Content · Editorial Strategy · Attention Economy
Your Editorial Plan Produces Noise
How to build an effective content strategy in 2026 — when 74% of new online content is already AI-generated and the attention market selects costly signals, not volumes.
A B2B consulting firm has been publishing three pieces of content a week for two years. Organic traffic exists. The CRM has not a single lead attributable to the content channel. The marketing director asks the agency what isn’t working. The answer is always the same: more content, higher frequency, better-optimised titles. The logic is hydraulic: increase inflow, increase output.
The problem is that the attention market does not work like a hydraulic system. It works like a signal market in which credibility is selected through the cost of production. In 2026, with generative AI having driven the marginal cost of any piece of content to fractions of a cent, the distinction between content strategy and background noise production has become the difference between building positioning and dissipating budget.
What follows is a structural analysis of the problem — with the data, the cognitive mechanism, and the operational implications that most editorial plans systematically ignore.
The company publishing into the void
What is an editorial plan and what is it for?
An editorial plan is the operational document that schedules which content an organisation will publish, on which channels, at what frequency, and with which declared objectives. It defines formats, topics, responsibilities, and timelines. It exists to give coherence to production, prevent publication gaps, and coordinate writers, approvers, and distributors. It is a production tool, not a strategy: it answers “what do we publish and when”, not “why do we publish it and for whom”.
Mark Schaefer formalised this in 2014 with the concept of Content Shock: when the supply of content exceeds the human capacity to consume it, the market seizes. Free content already carries the lowest possible price, so to get it read companies must pay in distribution — and that cost grows with every new piece of content published by any other actor in the market. In 2025, with AI production driving marginal cost to fractions of a cent, that curve reached a slope Schaefer had not anticipated even in his worst-case scenarios.
This is not hyperbole. Graphite data across 65,000 URLs published between 2020 and 2025 shows that the share of predominantly AI-generated articles rose from roughly 10% in 2022 to over 50% by the end of 2024. Ahrefs, analysing nearly one million new web pages published in April 2025, found that 74.2% contain detectable AI content. The web is filling with zero-cost signals.
The problem is not that this content is necessarily poor quality. It is that it carries no trace of the cognitive cost sustained to produce it. And the attention market, like any signal market, selects for credibility. Credibility is a function of cost.
80% of corporate content loses money. 20% generates returns above 500%. There is no middle ground. The distribution is bimodal, not normal.
| Indicator | Figure | Source | Implication |
|---|---|---|---|
| Web content with generative AI (Apr. 2025) | 74.2% | Ahrefs, 900k pages | 3 in 4 new pages are indistinguishable in signal |
| Online articles written primarily by AI | 52% | Graphite / Surfer, 2025 | Parity reached for the first time Nov. 2024 |
| Corporate content that loses money | 80% | Analysis of 15k companies | Bimodal distribution, not gaussian |
| Average content marketing ROI (when it works) | $7.65/$1 | Ranktracker 2025 | Nearly 2x the average ROI of traditional marketing |
| B2B marketers not measuring ROI accurately | 64% | ProperExpression | Producing without knowing what they produce |
| Brand awareness via B2B content marketing | 87% | CMI 2025 | Awareness ≠ leads. The gap is structural. |
| Marketers using generative AI tools | 89% | CMI 2025 | Everyone uses the same tools. Zero signal differentiation. |
Sources: Content Marketing Institute 2025, Ahrefs AI Content Prevalence Study Apr. 2025, Graphite / Surfer URL Study 2025, Ranktracker Content Marketing ROI 2025
Why isn’t your content being read, even when it’s well made?
In 2001, Thomas Davenport and John Beck published “The Attention Economy”. The thesis: the scarce resource in the information economy is not information, which has become abundant and immediately available, but the human attention required to process it. Attention as “focused mental engagement on a particular piece of information” carries real physiological costs that cannot be scaled.
Herbert Simon had intuited this thirty years earlier, in 1971: “A wealth of information creates a poverty of attention.” Not as a metaphor. As a structural implication of any system in which the supply of information exceeds the processing capacity of its recipients. In that context, the human brain develops automatic selection mechanisms: it recognises familiar patterns, classifies them as “low additional information”, and ignores them before the conscious system has time to evaluate them.
This has a direct implication for content strategy that most editorial plans never consider: format familiarity is already, in itself, a signal of irrelevance. An article that resembles other articles in its field does not compete with them on content. It is filtered out earlier, at the level of pattern recognition. It never enters the attentional field.
That estimate did not account for Slack, WhatsApp threads, automated AI summaries,
push notifications, algorithmic social feeds. In 2026 the real number is no longer calculable.
The cognitive processing capacity has remained unchanged.
The peacock and the LinkedIn post
In 1975, Israeli biologist Amotz Zahavi proposed the handicap principle. The peacock’s tail is an elaborate, costly signal to maintain: it reduces mobility, increases visibility to predators, consumes energy. And yet it evolved. Why?
