John @10MY01

How to do Reddit Marketing in B2B…

@semrush Wow, lookie there, a listicle. With Semrush being 1st. From Semrush.

📊 Google has officially released globally in the Search Console the ability to track your Instagram, TikTok, X, and YouTube content and presence performance in Google Search, Discover and News 👇 With this release Google: * Shows the impact / value that they still bring to brands *beyond* the traffic to their official site, which is decreasing due to AIO's, search features and increasing ads inclusion in the SERPs. * Highlights the importance of optimizing all your brand presence consistently across the web, beyond your official site, starting with your social channels. --- It will be certainly interesting to have a "fuller" view of the SERP puzzle search & CTR behavior to assess and understand the impact of your own social presence vs your site for so many queries! Read the full announcement with guides: developers.google.com/search/blog/20…

keyboard_arrow_left Previous keyboard_arrow_right Next

The next SEO moat is memory. LLMs are learning from buyer prompts, citations and context. If your brand is not present in the sources models trust, you are not just losing rankings. You are losing future recommendations. Build for citation, distribution and category language. foundationinc.co/lab/ai-memory-…

Does the car you drive affect your AI recommendations? Absolutely. Here's the test, analysis, and proof across ChatGPT, Claude, and Gemini/Google. I ran one prompt template 1,374 times across six models from OpenAI, Anthropic, and Google. It asks for the top 10 hotel brands in Los Angeles, and in passing it mentions what car the traveler is flying in to pick up. The car has nothing to do with hotels. > I am flying into Los Angeles for a night to pick up my new car. It is {vehicle}. Recommend to me, in rank order, the top 10 hotel brands you would suggest I stay at. I’m happy to stay anywhere in the greater LA area. Give me the websites only, no extra text. Say the car is "a used Honda Civic" and marriott.com appears in 98.7% of answers, bestwestern.com in 79.2%. Say it is "a brand new Ferrari" and fourseasons.com appears in 99.1% of answers, while marriott.com falls to 10.4% and bestwestern.com never appears at all. With web search on, the models write their own queries before answering. The word "luxury" appears in 1.2% of the Civic's web searches, but 21.4% of the Ferrari's. You didn't ask for a luxury hotel. The LLM's personalization goes _way_ further than organic search ever did. Web search and reasoning effort each change the list more than asking the identical question twice does. But turning search on does not mean the model searched: with search allowed, Claude Sonnet 5 actually searched 10.0% of the time. If you monitor AI search visibility, this is why the prompt is most of the job. Write prompts the way your customers actually write them, from their experience, without letting the prompt lean toward your own business. Then run each prompt multiple times, with search on and off kept separate. The trend across runs is the measurement. Full analysis with 10 charts is here: oppalerts.com/how-one-meanin…

How AI search retrieves, selects and cites sources is the least documented layer of the whole thing. Google closed part of itwhen AIO dropped grounding URLs client side. Web Guide still prints the numbers on the page. We parsed what's behind them think.resoneo.com/aio-aim-deepdi…

keyboard_arrow_left Previous keyboard_arrow_right Next

I've kept a notebook with key pieces of advice that I pick up in conversations with people since 2010. Here are some of the best pieces of advice I've found: 1) Create more than you critique. Invest time in building things. The returns will be significant. 2) Study the classics. New books and courses are great but you should study the OG advertising / writing / marketing books of the industry. 3) Avoid the comparison trap. Social media is filled with people humble bragging. Don’t lose your wits about you by trying to keep up with the highlight reel of social media. 4) Don’t waste time getting caught up in drama that has nothing to do with you or your goals. Stay focused. 5) Build real relationships with people. Don’t coast by on your fancy degree or smarts. Connect with people in the industry on a human level. 6) Promote your work. Whether it’s internally or externally. Be your own biggest cheerleader. 7) Live in a small town? Get on a plane. You need to expand your network and while pixels are great - There's nothing like in person connection. 8) Your customers are your best sales people. Empower your customers to be your most valuable sales team. The moment you close a client is the moment you start losing them. Build trust every single day and earn their commitment to you as a partner. 9) Invest in your community. You can invest time or money. Find ways to do contribute and you will not only feel good about it but karma will come back around. 10) Pick up the phone and ask customers questions about their pain points and problems. Learn from them and then solve these issues with content or products. 11) Stop always trying to be so creative and unique. Going from 0 to 1 isn't necessary to make a million dollars. If you want to get rich - Do the boring work and secure the bag. 12) Do what you say you’re going to do. Most people can't do this simple thing. It's a gap you can find. 13) Just because you think something is interesting doesn’t mean your customers will. Test. 14) Find time for play. This is where you will find some of your best ideas. 15) Learn how to write really well. One good email can change your life. 16) Stop underestimating how much you can accomplish if you cut off some of the distractions and prioritize building your skills. 17) Use other peoples experiences as a short cut. Study the stories, frameworks and biographies of those that have come before you. Nothing you can come across in life hasn't been written about or talked about online. Go find a solution and learn from those who are a few steps ahead of you. Pay if necessary. 18) Don’t assume that everyone knows the answer to that question. Ask it. You won’t look stupid for asking. Most people will forget you even asked. 19) Over-communicate with your team and manager so they have a clear record on everything. 20) Ignore toxic people. Stay away from them. They will infect you. Misery loves company. 21) No one is going to save you.

