; they avoid effort, uncertainty and risk. We are the same creature we were a century ago, equipped with the same attention budget, the same cognitive shortcuts and the same social cues. When data becomes turbulent, these truths become the bedrock.
Artificial intelligence (AI) is transforming our industry – no doubt. Agencies and holding companies are building secure, privacy-safe AI stacks; wiring planning, media and production workflows with models; partnering across clean rooms and retailers; and upskilling talent so creators, strategists and traders have a capable co‑pilot.
In the Middle East, a digitally native, youthful population and ambitious national AI agendas mean adoption is happening at speed. Hyper-local nuance, Arabic language models and commerce integrations are moving from experiment to infrastructure. But embracing AI doesn’t absolve us from judgement. In fact, it raises the premium on judgement. The models are only as good as the questions we ask, the data we feed them and the constraints we impose. Treat AI not as autopilot but as an instrument panel. It accelerates what you already are – disciplined or distracted, customer‑centric or channel‑centric.
If you’re a marketer seeking a way through the fog, a few principles help turn the ocean into drinkable drops. Begin with enduring human truths. Jobs to be done, not just demographics. Map the anxieties, aspirations and frictions in the buying moment. This is where creative advantage lives. Build memory and reduce friction. Distinctive brand assets, fluent devices, and mental and physical availability still correlate with growth. Make it easy to recognise, choose and use.
Make first-party data your fresh water well. Consent, clarity and value exchange matter. Invest in quality over quantity; sparse, accurate signals beat oceans of sludge. Demand causality, not just correlation.
Blend experimentation such as holdouts and geo-lifts with marketing mix modelling (MMM) and platform signals to understand incrementality. Reward learning velocity, not vanity metrics.
Put guardrails on generative AI. Red-team outputs, verify facts, protect intellectual property (IP) and define brand tone. Use models to augment insight and craft, not replace them. Have fewer dashboards and more default actions. If a metric doesn’t change what you do on Monday morning, it’s probably theatre.
Yes, LLMs can write a media plan synopsis, translate a brief, localise copy for 12 markets overnight and help a strategist prototype segmentation in minutes. They can fuel scenario planning, creative variation testing and attention optimisation. They can flag outliers in campaign performance or elevate emerging cultural signals before they crest.
These are real, material gains but they don’t change why people buy. And they won’t fix a weak value proposition, muddled positioning or forgettable creatives.
The other uncomfortable truth is that much of our data is salted. Identity is fragmenting. Signals are obfuscated by privacy controls, platform policies and the sheer noise of automated content. You can’t out‑model bad measurement hygiene. This is why discipline around clean rooms, consented data partnerships and robust experimentation is non‑negotiable.
It’s also why creative excellence matters more than ever; the single largest lever on effectiveness remains the work itself. Operating in a world that defies neat priors also argues for a different managerial rhythm: shorter planning cycles and more small bets.
Bayesian thinking doesn’t go away because the world is weird; it becomes the norm. Update beliefs weekly. Build portfolios that hedge. Reward teams not for being right, but for being rigorously curious and fast to adapt.
From where I sit – in a region where conflict and innovation often coexist – the marketer’s responsibility is clarity. To make sense without oversimplifying. To use AI to compress the distance between a real human need and a useful, memorable response. To protect attention, not exploit it. And to remember that in the long run, trust is the compounding asset: earned slowly, lost quickly and rarely rebuilt by an algorithm.
The Mariner’s lesson wasn’t to fear the sea. It was to respect it – and to atone by telling the tale. Our tale, as an industry, is that technology keeps changing and people mostly don’t. Let AI make you faster, broader and bolder. Let data make you humbler and more precise. But let the human stay in charge.
Amid data, data everywhere, the drops to drink are the ones that help a person feel seen, make a decision with ease and come back again. Keep your hands on that wheel, and you won’t just survive the voyage – you’ll chart a course others can trust.
By Chris Solomi, Chief Digital Officer, Publicis Groupe Middle East
