YT YouTube2Text

YouTube2Text → Transcripts → Google’s Secret “Content Effort” Score: Here’s What It Looks For

Google’s Secret “Content Effort” Score: Here’s What It Looks For — Transcript

by Edward Sturm · 2,425 words · 382 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00If you've ever wondered if Google looks
0:02at how much effort you put into your
0:04content, well, this is a write up from
0:06Cyrus Shepard, really well known in SEO.
0:09He's been on the podcast before. He
0:10wrote this article, content effort, the
0:13Google ranking feature nobody talks
0:15about. I read it, very interesting. So,
0:17I'm going to share the most important
0:19parts with you and then give some extra
0:21information that will help make sense of
0:23this. Cyrus writes, "When I worked as a
0:25Google quality raider, we were required
0:28to evaluate every web page we scored on
0:30effort alongside originality, talent,
0:33skill, and accuracy." What exactly is
0:36effort, and why is it so important?
0:38Google's definition isn't exactly what
0:40you think it is. Effort does not equal
0:42work. So, before I go deeper, if you
0:45don't know, Google has these things
0:46called quality raers. These are human
0:49contractors who manually review and
0:52score search results to help Google
0:54figure out how well its algorithms are
0:56doing. The feedback these quality raers
0:58give is just one signal out of many that
1:01influence the changes Google makes to
1:04its algorithms. And we see those changes
1:06in spam updates, core updates, many
1:08updates that Google pushes out every
1:10year. Back to Cyrus's article, he says,
1:13"It may seem strange for Google to
1:15evaluate pages on subjective factors
1:17like effort and talent. How can Google
1:20possibly know how much effort you put
1:21into a page? In truth, they don't. But
1:24they can very easily estimate it. And
1:26with a high degree of accuracy based on
1:29strict standards and the evidence
1:31visible on your page, Google can draw on
1:34thousands of humanrated documents,
1:36machine learning trained at scale, and
1:38now AI to assess the quote unquote
1:41quality signals of virtually any web
1:44page you publish. And over the past 20
1:46years, they've gotten very, very good at
1:49it. SEO tools typically focus on things
1:51like keywords and topical relevance. But
1:53there's a strong case that subjective
1:55features such as effort are equally and
1:58possibly more important in Google's
2:00eyes. In the age of loweffort scaled AI
2:03content getting algorithmically
2:05decimated, understanding high effort is
2:07increasingly relevant to brands. This is
2:10how Google defines effort in its quality
2:12raers guidelines. Effort. Consider the
2:15extent to which a human being actively
2:17work to create satisfying content.
2:19Effort may be direct, such as a person
2:21translating a poem from one language to
2:23another. Effort may go into designing
2:26page functionality or building systems
2:28that power a web page, such as the
2:30creation of a page that offers machine
2:32translation as a service to users. On
2:34the other hand, the automatic creation
2:37of thousands of pages by running
2:39existing freely available content
2:41through existing translation software
2:43without any oversight, manual curation,
2:46etc. would not be considered to have
2:48effort. The Raider guidelines then go on
2:50to give exhaustive examples of both high
2:53and loweffort web pages. Effort is
2:55mentioned over 100 times. Most
2:58interestingly, the May 2024 Google API
3:01leak exposed two data points that
3:03specifically reference effort scoring.
3:06There's content effort, which is an LLM
3:08based effort estimation for article
3:10pages, and UGC discussion effort score,
3:14which is UGC page quality signals. To be
3:17fair, we don't know how or even if
3:19Google uses these metrics, but it's
3:21obvious from documentation, public
3:24statements, and data leaks that Google
3:26spends a lot of time thinking about
3:27evaluating and scoring content effort.
3:30Understanding how Google evaluates
3:32effort can help us produce content with
3:35a better chance of ranking highly. That
3:37I really agree with. This is a cool
3:39section. Cyrus says, "Before we
3:40understand what effort is, it is
3:43important to know what it is not. Effort
3:45is not word count. It's not time spent
3:48creating a page. It's not human written
3:50content is better than AI. The key is
3:52this. Effort is not how much work you
3:55put into a page. It's more about the
3:57evidence of useful work that ends up on
4:00the final page. And Cyrus gives five
4:02really great examples. So there's
4:04curation and editing. Loweffort curation
4:06is just a large collection of chicken
4:09recipes scraped from other sites. High
4:11effort curation is a collection of
4:13chicken recipes personally tested by the
4:16author organized by popularity and
4:18taste. Then there's lowquality
4:20automation versus added value. Cyrus
4:23writes highef effort pages typically
4:25contain at least some elements that AI
4:28or automation can't easily replicate
4:30like information from proprietary
4:32databases or human evaluation and
4:35perspectives or highquality userenerated
4:38content unique images and other media.
4:40authentic user reviews. And an example
4:43of loweffort automation is 5,000
4:46location pages generated from the same
4:48template with city names swapped in. So
4:50think you have a page for digital
4:52marketing expert in New York, digital
