How does your store show there is someone behind it? This is what Google calls effort.
Miguel Nicolás
27/10/2026
-
Google looks for effort in every piece of content and turns it into a variable that is, at the very least, good practice what about AI? what about SEO?
-
-
Google has just included effort in its official guide to helpful content, and makes it clear that churning out AI texts without even the slightest review counts as zero effort. Do not panic yet: we explain what has changed, what it means for an online store and how to review your product pages, categories and content step by step.If you have a store with hundreds of products, these days, you have more than likely already used AI to create/optimise descriptions, category text or metadata. If you ask me, I think it makes perfect sense: writing 500 product pages by hand is unfeasible for a small team (and before Artificial Intelligence, we were already using templates and other tricks). No objections here.The problem arises when that text is published exactly as it comes out. It looks too much like that of any other store selling the same thing, tells buyers nothing they could not read on the product's box and, worse still, even hallucinates and makes up critical facts.In early October 2026, Google has given a name to what separates a page like that, from one that is worthwhile. It calls it effort (effort), and the good news is that, for once, small or medium-sized stores, find it much easier to demonstrate than their big competitors.
-
What exactly has Google changed in its guidelines?
-
On 1 October, Google updated its page on how to create helpful, reliable content that's designed for people (Google Search Central). It added a section stating that the quality of the main content (main content) is one of the most important factors in assessing a page. It is a somewhat subjective concept, but it went into enough detail to cite the four variables its quality raters look at:
- Effort
- Originality
- Talent or skill
- Accuracy.
But what does effort mean to Google? It defines it as the amount of human work devoted to creating the content or the systems that generate it. I think it is fairly clear what it is talking about, but, just in case, the text adds that misusing generative AI to produce large amounts of text without supervision or manual curation amounts to little or no effort (Lily Ray).That same day, Google also revised its guide to AI-generated content. It is now much clearer that you need to manually review and check everything AI produces before publishing it, take note: this includes titles, meta descriptions, structured data and image alt text (Google Search updates).From the perspective of stores like ours, there is something particularly interesting: for Google, main content is not just the text you write on your website: it also includes customer reviews, specification tabs, tools such as calculators or search engines, and also the page's titles and headings (Relevant Audience). In other words, it sees a product page as integrated content, rather than separate elements.All this sounds relevant, and it clearly is, but let's keep things in perspective: we are talking about good practices that increase relevance, rather than potential penalties.Google has not announced any algorithm update linked to this change, or published an effort score you can measure. What it has done is move criteria that were previously in its raters' handbook into its public documentation. So it is still a clear indication of where it will focus in the near future, and it is worth taking seriously, though without panicking. -
How does effort show in an online store?
-
The effort that counts is what you can see on the page, it is intangible, but we can all sense when there is work behind an eCommerce business. It is not just about spending time on development, in fact, the hours you spent putting the catalogue together mean nothing to anyone, if the product page the user ultimately sees, is identical to the manufacturer's. What matters is whether visitors find something they could not read in another store, something that shows you have taken the trouble to think and give them something unique.On product pages it usually shows in very obvious details. Photos you have taken yourself, with the product in your hands or while using it, rather than the catalogue images forty other stores use. Dimensions checked with a tape measure, rather than copied from the technical sheet. A paragraph saying who the product isn't for, or which other item in your store it goes best with... a human touch.Categories are the other major area that demands effort, which makes sense because there is always been a push to rank them in SEO with optimisations specific to this type of content.A generic 200-word text about "the importance of choosing the right backpack" adds absolutely nothing, it is just a jumble of keywords. However, one explaining the differences between the 5 brands you sell, or what capacity you need depending on use, does.Then there is everything only you know because you are the one who talks to customers: the questions you get on WhatsApp, the most common reasons for returns, the feedback you have received in the physical store or the usage tips customers have shared. If we explained in a recent article how to make the most of your competitors' reviews, here is a good place to apply everything you discover.A small store has an advantage here. A large marketplace can't test each of its millions of products, but you might be able to do so with, for example, your 50 bestsellers.