Zahavi’s answer is that the cost is the point. In a system where everyone has an interest in appearing stronger and more fit than they are, false signals proliferate. The only way a signal becomes credible over time is if it is genuinely costly to produce: weaker individuals cannot afford to sustain it. The signal is honest because it is costly.
Applied to content marketing: a piece of content produced without real cognitive cost — generated by AI on a template, published in twenty minutes, optimised for keywords without a point of view — does not emit a credible signal because it carries no trace of the cost sustained to produce it. Any competitor can replicate it identically. The attention market recognises this and filters it out. Not by conscious choice, but by evolutionary mechanism.
Costly signal (credible):
An original research paper with primary data collected by the company over 12 months
→ Visible production cost. Not replicable at low cost by competitors.
→ The market perceives quality through the trace of cost.
Zero-cost signal (not credible):
An SEO-optimised article on “5 industry trends in 2026” generated by AI
→ Marginal cost: fractions of a cent. Any competitor can replicate it in minutes.
→ The market receives no information about underlying quality because anyone can emit that signal.
Structural implication:
In a market saturated with zero-cost signals, the cognitive system automatically calibrates to ignore everything that carries no trace of the cost sustained to produce it.
| Content type | Production cost | AI replicability | Credible signal |
|---|---|---|---|
| Original research with primary data | High (time, access, analysis) | None | Yes |
| Case analysis with direct source access | Medium-high (relationship, exclusive context) | Very low | Yes |
| Signed, contestable opinion | Medium (real reputational risk) | Low | Partially |
| SEO-optimised thematic article | Low (template + AI) | Very high | No |
| Listicle / “N things to know about X” | Minimal | Universal | No |
| AI-generated post on industry trends | Near zero | Infinite | No |
How to build an effective content strategy: start with the channel, not the format
Neil Postman wrote in 1985 that every medium has its own epistemology: it is not neutral with respect to the content it carries, but shapes it, selecting which forms of thought it can host and which it cannot. Television could not carry complex thought because its internal logic rewarded simplification. The point was not moralistic. It was structural.
The same logic applies to every digital platform. LinkedIn is not a neutral container: it has an algorithm that rewards certain behaviours, a user culture that has developed its own implicit norms, a specific attention threshold. Content optimised for LinkedIn is structurally different from content optimised for an industry newsletter. Which is structurally different from content optimised for a podcast. Not in terms of tone or length — in terms of the type of credibility the format can host.
A content strategy that ignores the specific ecology of each channel produces abstract content: designed for a generic audience on a generic platform. The result is that it is ignored by specific audiences on specific platforms, who have learned to recognise and discard generic content with the same speed as the immune system neutralises a known pathogen.
Choosing where to emit the signal is more decisive than optimising the signal. Most content strategies work in the wrong order.
What is the difference between an editorial plan and a content strategy?
The most common product that content marketing agencies sell to companies is not a strategy. It is a production plan with a “strategy” label. The difference is substantial: a production plan answers “what do we publish and when”. A strategy answers “which cognitive space do we want to occupy in people’s minds, and through which system of credible signals”.
What is the difference between an editorial plan and a content calendar?
An editorial plan and a content calendar are distinct tools that are often confused or used as synonyms. The editorial plan is the strategic-operational document: it defines objectives, target audience, core themes, tone of voice, formats, and channels. It is the “why and how” of content production. The content calendar is the temporal planning tool derived from the plan: it lists publication dates, planned titles, authors, progress status, and distribution channels. It is the “when and who”. A company can have a calendar without a plan — and many do — but the result is a sequence of dates without a direction. A plan without a calendar remains intention without execution.
What should an effective editorial plan contain?
An effective editorial plan contains at least six elements: a specific audience definition (not “Italian SMEs” but “the marketing manager of a manufacturing SME with a team of fewer than five people”); measurable objectives for each format and channel; core themes with a horizon of at least 12 months; the mapping of formats onto channels with the logic of adaptation between one channel and another; evaluation metrics distinct for awareness, engagement, and conversion; and a frequency calibrated to the actual resources available, not to industry benchmarks. The most common mistake is building a plan that assumes resources that do not exist — and that collapses after six weeks from operational overload.
The content calendar solves the operational problem of regularity, which is a real problem. But it takes the place of the strategic question instead of being its implementation tool. The result is that many companies have a regular cadence without positioning, a presence without identity, a voice without a point of view that could not belong to any of their competitors.
| Dimension | Editorial Plan | Content Strategy |
|---|---|---|
| Central question | What do we publish? | Which cognitive space do we occupy? |
| Unit of measurement | Content produced | Perceived positioning |
| Time horizon | Month / quarter | 12–36 months |
| Critical variable | Frequency | Signal credibility |
| Main risk | Irregularity | Indistinguishability |
| Measurable output | Traffic, impressions | Durable mental association |
| Where it produces value | Short-term presence | Competitive positioning over time |
The durable mental association is the metric no one measures and the only one that matters in the long run. When a decision-maker thinks of a supplier for a specific category of problems, the process is not rational in the strict sense: it is the result of an accumulation of signals that over time has built an association between that category and that name. Building that association requires consistency, specificity, and signal credibility. Not frequency.
