← Journal

Guides

How to write people-first content, using Google's helpful content guidance

By Thomas Lovas Lokoy · Published 16 October 2026 · 6 min read

A person reading a book by the water
Photo: Reading Book by Negative Space · CC0

"Helpful content" is Google's shorthand for pages written to serve a reader rather than to catch search traffic. Google calls the same idea people-first content, and its guidance page on the subject opens by saying its automated ranking systems are designed to prioritise "helpful, reliable information that's created to benefit people, and not content that's created to manipulate search engine rankings." This guide turns that page, and the Google posts around it, into a working method: what the guidance actually says, what has changed since 2022, and the editing passes that make a draft more useful to the person who will read it.

What Google means by helpful content

The guidance is short and mostly made of questions. Google suggests creators evaluate their own work against them, and, beyond that, consider asking people they trust but who are unaffiliated with the site for an honest assessment. The questions fall into a few groups: content and quality, expertise, page experience, and a pair of contrasting lists headed "people-first" and "search engine-first".

The core of the people-first list is five questions. Would an existing or intended audience find the content useful if they came to you directly? Does it show first-hand expertise and depth of knowledge? Does the site have a primary purpose or focus? Will a reader leave feeling they have learned enough to achieve their goal? Will they leave feeling they have had a satisfying experience?

None of this rules out SEO. Google's page says SEO "can be a helpful activity when it is applied to people-first content, rather than search engine-first content."

From a single update to core ranking

The idea arrived as a named update. Google's Search Central blog announced the "helpful content update" on 18 August 2022, describing it as part of an effort to ensure people see "more original, helpful content written by people, for people". That post introduced a site-wide signal: sites judged to have relatively high amounts of unhelpful content could see all their content perform less well, assuming better content existed elsewhere. It also said that "content created primarily for search engine traffic is strongly correlated with content that searchers find unsatisfying."

That arrangement has since changed. Google's guide to its ranking systems now lists the helpful content system under retired systems and says that "In March 2024, it evolved and became part of our core ranking systems". The blog post announcing the March 2024 core update put it this way: "There's no longer one signal or system used to do this". If you read older advice about a single helpful content classifier, treat it as history.

The search engine-first warning signs

Google lists habits it treats as warning signs. Answering yes to some or all of them, the page says, means you should reevaluate how you are creating content. In practical terms:

  1. Writing mainly to attract search visits. The test is whether you would publish the piece if search did not exist.
  2. Covering many unrelated topics in the hope something ranks. A site with a focus is easier to trust and easier to keep accurate.
  3. Using extensive automation to produce content on many topics.
  4. Summarising others without adding much value. A rewrite of the top three results is still only those three results.
  5. Chasing trends your audience would not care about.
  6. Leaving readers needing to search again for better information elsewhere.
  7. Writing to a word count. Google's page asks whether you are writing to a particular length because you have heard it has a preferred word count, and answers: "(No, we don't.)"
  8. Entering a niche without real expertise, mainly for traffic.
  9. Promising an answer that does not exist, such as a release date nobody has confirmed.
  10. Changing dates to look fresh when the content has not substantially changed, or adding and removing content in bulk to make the site seem fresh. Google's parenthetical on the latter is "(No, it won't)".

Who, How and Why

Google offers three questions to keep content on course.

Who created it. Make authorship obvious to visitors, with bylines where a reader would expect one, leading to information about the author. Google is equally clear that fabricated profiles, such as AI-generated headshots, made-up names or false credentials, are deception. Our E-E-A-T explainer covers authorship and trust in more depth.

How it was created. Readers benefit from knowing how a piece was produced. For product reviews, Google suggests stating how many products were tested, the results and how the tests were run, with evidence such as photographs. Where automation or AI substantially generated the content, it suggests making that clear and explaining why it was useful, where readers would reasonably expect to be told.

Why it was created. Google calls this "perhaps the most important question to answer about your content." The answer should be that it helps people who would find it useful if they came to your site directly. If the answer is that you mainly want search visits, that is not what Google's systems seek to reward, and using automation to manipulate rankings is a violation of its spam policies.

What raters are told to look for in the main content

The same page summarises four qualities that search quality raters are trained to evaluate in a page's main content:

  • Effort: how much human work went into the content or the systems behind it. Using generative AI to produce large amounts of text without manual oversight is given as an example of little to no effort, and the page adds: "Attribution or giving credit to other sources doesn't replace the need for original effort."
  • Originality: information or perspective not already available elsewhere.
  • Talent or skill: clear writing, well-produced media, working tools. Not every topic needs specialised expertise.
  • Accuracy: informational pages should be factually accurate, and on topics that could significantly affect people's lives, highly accurate and consistent with expert consensus.

Raters do not set rankings. Google says rater data is not used directly in its ranking algorithms; it is feedback on whether the systems are working.

A people-first editing routine

Use the guidance as a series of passes on a draft rather than a checklist to tick at the end.

  1. Name the reader and the goal. Write one sentence: who is reading, and what they should be able to do afterwards. Every section must serve that sentence. If you cannot write it, you do not have a brief yet; our briefing guide shows how to build one.
  2. Find what you add. List what this page offers that the current results do not: a test you ran, a worked example, a clearer explanation, a correction. If the list is empty, the page is a summary.
  3. Answer first. Put the direct answer near the top, then the detail and the conditions that change it.
  4. Cut to length, not to a target. Delete every sentence that only introduces the next one or restates a point. The page is as long as the answer needs.
  5. Check every claim. Link a primary source for each fact, figure or quotation, or remove it. Google's guidance on generative AI warns that models "don't retrieve facts, but predict a likely sequence of words based on their training data", and says the review should also cover titles, meta descriptions, structured data and image alt text. The fact-checking routine is a fixed order for doing this.
  6. Sign it and explain it. Add a byline, and a short note on how the piece was made if readers would reasonably ask.
  7. Run the questions. Go through Google's content, expertise and people-first questions against the finished draft, then give it to someone unaffiliated with the site and ask whether it answered their question.

Scale is not the problem; purpose is

Google's scaled content abuse policy, announced in March 2024, targets many pages generated "for the primary purpose of manipulating Search rankings and not helping users", and applies whether the content is produced by automation, by people, or by both. The line is not the tool. It is whether each page exists because someone needed it.

Sources