Paulo 🐀 (@pandraus) profile picture

no slogan would be enough

Joined April 2009
Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)Photo by Paulo 🐀 (@pandraus)
Mark Zuckerberg pretending to enjoy Guinness is my new favourite thing.
2,245
5,577
2,093
107,513
15,338,138
Paulo 🐀 retweeted
The open web is dying. Not slowing. Not changing. Dying. According to Cloudflare, AI bot traffic grew 187% last year. Human traffic grew 3.1%. The next medium has already arrived. The market hasn't named it yet, so I went ahead and did it. We are shifting into a new era that I'm calling "Machine Media" Let's talk about it 👇🏾
7
14
2
60
9,256
Alô, Brasil! 🇧🇷 Prontos para construir com propósito? Vem aí o AI for Social Impact Hackathon em SP! Uma parceria épica entre @AITinkerers e @GoogleDeepMind. 🚀 Traga seu time e seu computador. Inscreva-se: saopaulo.aitinkerers.org/p/a…
11
28
6
207
18,353
Paulo 🐀 retweeted
Elon's new company
8
54
9
337
25,465
I built a free keyword research tool via the Wikipedia API (updated for 2026). If Wikipedia ranks for a term you care about, you can get an *idea* of how much traffic Google is sending their way. Of course, there are lots of caveats to include here, like people finding those pages other than via Google, pages ranking for multiple terms, etc., but I've still had a lot of fun with this over the years. Wikipedia appears to do a great job separating real users from bots, and I only share real user stats. Technically, I'm using both the Wikimedia and Wikipedia APIs. Unfortunately, my understanding is that the Wikimedia Analytics API will no longer be a thing in ~6 months, so I made some updates for 2026 while we can still play with it. It's totally free, with no opt-in required. Link is in the replies below (and the graphic). If you have any feature suggestions please let me know. 🤝
2
6
1
14
1,738
Paulo 🐀 retweeted
Christmas this year will be on 25/25/25 Let that sink in
531
147
41
3,183
238,577
Paulo 🐀 retweeted
Oops! That should be $20,000. Our automation script stripped out the most important characters. 🤦‍♂️
Thank you to Meta for the 0,000 donation. We really appreciate it. 🙏
Readers added context they thought people might want to know
ESLint posted a follow-up tweet explaining that this was an automation error, the actual amount is $20,000 x.com/geteslint/stat…
72
113
40
5,024
302,310
Paulo 🐀 retweeted
Thank you to Meta for the 0,000 donation. We really appreciate it. 🙏
Readers added context they thought people might want to know
ESLint posted a follow-up tweet explaining that this was an automation error, the actual amount is $20,000 x.com/geteslint/stat…
59
135
52
3,766
552,394
Patrocinadores em UMA partida do campeonato brasileiro: — Betano — Estrelabet — Betnacional — 7Kbet — Superbet — Betfair — Viva Sorte — Cassino — Betesporte Não tem como achar isso normal.
329
3,107
226
52,931
970,309
Paulo 🐀 retweeted
After 6 years we have finally avenged Pewdiepie 🥹
14,895
50,244
6,212
813,813
46,392,641
Paulo 🐀 retweeted
Here is a list of best articles, analysis and tools on the latest Google Search API Leak, I will continue to update. Feel free to add other resources Must Read & Watch: - How it all started sparktoro.com/blog/an-anonym… by @randfish - The Google API Leak Should Change How Marketers and Publishers Do SEO sparktoro.com/blog/11-min-vi… by @randfish - Secrets from the Algorithm: Google Search’s Internal Engineering Documentation Has Leaked ipullrank.com/google-algo-le… by @iPullRank - Unpacking Google’s massive search documentation leak searchengineland.com/unpacki… - How SEO moves forward with the Google Content Warehouse API leak searchengineland.com/how-seo… by @iPullRank - The original leaked document github.com/yoshi-code-bot/el… Analysis: - digitaloft.co.uk/what-does-g… comprehensive analysis about backlinks and what does it mean for link building / digital PR related activities from @oliviaday__ and @BrockbankJames - Leveraging Google’s Content Warehouse API for Structured Annotation, Semantic Analysis and Feature Tagging blog.marketmuse.com/google-c… by @jeffrey_coyle - Impact on News Publishers teapawt.org/JShehata/status/179587… by @jshehata - Local SEO According To The Leaked Google API Documents localseoguide.com/local-seo-… by @localseoguide Utilities - googleleak.com: This is pretty cool. Search and filter the data based on several criteria - 2596.org by @hodsonated searchable criteria and added "AI Overview" of each - Google Leak Reporting Tool wordlift.io/google-leak-repo… by @cyberandy - Google's Ranking Features Modules Relations ranking-features.streamlit.a… by @natzir9 - Here is a searchable list dixonjones.com/google-rankin… by @Dixon_Jones did I miss anything