4:54marketing expert in Los Angeles, digital
4:56marketing expert in Philadelphia. The
4:58only thing that's changed is the city
5:00names. Everything else on the pages are
5:02the same. Those are doorway pages. That
5:04goes against Google's guidelines. High
5:06effort automation would be automated
5:08movie recommendation pages generated
5:11with real user reviews from an owned app
5:14along with proprietary realtime ticket
5:16data. That's a great example. Next,
5:18there's recycles information versus
5:21original facts and perspectives. To put
5:24it simply, loweffort content uses the
5:26same facts and information that can be
5:28found elsewhere on the web. It's the
5:30type of content most easily reproduced
5:32by anyone with access to AI, Google, and
5:35Wikipedia. High effort content, on the
5:37other hand, contains original facts
5:39gained through your own experience and
5:41unique perspective/opinions.
5:43Importantly, images and video count as
5:46content, too. Recycled images and videos
5:48borrowed from other sites and/or stock
5:51images and video can be considered
5:53loweffort. This applies to images or
5:55videos that have been slightly altered.
5:57High effort media includes detailed
5:59high-quality photography and original
6:01video. High-quality photography doesn't
6:04guarantee high rankings, but it does
6:06show you put the work in. I have an
6:07episode about this. Episode 1,56 of this
6:11podcast, backlinks aren't enough. How to
6:13win competitive SEO. If you are in a
6:16competitive niche, one of the best
6:17things that you can do is have an image
6:19at the top of your page, especially on a
6:22commercial page that explains things so
6:24clearly without having text in it, that
6:27all you need to do is look at the H1 on
6:29the page and then at the image and you
6:31just understand what the page is about.
6:33You don't even have to read the
6:34paragraph text. So, if you are a SAS,
6:37for example, and you're targeting the
6:39keyword room planner app, then there's
6:41an image right next to it that shows
6:43exactly how the app works and you can
6:45look at this image, understand it
6:46immediately, and then click the use now
6:49button without having to read any other
6:51text cuz you just understand this is how
6:53this works. That would be a custom
6:55image, too. Cyrus gives an example of
6:57loweffort information. A page titled 10
7:00best protein powders describes 10
7:02popular powders without any original
7:04research or testing. High effort
7:06information. A similar page titled 10
7:09best protein powders we've tested
7:11describes the user's experience, testing
7:13methods, experiment results, opinions,
7:16and richly detailed photos of each
7:18product. I will add you want to give the
7:22results upfront before going deep into
7:25the testing methods, the users
7:27experiences, opinions. You give the
7:30answer upfront. Everything else is extra
7:33information and I'll explain why that's
7:35so important in a moment. The next one
7:37is filler content versus prominent
7:39helpful content. This is about fluff. So
7:42loweffort filler is a page titled how to
7:45boil an egg. It begins with 600 words
7:48about the history and nutritional value
7:50of eggs. High effort content is a
7:52similar page titled how to boil an egg
7:54perfectly every time. It gives important
7:56facts up front followed by useful
7:59supplementary information such as egg
8:01size, altitude adjustments, and yolk
8:03firmness, but you still got to say how
8:05to boil an egg perfectly right at the
8:07beginning. You got to give people what
8:09they want. The last thing on this list
8:11is shallow discussion versus quality
8:14discussion. And this one could help
8:16reflect why some threads on Reddit rank
8:19while others don't. Cyrus writes, "For
8:21pages like social media posts or forum
8:24discussions, the level of participation
8:26and depth of conversation is an
8:28important part of effort. Contributions
8:30from multiple individuals on such pages
8:33can add up to a significant amount of
8:35total human effort. A loweffort
8:37discussion is a Reddit post titled,
8:39"Anyone else allergic to butter?" And it
8:41gets a single reply with somebody
8:43saying, "Yep, me. High effort discussion
8:45is a page titled best hiking boots for
8:47women which describes features and
8:49experiences from wearing each boot along
8:52with in-depth perspectives and opinions
8:54which inform the final recommendations.
8:57So that one's an actual article but it's
8:59giving opinions or judgment based on
9:02evidence. Cyrus gives a twominute audit
9:05to create higheffort pages. Before I
9:08share this audit, we got to go back to
9:10this internal Google presentation from
9:1310 years ago that was dug up during
9:15Google's antitrust trial. In this
9:17presentation, Google admits this is
9:19Google's language. We do not understand
9:22documents. We fake it. Our ability to
9:24understand documents directly is
9:26minimal. So, we watch how people react
9:28to documents and memorize their
9:30responses. Let's start with some
9:32background. A billion times a day,
9:34people ask us to find documents relevant
9:36to a query. What's crazy is that we
9:38don't actually understand documents.
9:40Beyond some basic stuff, we hardly look
9:42at documents. We look at people. If a
9:45document gets a positive reaction, we
9:47figure it is good. If the reaction is
9:49negative, it is probably bad. Grossly
9:51simplified, this is the source of
9:53Google's magic. Each searcher benefits
