Commodity Content: Am I Already Replaceable? The SEO Self-Test

Here’s a question I ask myself before finalizing any article for this blog: Could ChatGPT write this in 90 seconds – and would the result be essentially identical? If the answer is yes, I rewrite the article. Or I omit it entirely.

That sounds strict. And it is. But this is exactly the standard Danny Sullivan set on stage at the Google Search Central Live event in Toronto on April 21, 2026 – according to JC Chouinard’s slide documentation of the event, he directly asked the audience on which side of the line their own blog stands. Commodity or Non-Commodity. Interchangeable or indispensable.

The frustrating thing: most SEO blogs in the DACH (German-speaking markets) region that I know would fail this test. Not because the texts are bad. But because they are good and yet interchangeable. This is the point almost all explanations of Commodity Content miss.

This article is my attempt to make the concept concrete for the DACH context: with a manageable self-test, with German industry examples – and with an honest look at my own content production here on seo-kreativ.de. You can find the event overview from Toronto, to which I refer here, in my Toronto follow-up report.

In May 2025, Google explicitly wrote in the Search Central Blog: „Focus on making unique, non-commodity content” – and emphasized that unique, valuable content remains relevant for both traditional search and AI experiences. Search Engine Roundtable, among others, reported on this. This is not a ranking formula, but a recommendation – which, however, in combination with Danny Sullivan’s Toronto slides from April 2026, yields a clear strategic statement.

The precise classification: Commodity Content is not an officially confirmed ranking term – but an interpretive model derived from Google’s recommendations for helpful content. Google’s guide to helpful content has long warned against highly automated, summarized, and search-engine-driven content without original informational value. The recommendation for „unique, non-commodity content” from May 2025 is a particularly clear formulation of this stance – but not the only one, and not to be understood as a definitive definition. Crucially: You cannot copyright facts. A text that contains correct information, but no unique perspective, no unique data, no unique decision – is Commodity. Period.

Shaun Anderson from Hobo puts it well: This is essentially Helpful Content 2.0. What Google previously vaguely described as „people-first” gains a sharper edge with the Commodity-term. The benchmark is no longer „is this helpful?”, but „would this be irrevocably lost if this page disappeared tomorrow?”

This is the point where I regularly hit a wall in my consulting work. I show a client their blog. The texts are tidy. The structure is sound. H2, H3, bullet points, internal linking – all there. And yet: interchangeable.

Why? Because the author did everything right – just with the wrong source material. They read the top 5 results for their keyword and created a new article from them. Better formatted, perhaps. A little more detailed. But in terms of content: a summary of what already exists. Mark Williams-Cook essentially summarized this in the community discussion after Toronto: Content that looks like the top 5 results, only slightly rephrased, is Commodity – no matter how well it is written.

The ALM-Corp team added another important nuance that I find crucial for DACH practice: In many companies, the people with the most valuable knowledge are too busy to write content. The tax advisor who handles truly fascinating borderline cases has no time for the company blog. The master craftsman who knows dozens of difficult installation situations delegates the writing to someone who doesn’t know the cases. The result is inevitably generic.

Two niches where this particularly strikes me – and which I know from my own observations:

iGaming: Anyone clicking through German-language casino or slot reviews sees the same thing a dozen times: Bonus-Übersicht, RTP-Angabe, Screenshot von der Lobby, Bewertungssterne. Not a single piece of proof that the author actually used the provider. No screenshot of a real deposit or withdrawal. No information on how long the withdrawal actually took. No personal assessment of whether support answered a specific question. This is Commodity in its purest form – and simultaneously one of the most competitive content niches on the German web. Google has recognized this: Affiliate pages without real added value are increasingly being sorted out – and iGaming reviews without user experience are the prime example. More on the background of this niche: Black Hat SEO im Glücksspiel-SERP. The only way out: genuine proof. A screenshot of the withdrawal confirmation with a timestamp is worth more than 1,500 words of generic review text.

Template-based Content: Aside from iGaming, I encounter a pattern in audits that I now call „Who-What-How Syndrome”: every article follows the same skeleton – section 1 explains what the topic is, section 2 why it’s important, section 3 how to implement it, section 4 some tool recommendations, conclusion. The template is not wrong. But when a hundred articles in the same niche have exactly this structure, the structure itself has become Commodity – regardless of the content. Readers notice this, even if they can’t explicitly name it: The text is skimmed, not read. Whether Google evaluates structural similarity as a quality signal is not proven – but Google’s guidelines explicitly warn against content that primarily summarizes rather than provides original informational value. I described how AI exacerbates this problem in my article on the AI Content Trap.

This is where it gets technical – and I must say upfront: what follows is based on Shaun Anderson’s (Hobo) analysis of the Google API leak and has not been officially confirmed by Google. I consider the interpretation plausible and well-founded. But I don’t want to turn inference into fact.

In May 2024, parts of Google’s internal Content Warehouse API documentation became public. Among the approximately 14,000 documented attributes, one named contentEffort was found. Analysts – primarily Shaun Anderson (Hobo) – interpret it as an LLM-based effort estimation for article pages: an algorithmic attempt to measure how much original human work, expertise, and originality is contained in a piece of content. This is an analytical hypothesis, not a Google-confirmed specification.

What this means: From my work in technical SEO audits, I’ve observed for a while that simply more text and more keywords alone are no longer enough. The contentEffort attribute, if it plays the described role, would be a possible technical explanation for this. Google itself says there is no preferred word count – the focus is on quality and relevance, not length. I explain how Google specifically evaluates AI texts in my article How does Google evaluate AI texts?