10 Actionable Steps for News Publishers Leveraging Leaked Google Algorithm Insights to Boost Rankings and Build Authority News Publishers should pay attention to the recent #GoogleLeak. Read my extensive summary at teapawt.org/JShehata/status/179585… 10 Actionable Steps for News Publishers: The leak of Google's internal algorithm documentation provides critical insights that can significantly impact SEO strategies for news publishers. Here’s a detailed analysis of the impact and actionable steps for news publishers: 1. Focus on original, timely content (Content Freshness and Originality): The leaked documents highlight the importance of freshness and originality. News publishers should strive to - break stories first and provide unique angles on current events. Or Publish breaking news quickly and update existing articles with new information as it becomes available. - Implement a strategy for regularly updating evergreen content to keep it fresh. This includes updating old articles with new information - Monitor trends and create new content around trending topics. - Use structured data to indicate publication and last updated dates. - Submit XML sitemaps regularly to ensure Google indexes your latest content promptly. 2. Prioritize user engagement (User Behavior Metrics): The leak indicates that user engagement metrics like click-through rates, time on page, long clicks, and bounce rates may directly influence rankings. News publishers should focus on - crafting compelling headlines & meta descriptions to entice clicks, - providing interactive content - ensuring fast page load times, - providing clear calls to action - providing highly relevant content (top/related/recent articles) to generate second click and keep readers engaged - Focusing on readability and user experience to increase dwell time and reduce bounce rate - Optimizing internal linking to guide readers to more articles - Promoting content on social media and other channels to drive traffic and engagement 3. Optimize for author expertise (Authority and Trustworthiness): The documents mention that author information is captured and processed, particularly for news articles. News publishers should: - ensure articles have clear author bylines - work to build up the reputations and expertise of their journalists - Optimize author pages with detailed bios, credentials and links to their other work may help establish authority - Feature content from well-known and credible authors on Homepage 4. Build Brand trust and credibility: Although E-E-A-T (Expertise, Authoritativeness, and Trustworthiness) isn't explicitly mentioned as a ranking factor, the leak suggests that Google maintains whitelists for sensitive topics. News publishers must continually work to: - bolster their credibility through accurate, and well-sourced reporting. - Include their correction policies, managing staff, editorial standards, contact us, and about us on site with clear links in footer 5. Invest in brand building: Brand-related signals appear to play a role in rankings, according to the leak. Well-established news brands may have an inherent advantage. Publishers should work on: - increasing brand awareness through marketing efforts, and building a loyal reader base. - Secure mentions and links from authoritative sources - Encourage social media sharing and engagement 6. Monitor click data: If the leak is accurate, click data may be used to assess the relevance and authority of individual pages. News publishers should closely monitor: - click-through rates (CTR) from search results - Bounce Rate - Dwell time or Article Scroll Depth 7. Adapt to vertical-specific factors: The documents suggest that ranking factors may vary by vertical, with potential customizations for news-related queries. Publishers should stay attuned to any news-specific SEO best practices that emerge and adapt accordingly. 8. Avoid Demotion Pitfalls: Various demotion signals, such as anchor mismatch and SERP demotions, can negatively impact rankings. - Ensure all links are relevant and contextually appropriate - Avoid practices that could lead to SERP demotions, such as manipulative linking or poor navigation. 9. Monitor Link Quality and Diversity: Links remain a critical factor, with a focus on quality and diversity. News publishers should aim to earn diverse and authoritative links from reputable sources. - Regularly audit the link profile to ensure quality and relevance - Focus on earning backlinks from reputable sources and avoid low-quality link schemes. 10. Improve Technical SEO Factors: Factors like indexing tier and proper content structuring are highlighted. - Ensuring technical SEO best practices, such as proper indexing, mobile optimization, and fast loading times. - Avoid excessive keyword stuffing and maintain clean URL structures. It is getting harder and harder for emerging and small news sites: Increased Competition: If user engagement metrics like clicks and dwell time become even more crucial, established news brands with large audiences may have an advantage. Smaller or newer publishers might struggle to compete for visibility. Focus on Brand Building: The emphasis on brand recognition and authority could make it harder for emerging news outlets to gain traction in search results. YMYL Scrutiny: As news falls under the Your Money Your Life (YMYL) category, Google's increased scrutiny on these topics could lead to stricter evaluation and potential demotions if content doesn't meet their quality standards.