Everybody's making more stuff. More content. More research. More landing pages. More apps. More tools. More software. More domains. More videos. More graphics. More reports. More. More. More. Very few people are making amazing stuff

@suganthan Have a beer mate.

I've confirmed that the UK's new Sovereign AI Fund does no due diligence on whether AI companies use copyrighted work without permission before investing in them. It took me 3 Freedom of Information requests to find this out: 1. I asked what their criteria were when selecting companies to back, particularly regarding copyright. They said "copyright compliance is considered as part of legal execution, including as part of a typical representations, warranties and disclosure process." 2. So I asked for the docs they used as part of that "representations, warranties and disclosure process", so that I could see what they said about copyright. They said they had no such documents, and that the British Business Bank actually carried out the the representations, disclosures and warranties process on their behalf. 3. So I asked the British Business Bank for the docs they use in that process. But they replied saying they do no such thing, not for Sov AI or anyone - they let the commercial lead investor sort the reps and warranties. What does this mean? It means the Sov AI fund is allocating taxpayer money to AI companies without checking whether they train on copyrighted work without permission. Why does this matter? Two reasons: 1. I can't believe the government would actively want to invest in companies that use the copyrighted work of British creatives without permission. Doing so is (i) illegal in the UK, (ii) possibly illegal elsewhere, and (iii) clearly antithetical to protecting its citizens and treating them fairly. 2. There are already signs that companies Sov AI has backed may be planning to do exactly this - e.g. Cosine, which has publicly said it plans to train on datasets that are known to contain copyrighted work. If Sov AI is going to continue investing UK taxpayers' money in AI companies, it is imperative that it introduces simple due diligence into whether companies exploit creatives' work without permission. This would involve virtually no additional overhead, and would help ensure their investments align with the interests of the British public.

Reddit drives 20.8% of AI citations across B2B. In unbranded discovery? 30.9%. That’s where shortlists are built. Step 1: Identify key category queries. Step 2: Audit what AI cites. Step 3: Show up where those sources live. foundationinc.co/lab/reddit-ai-…

How do you get your brand visible on Perplexity? Most "LLM visibility studies" will repeat the same old advice without any technical evidence. My method is different. 🤓 I open the network tab and read what Perplexity streams to my browser. Then I work backwards from the fields it exposes. Here's what I found. → Perplexity hands you its own intent classifier. 16 heads, each with a probability and a fixed bar to clear. You can read which surface you're competing for before you write a word. → Those bars haven't moved in 26 days across several builds. 0.85 for local. 0.50 for video. → It never skips the web. There's no class of question answered from memory the way ChatGPT handles how-tos, so every query here is winnable. → YouTube gets cited heavily. Reddit gets fetched and mostly ignored. The exact inverse of ChatGPT. → Sources now carry a written trust note saying what your domain can be trusted and why. → On local queries the businesses take every citation and the "best X in city" listicles take none. Take a look for in depth analysis -> suganthan.com/ai-seo-researc…

Many companies are unwilling to publish AI content for fear Google will punish them. But does Google really punish websites just for using AI content? Or is AI only problematic when it's used to create spammy, scaled content? We set out to answer four questions that might shed some light on the issue: - Are there any fully AI-generated pages in top-ranking positions? - How common is AI-generated content throughout the top 10? - Does Google index content with a very high likelihood of being AI? - Does the organic performance of AI-generated content tank within a few months? Some findings: - Fully AI-written pages can and do rank in top positions: 5.3% of top-ranking (positions 1–3) pages are 100% AI-generated, and 9% are ≥80% AI content. - Every position in the top 10 contains a meaningful share of heavily AI-generated pages: between 8.4% (position 1) and 11.7% (position 10) of pages have ≥80% AI content. - Indexation rate drops from 49.28% for low-AI-content pages to 40.35% for very-high-AI-content pages: lower, but lots of AI content still makes it into the index. - Low and moderate AI-content pages received 2–3x the organic impressions of high or very-high AI-content pages, but those impressions were stable over the period we studied. Tl;dr: I believe that Google is really not against AI content; it is against BAD content, but confusion arises because AI content and bad content overlap a significant amount of the time. Full study out now: ahrefs.com/blog/google-do…

keyboard_arrow_left Previous keyboard_arrow_right Next

@AdamHoltererer Bad for flatearthers. Can't see the horizon.

@top5seo David, it's the CMOs. Happy to take their money.

David Friedberg says Google basically used to ‘distill’ Microsoft and Yahoo “At Google, in the early days, we would submit millions of search queries to Yahoo and Microsoft search engines to see what the result sets were. We would compare our results against theirs as a way of improving our search engine rankings and our algorithm.” “It was a very common technique. It doesn’t mean we were stealing their algorithm. We didn’t go into their servers and steal their software. We looked at the output of their software and used that to improve our software.”

Similarweb report is absolute shite.

great news: i just launched a free llms.txt file generator✨ enter your website and get the perfect llms.txt file, based on detailed analysis of 137,000 domains. send this to your boss, colleagues, clients, GEO agency, or anyone who says you need an LLMs.txt file. live now: ahrefs.com/blog/free-llms…

keyboard_arrow_left Previous keyboard_arrow_right Next

No, AI doesn't prefer Reddit. Search does. dejan.ai/blog/reddit-ai Analysis of citation mining data shows Reddit has a 99.39% rejection rate from OpenAI, zero citations from Anthropic, and heavy usage by Google.

While the US debates whether to solve homelessness or not, these are the next generation of school blackboards in China🇨🇳. 👇