Cyrus Shephard, who further elaborated on the topic at Ahrefs Evolve 2025, puts it this way: Effort is original information, useful information, own images, own research – and it’s all of it together. Word count alone says nothing. I elaborated on how the Google leak links user signals and content quality in my Google Leak Analysis.

I use this test for my own blog and for client projects. The questions sound simple. They are not.

Question 1: Could ChatGPT write this article in 90 seconds – and would the result be essentially identical?
Not „similar”. Identical in information content. If yes: You have Commodity Content. The test is brutal, but reliable. You can even do it directly: prompt ChatGPT with your main keyword and compare.

Question 2: Does the article contain at least one data point that only you could have collected?
This could be your own measurement, a screenshot from your Search Console, an observation from a real client project, an anonymized case study. Something that comes from your practice and is not already reflected online. If you find nothing: Commodity. In the iGaming sector, for example – one of the most commodity-dense niches on the German web – a single genuine screenshot of a withdrawal confirmation with a timestamp would already suffice. Not because it’s technically impressive, but because it’s the only proof that the author actually used the provider. No one else can produce this screenshot.

Question 3: Would an experienced industry expert know something they didn’t know before reading it?
Not a beginner – a professional. This is the toughest filter. If the article is only useful for beginners because it explains basics that an expert already knows: that is by definition interchangeable basic knowledge.

Question 4: Does the article contain an opinion or decision that you personally stand by?
Not „experts recommend…” or „generally applies…”. A clear position: „I would do X in this situation because…” If the article consistently weighs both sides and reaches no conclusion, it is a diplomatic Non-Commitment – and thus interchangeable.

Question 5: If this page went offline tomorrow – would anyone miss it?
Or would there immediately be ten identical alternatives? This is not rhetoric. It is the core test. Non-Commodity Content is what people bookmark, share, and link to because it isn’t stated anywhere else in the same way.

The following table shows my own examples from industries I am familiar with from my project work. The Commodity headlines have one thing in common: you could swap them out, and no one would notice the difference.

The pattern: The Commodity headlines could come from a hundred different websites. The Non-Commodity headlines bear a signature. They tell of a concrete situation, a real decision, a real result – and that can only be written by someone who has experienced it themselves.

The iGaming example illustrates this particularly clearly: if twenty pages review the same provider with the same table structure and not a single one shows genuine proof of payment, then the lack of evidence is the actual Commodity characteristic. Non-Commodity here doesn’t necessarily mean an extraordinary experience – but simply proof that one has taken place at all.

The same applies to the Who-What-How structure: it’s no coincidence that template-based content is rarely linked and rarely shared. The structure itself signals to the reader – and Google – that someone has filled a format, not documented knowledge.

I approach this in three stages when I sit with a content team:

Stage 1: Prioritization. Take your 20 highest-traffic pages and perform the self-test. Divide them into three groups: clearly Non-Commodity (do nothing), gray area (expand), clearly Commodity (decide: refactor or assign a deliberate supporting role). I explain in detail how to proceed systematically and separate dead content from valuable content in the guide to Content Audit & Pruning. Pay particular attention to two warning signs: (a) all articles follow the same structural setup – Who, What, How, Tools, Conclusion – without a single article breaking this logic; (b) all articles could also come from another provider, and no one would notice.

Stage 2: Finding the non-replicable core. For each page in the gray area: What do you know that your competition doesn’t? This could be a number from your internal statistics, a screenshot from a real project, a decision you can justify. In the iGaming context, this is often surprisingly simple: A screenshot of the actual withdrawal confirmation with a timestamp – something every reviewer could have, but hardly anyone shows – would already be a strong Non-Commodity signal. Incorporate this core, not as an appendix, but as the heart of the article.

Stage 3: Adjusting the headline. This sounds cosmetic, but it’s not. A Non-Commodity headline is a promise of real added value. „5 Tipps für besseres E-Mail-Marketing” is a promise of general knowledge. „Warum meine Open-Rate nach diesem einen Fehler von 18% auf 31% gestiegen ist” is a promise of a specific experience. Different expectations, different click-through rate, different engagement.

What I observe in my projects: Often, a single really good unique data point or a single concrete decision-making situation is enough to elevate an article out of the Commodity zone. This is less work than a completely new article – and has significantly more impact. I have extensively described how Google weights content quality in its indexing decision in the article on Crawling & Indexing. And I explain how this translates to AI visibility in the guide to SEO, AIO, GEO, and LLMO.

Kevin Indig already formulated this in 2021 – long before Danny Sullivan took the stage in Toronto: Content exists on a spectrum of interchangeability. If it’s easily replicable, you have a problem. If it’s differentiated, it’s your competitive advantage. Shaun Anderson from Hobo describes the current stage as „Helpful Content 2.0″ – an intensification of the old people-first logic with a sharper edge.

What has changed since 2021: The signals are consolidating. Google’s recommendation for „unique, non-commodity content” from May 2025, Danny Sullivan’s Toronto slides from April 2026, and the leak analyses of contentEffort – these are not random isolated events. This is consistent communication about the same quality principle, even if the technical implementation remains behind Google’s scenes.

For DACH practice, this means: The cheapest way to produce content is the most expensive way to operate content. Those who produce mass content with AI tools or outsourced text agencies without inputting their own knowledge will pay the price with the next Core Update. This is not new. But it is being enforced more strictly.

I don’t always do this perfectly myself. Even on seo-kreativ.de, there are articles that I would rate worse with the self-test today than at the time of publication. The difference is: I revise them. That’s the only thing that counts in the long run.