9:55from the responses of past users and
9:58contributes responses that benefit
10:00future users. So, if you search right
10:02now, you'll benefit from the billions of
10:05past user reactions we've recorded.
10:07Google records these reactions with
10:08Chrome and with tons of different pieces
10:11of software that they own. Your
10:12responses will benefit people who come
10:15after you. Search keeps working by
10:17induction. In designing user
10:19experiences, serving the user is not
10:21enough. We have to design interactions
10:23that allow us to learn from users
10:26because that is how we serve the next
10:27person, keep the induction rolling and
10:30sustain the illusion that we understand.
10:32So this was in 2016, Google's quality
10:35raider program started since at least
10:382008. So even 8 years after Google's
10:41quality raider program, Google is still
10:44saying in an internal presentation, we
10:46don't understand documents, we look at
10:48behavioral signals. This is why giving
10:51the searchers what they want upfront
10:54before getting into the extra
10:55information is so important because if
10:57you don't give that information that the
10:59searcher is looking for right away, the
11:01searcher is just going to go back to the
11:02results. And if that happens too many
11:04times, it's going to say to Google,
11:05"This content isn't very good. Let's
11:08decrease the ranking for this content."
11:10And if that happens across your site,
11:11now Google's going to think, "Well, this
11:13site isn't very good. Let's stop showing
11:15this site as often. Let's stop trusting
11:17new content on this site." Google search
11:19advocate John Mueller recently said our
11:23systems can lose faith in websites.
11:26That's his language. But Cyrus's
11:292-minute audit for how to create high
11:31effort pages. It's great for actually
11:33checking if the content we make is going
11:36to throw off the positive user signals
11:39Google wants to see. This audit is just
11:41five questions. Is the content edited
11:43and curated thoughtfully? So, is the
11:45content organized around the user's goal
11:47with the most valuable content easy to
11:49find? Number two, did you add something
11:52unique that doesn't exist elsewhere?
11:54First-party information is critical.
11:55This can include original data, facts,
11:58observations, and even opinions and
12:00perspectives. Three, does the content
12:02show evidence of work? Explaining your
12:04methodology, highquality photos,
12:06screenshots, interviews, and
12:08observations all provide evidence that
12:10you actually have experience. It's like
12:12showing a screenshot of your app on a
12:14bottom offunnel SAS page. Number four,
12:16does the content go beyond common facts?
12:19If someone else could easily reproduce
12:21your content simply by researching the
12:23internet, it's likely considered
12:24loweffort. Like if somebody all all
12:27somebody had to do was do a one single
12:29Google search off the top of their head
12:31and that was enough. Cyrus writes,
12:33explain details or edge cases if you
12:35need to dive deeper into subject matter.
12:38Number five, does the content provide
12:40deep discussion? To be clear, not every
12:43web page requires discussion, but for
12:45many topics, personal perspectives that
12:47explore evidence and offer carefully
12:50crafted perspectives can elevate content
12:52to quote unquote high effort. I'll also
12:54say that not all of these apply to all
12:56types of content. You don't necessarily
12:58need personal perspectives on a bottom
13:00offunnel SEO sales page, but this is
13:03still a great checklist for making
13:04content which that to quote Google gets
13:07a positive reaction cuz that's what
13:10Google says that they measure reactions.
13:12The write up for this will be in the
13:14description for this episode if you want
13:15to hear Cyrus Shepard's first time on
13:17the podcast. Episode 87, Link Building
13:212025, first 10 links, linkable assets,
13:24and HCU wins with Cyrus Shepard. He also
13:27talks about his time as a Google quality
13:28raider in this episode. That was a great
13:31episode, Cyrus. I hope you'll come back
13:32on for many more. And this is episode
13:351,195
13:37of the podcast 1,195 days in a row doing
13:40this show. If you want to learn how to
13:42make those bottom ofunnel SEO landing
13:44pages I talked about in this episode,
13:46how to find the keywords where people
13:48are searching, looking to become paying
13:50customers, users to call you up as a
13:52warm lead, how to do link building that
13:54gets meaningful links, how to structure
13:56your site for these conversion-based SEO
13:58landing pages, examples in every niche,
14:00stepbystep breakdowns and walkthroughs
14:04and guides, how to do an SEO technical
14:06audit, so much more. This is my SEO
14:08course, Compact Keywords. It's 13 and a
14:10half hours long. I spent a year making
14:12it and you are going to love it. That's
14:14at compactkeywords.com. If you watch
14:17this episode on YouTube, thank you so
14:18much for watching. If you listened on
14:20Spotify or Apple Podcasts, thank you so
14:22much for listening and I will talk to
14:24you again tomorrow. Bye now.

About this transcript

This page contains the full transcript of Google’s Secret “Content Effort” Score: Here’s What It Looks For by Edward Sturm, generated from the public captions YouTube serves with the video. The transcript has 2,425 words across 382 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.