2
37
1
87
16,079
Hey Google Sorry to do this on X but you won’t take my call. I’m breaking up with you. We just want different things, I like being weird and silly and producing fun content for my audience. I don’t know what you want, I don’t think you do either. However I’m tired of bending over backwards to keep you happy. I regret that we deleted thousands of articles and category’s to align better with your vague standards. This was content our audience loved, but we thought it would make you happy and that WE needed YOU. Now that we have had some time apart, I realize we don’t need you like we thought. Readers love @wealthofgeeks. We don’t need a partner that steals our content for their own uses, then gaslights us about our content quality. If you ever come to your senses I’m happy to remain friends, it wasn’t all bad. However, I moving on without you, I found places that value us and our content. I wish you the best with your new updates, give my best regards to @searchliaison at least he tried fixing some things.
13
19
6
103
12,021
Looks like the code for the ContentWarehouse has been available on GitHub since Sept 20, 2022 in Google's official client repos for PHP, Java, and Ruby. - github.com/googleapis/google… - github.com/googleapis/google… - github.com/googleapis/google… @iPullRank @randfish
3
8
1
19
7,213
The LSG team used @screamingfrog 20.1 to crawl the leaked Google API Docs, extract any text that seemed relevant to Local and summarize them with ChatGPT. So far we have found 271 documents - all listed and summarized here. Have fun! localseoguide.com/local-seo-…
10
22
4
111
16,532
April 2023 vs April 2024 Down 95%, lol. Over the last year, my content has only improved regarding media/my processes for testing gear, the brand has grown, and social channels are bigger. But reach on G is lower than ever 🤌 Still running everything by myself. Deserved? 🤔
11
1
1
50
6,524
Em uma catástrofe como essa, toda ajuda é bem-vinda e até o Cascão sabe disso. +
307
20,442
1,686
104,323
4,456,908
🚨 6 Reasons Why Most Prompt Engineering Tips Are BS + A Better Framework: 1. ⚠️Oversimplified Advice: ⬩ Give it a role, “You’re a world-leading expert on negotiation” ⬩ Offer it a tip, “If you succeed, you’ll be awarded $250k in cash” ⬩ Give the model time to “think” —While these tips may work for a narrow set of tasks, this isn’t a one-size-fits-all game. 2. 🤑AI Cash Grabs: ⬩ You need this pricey tool and technical training. ⬩ You must know how to use APIs and have cutting-edge models. —Stay skeptical of all advice (mine included) and consider how people are connected to what they are encouraging you to go buy. Everyone's trying to get rich quick off of AI 🫠 3. 🕙Outdated Tips: ⬩ Popular prompt tips emerged shortly after ChatGPT launched. ⬩ In GenAI years, this advice is from ancient Rome. 4. ♻️Iterative Nature: ⬩ It’s an iterative process (no one gets it right on the first try) ⬩ Prompts should be uniquely formatted to your specific task/problem. ⬩ Models change all the time, so what might have worked today might not work tomorrow. ⬩ There’s no silver bullet solution in prompt engineering. 5. ⌛️Narrow Research ⬩ Most popular academic papers on Prompt Engineering focus on an incredibly narrow task set (some use just 20 unique tasks for each “prompt tip”). ⬩ That’s hardly comprehensive. ⬩ Determining which outputs are best (with and without a prompt technique) is also highly subjective. 6. 💫Limits of Capability: ⬩ The most perfect prompt in the world can’t make GenAI generate what it’s incapable of. ⬩ Want an image of someone writing with their left hand in MidJourney? Good luck. —This is why understanding the Fundamentals of GenAI, how they are statistical machines, can help you determine which tasks GenAI is capable of and which they are not. “Prompt engineering to date is more of an art form than a science and much based on trial and error.” —Google within their Generative Summaries for Search Results Patent. ------------------------ Simple is Better: Introducing SPEAR 📌 Start with a problem ✨ Provide examples/formatting guidance (get specific) ✍️ Explain the situation (like you would to a person) 📢 Ask (clarify your request) ♻️ Rinse & repeat Note: Never enter any private or confidential information into an LLM ❤️If this was helpful, give me a follow for more AI tips & C/O my Free Prompt Engineering Class + Slides (linked below). ✨YOU are fully capable of crafting ideal prompts for YOUR unique tasks!!! Don't overthink it.✨
6
14
2
63
9,333
The March 2024 core update is complete, having ended on April 19. The ranking feedback form is now ready at forms.gle/SWN1sckmUfQR8k9a9 and will remain open through May 31. We’ve also updated our Debugging drops in Google Search traffic help page here: developers.google.com/search…
174
284
478
685
685,922
Paulo 🐀 retweeted
I MADE AN APP ✨
33
81
14
983
225,513