Your Content Strategy Is a Landfill — And AI Just Made It Worse

Between 2023 and 2025, the volume of published content online increased by an order of magnitude. Most of it was AI-generated. Most of it was indistinguishable from everything else on the same topic. And most of it achieved precisely nothing.

This wasn't a failure of AI. It was a failure of judgement. Organisations treated AI as a way to do more of what was already not working. If your content wasn't generating leads at fifty pieces per month, producing five hundred pieces per month doesn't fix the problem. It multiplies it.

The hard truth is that most corporate content has never worked. It was always vanity publishing dressed up as marketing. The metrics were reach and impressions — which measure distribution, not impact. AI just made it possible to produce vanity content without the inconvenience of actually having to write it.

Content intelligence starts with honest measurement, and honest measurement is what most organisations are desperate to avoid.

I've seen content teams report 'engagement' based on page views. Page views measure whether someone clicked, not whether they cared. I've seen teams celebrate social shares without tracking whether a single share led to a conversation, let alone a commercial outcome. I've seen quarterly content reports that would make a Soviet statistician blush — all green, all up, all meaningless.

Real content intelligence means answering uncomfortable questions. Did this piece generate enquiries? Did it influence a deal? Did it earn citations from credible sources? Did anyone with purchasing authority actually read it? For most organisations, the honest answer to all four questions is 'we don't know' — which functionally means 'no.'

Every content intelligence programme I've been involved with reaches the same conclusion: organisations that produce less content with higher quality consistently outperform organisations that produce more content at lower quality.

This is true for traditional search. It's dramatically more true for AI surfaces. AI models don't reward volume. They reward authority. A brand that publishes one definitive, well-structured, frequently cited piece on a topic will outperform a brand that publishes fifty shallow variations. The model doesn't count your pages. It assesses your credibility.

But try telling a marketing director to produce less. The entire incentive structure of content marketing is built around volume. Agencies bill by output. Internal teams justify headcount by throughput. The metrics reward production. Cutting volume feels like cutting effort — and nobody wants to go to the board and explain that the strategy is to do less.

The organisations that get past this are the ones where someone senior enough has the nerve to say: most of what we publish is a waste of money, and we're going to stop.

Here's an exercise that most content teams resist: take everything you published in the last twelve months and answer three questions for each piece. What was it supposed to achieve? Did it achieve it? Is it still performing?

In every audit I've seen, 70-80% of the content library is dead weight — producing no measurable value. Not underperforming. Producing nothing. Zero leads, zero citations, zero influence on any commercial outcome. It exists because someone had a content calendar to fill.

The 20-30% that works tends to share three characteristics: it takes a genuine position that not everyone agrees with, it's based on first-hand experience rather than desk research, and it's structured in a way that AI models can parse and cite. In other words, it's opinionated, original, and technically sound.

The implication is brutal: most content teams should fire their content calendar, audit what actually works, and rebuild around the 20% that isn't landfill.