-
-
How to review the effort in your product pages and categories step by step
-
I hope you have not already started rewriting your entire store, because that is not what this is about. It is simply about knowing which pages have the most to lose and starting with them. The best part is that we can do it without tools, with Search Console or Analytics and a spreadsheet, you have more than enough.1.- Export your most visible pages. In Search Console, go to Performance > Search results, select the last three months and open the Pages tab. Add a page filter containing the part of the URL your product pages or categories use (for example, /producto/ or /categoria/).2.- Export the list to Google Sheets using the button at the top right. Keep the top 30 or 50 by impressions. These are the ones Google is already showing people and, therefore, the ones that would be most affected if their rating dropped. If you only have a few, all the better: you will finish sooner.3.- Run the duplicate text test. You already know this is very negative, in fact, duplicate content has been subject to SEO penalties for many years. Even so it is better to clear up any doubts because it only takes a moment: copy a sentence from each product page's description, paste it into Google in quotation marks and search. If ten stores are shown with the same sentence, that text comes from the manufacturer or from an AI that told everyone exactly the same thing. Note it in a column.
4.- Score each page using a simple criterion. Add five columns and enter a 1 if it meets the criterion and a 0 if it doesn't:CriterionWhat you check
Original text
The description does not appear identically on other websites
Original image
There is at least one photo or video you've made yourself
Verified fact
Dimensions, materials or compatibility checked manually
Usage experience
Something only someone who has used or sold the product knows: who it is for, how it fits, what it goes with
Human review
A person has read and corrected the text, title and meta description
5.- Better with a formula. Do not worry, that is simple: add up the five criteria in a score column (from 0 to 5) and sort the sheet by impressions, highest to lowest. Pages with lots of impressions and a score of 0, 1 or 2 are your work list. If you want to refine it, add a column with =impressions*(5-score) and sort by it.
6.- Do not try to do everything at once. Work on 5 pages from the list each week. Start with what makes the quickest visible difference: an original photo and a couple of sentences with real information are worth more than rewriting the whole product page.
7.- Measure the results. Record the date of each change in the sheet. After 6 or 8 weeks, return to Search Console, filter by those URLs and compare clicks and impressions with the previous period. Bear in mind seasonality and that Google takes time to recrawl pages, so do not draw conclusions after a fortnight.If you are already using AI to help generate content, great, but you can only fill in the "Usage experience" column manually. This is the part Google will not find on any other website. Combining AI's ability to scale, with human judgement, is what makes the real difference, blending productivity with judgement. -
-
What should you do with content you have already created using AI?
-
Let me stress again: you DO NOT need to delete it, do not remove it... Google accepts the use of AI without any problem (it would be rather cynical of it not to), what it considers low effort is publishing without reviewing.You simply need a human filter before anything goes onto the website, and to apply it to what is already published too. In fact, you are very likely already doing it.Google highlights four elements worth reviewing one by one:
- Titles, because AI tends to repeat the same structure across every product page.
- Meta descriptions, which often promise things the product does not have.
- Structured data, where an invented price, availability or rating can end up appearing in search results.
- Alt text for images, which AI often fills with vague descriptions such as "high-quality product image".
In a store, the most expensive mistake is a false fact. An AI can claim a case is compatible with a phone model that does not exist, that a garment is organic cotton when it is not or that a watch is water-resistant to 100 metres when getting it wet will ruin it. These things generate returns and complaints, as well as reducing the quality of the page and the user experience. If a text includes dimensions, materials, compatibility or timeframes, always check them against the supplier's specifications or the product itself.The other new development concerns the byline, the famous author. Google has added a warning against inventing author profiles, for example, with AI-generated photos, fake names or qualifications that don't exist, and considers it a form of deception (Relevant Audience). If your store's blog is signed "Laura, cosmetics expert" and Laura does not exist… change it today. Use your own name or that of the person who actually writes, or simply use the store's name. And on your "About us" page, show real faces, even if it is a photo taken on your phone in the warehouse.Relevance, authority… nothing new in the conversation, these are things we have been talking about for two decades. They simply adapt to a new reality, and AI is not something negative, but rather a tool that should be used to add value through initiatives that stand out. -
Automate with AI, personalise with “effort”
-
For a couple of years it seemed that whoever published faster, more often, with longer content..., won, there was even a “minimum amount of text” as an SEO best practice. With these new guidelines, Google makes it clear that volume without review is useless to it, and that what it is looking for is the mark of someone who knows what they sell.It's a mark that is quite hard to simulate. It is in the photo you took in the warehouse, the measurement you checked with a tape measure and the question you answered on WhatsApp yesterday. It takes more work, but no store can copy it from you.So, if you like it, use AI effectively to generate 70% of the volume, but always try to make a difference in that other 30%.
-
Are you clear yet on how to show Google the effort you dedicate to your eCommerce business?