How often should you publish content in 2026?
Optimal frequency is not a constant: it depends on the content type, the channel, and — above all — the actual availability of cognitive resources to produce something that is worth the attention cost of the person receiving it. As an operational rule: it is better to publish once a week with a piece that carries a perspective that is hard to replicate, than five times with content any AI tool can generate in thirty seconds. Frequency is an amplifier of the signal, not a substitute for it. A company that does not yet have a clear signal does not need to publish more — it needs to understand what signal it wants to emit.
How do you measure content marketing ROI?
Content marketing ROI is structurally harder to measure than other channels because its primary value — cognitive positioning — accumulates over time and is not directly attributable to a single session or a single piece of content. Vanity metrics (views, followers, raw engagement rate) do not measure ROI: they measure visibility. The metrics that approximate real ROI are: the number of leads who spontaneously cite specific content during the purchase process; the change in share of voice on strategic topics over time; the conversion rate from repeat readers compared to first-access traffic; and the length of the sales cycle for prospects who consumed content before commercial contact. The 64% of B2B marketers who do not measure ROI correctly do so not for lack of tools, but because the metrics they measure are not connected to the business objectives the content is supposed to produce.
A company that publishes twice a month on a very specific topic, with original data and a recognisable point of view, builds more effective positioning than one that publishes every day on everything relevant in its sector. Because the second case, for the reader’s cognitive system, is noise. The first is a signal.
Why does content marketing not work for most companies?
The second law of thermodynamics establishes that isolated systems spontaneously evolve towards states of increasing entropy: from concentrated energy to dispersed energy, from structured disorder to uniform disorder. Boltzmann formalises this with S = k ln(W): entropy is proportional to the number of microscopic states compatible with a given macroscopic state. The more states are possible, the more disordered the system.
A content system without strategy is a high-entropy system. Every piece of content published is a unit of energy that enters the media system and disperses uniformly instead of concentrating at a specific point. The heat generated is real, the resources consumed are real, but the gradient produced is zero. Without gradient there is no useful work.
The uncomfortable implication is that entropy is the default condition. You do not arrive at entropy by mistake. You arrive there by doing nothing specific to prevent it. A company that publishes content without a positioning strategy is not failing at a strategy. It simply does not have one. It is producing entropy by definition.
An editorial plan without positioning is a system
that produces waste heat instead of useful work.
The only way to reduce local entropy is to import energy
from outside: original thought, research, exclusive point of view.
Building order, reducing local entropy, requires a real energy investment. In a saturated media system that investment is called specificity: the content that reduces its own entropy is the content that positions itself on something precise enough to exclude everything else. Specificity is not an aesthetic choice. It is the physical mechanism by which gradient is created in a high-entropy system.
Does content marketing work for small businesses and SMEs?
Yes, but with a logical inversion from what agencies typically sell. For an SME with limited resources, content marketing works to the extent that it concentrates — rather than distributes — its cognitive budget. A large company can afford to emit signals on ten channels and hope one takes hold. An SME cannot. Its only competitive option is depth on a specific channel and a specific topic: becoming the recognisable reference for that problem, in that niche, for that audience. The advantage is that specificity does not cost more than generality — it costs less, because it requires less production and more thought. Thought is the only resource SMEs have in identical proportion to large companies.
In 2026, with 89% of marketers using generative AI to produce content, and 74% of new web pages already identifiable as AI-generated, the differentiating gradient is called real cognitive cost: the visible trace of thought, research, access, and point of view that no tool can replicate at zero cost. It is the only form of signal the attention market has not yet learned to ignore automatically.
// Final ripple
The question a company should ask its team is not “how much content do we produce this week”. It is “which mental association do we want to build, in which specific audience, through which signal costly enough to be credible over time”. The answer is not in a calendar. It is in a choice about who you are willing to be, and what you are willing to stop pretending to be.
Content that positions itself on nothing is not a failed strategy. It had already reached its thermodynamic equilibrium before it was published. The media system, like any physical system, does not preserve waste heat. It dissipates it.
Behind the Algorithm
The question this article leaves open
If real cognitive cost is the only differentiating signal, are companies that cannot afford original research structurally excluded from the attention market?
Zahavi’s thesis does not exclude subjects with limited resources: it obliges them to be more selective. Not more content at lower cost, but less content at genuinely higher cost. Frequency is a signal multiplier only if the signal exists. Otherwise it is a noise multiplier.








