''' Most marketing metrics are a comfortable lie. They are designed to make marketing managers look busy and justify agency retainers, not to measure what actually matters: profit.
For decades, the industry has been trapped in a cycle of flawed measurement. We’ve stumbled from an era of no data to an era of too much bad data. Now, we are entering the third era, the Prompt Era. This isn’t another incremental shift. This is a complete reset, driven by agentic AI, that will finally force marketers to measure what counts. The rest is noise.
I’m just old enough to remember the tail end of the analogue marketing world. The era of print, billboards, and television ads. The era of ‘Print and Pray’. The primary form of measurement was inference, bordering on guesswork. A full-page ad in The Sunday Times or a prime-time slot during Coronation Street was hideously expensive, and its success was measured in the vaguest possible terms.
Marketing directors would talk about ‘Opportunities To See’ (OTS) and projected readership figures. They would run brand tracking surveys, asking consumers months later if they recalled seeing an ad. It was a world of soft, lagging indicators. Did the ad campaign cause the 1.5% uplift in quarterly sales, or was it the sunny weather, a competitor’s supply chain issue, or a million other confounding variables? Nobody knew.
This era is best summarised by the famous quote attributed to department store magnate John Wanamaker: “Half the money I spend on advertising is wasted; the trouble is I don’t know which half.” It was a knowing shrug, an acceptance of the fundamental opacity of marketing effectiveness. The best you could do was buy the most prestigious slot you could afford and hope for the best. Accountability was minimal because genuine proof was impossible.
The internet arrived and promised to fix all this. The shift from print to pixel was meant to usher in an age of perfect accountability. Every click, every impression, every session, every conversion could be tracked. We were finally going to find Wanamaker’s wasted half.
What happened instead was that we swapped one problem (no data) for another, arguably worse, problem: a tsunami of meaningless data. The industry became obsessed with intermediate metrics, creating a whole new dictionary of jargon to justify its existence. Click-Through Rates (CTR), Cost Per Click (CPC), Engagement Rate, Bounce Rate, Video Completion Rate (VCR). The list is endless. These are vanity metrics.
They are easy to measure, easy to report on, and easy to manipulate. An agency can always find a way to get you a lower CPC by targeting irrelevant audiences. A social media manager can always boost ‘engagement’ by posting a picture of a cat. Do these activities drive profit? Almost never. But they look fantastic on a monthly report.
Take a major UK online retailer like ASOS. For years, the focus of its social media strategy, like many others, would have been on metrics like likes, shares, and follower growth. It’s a comforting dashboard. But what if the cost to acquire a customer (CAC) through that channel is steadily climbing? What if the lifetime value (LTV) of customers acquired via Instagram is a fraction of those from organic search? The focus on the vanity metric of ‘engagement’ obscures the terminal decline in channel profitability. This isn’t a theoretical problem. Research from analytics firms has suggested UK businesses waste over £1 billion annually on digital advertising that fails to deliver a return, often because it’s optimised for the wrong, intermediate metrics.
This led to the rise of the attribution model, a supposedly scientific method for carving up credit for a conversion between dozens of different touchpoints. First-touch, last-touch, linear, time-decay. It’s a multi-billion pound industry dedicated to creating complex spreadsheets that provide the illusion of certainty. In reality, it’s a sophisticated way to lie to yourself. The modern customer journey is far too chaotic to be neatly mapped. Trying to assign a precise value to a single tweet, blog post, or display ad viewed three weeks before a purchase is a fool’s errand.
This era, the Pixel Era, didn’t solve marketing’s measurement problem. It just made the reports longer and the lies more complex.
The Third Era: The Inevitable Agentic Prompt