AI Content Strategy / Management / Reviews
AI content strategy, built and tested in public.
Shannon Sheriff is an AI content strategist who plans, manages, and reviews content for the age of AI search. She has spent more than ten years in content strategy and management, runs content and website audits built around how AI engines read the web, and has been building with AI since 2023. She writes about answer engine optimization (AEO), social media algorithms, and blockchain tokenization, and is the author of NFTs: The Future is On Chain (2024).
What I have built with AI
Reporting agents 01 / 08
I got tired of pulling the same numbers every week, so I taught agents to do it. Now the reports just show up in my email and Slack.
Live from X
Algorithm shifts, AEO news, tokenization headlines, and the crypto keywords moving right now. Scroll sideways.
X just made 2 big video changes. 1. You can now clip 5 to 60 seconds from any video over 2 minutes. The clip credits the original, and both the creator and the clipper earn. 🔥 2. X's open-source code now labels AI video 2 ways: "ai_video_self" when the creator discloses it, and "ai_video_detected" when X catches it. → → Make your videos longer than 2 minutes. → → Label your AI video before X does.
Everyone is discovering a flag in X's open-source code called is_pasted. It tracks whether a reply was pasted instead of typed. It kicks in when you reply to accounts with 250K+ followers. A low score can push your reply down. Your own posts aren't affected. So if you reply to the big accounts, type it out.
Avalanche will run the ledger for the clearing firm on NYSE's planned 24/7 tokenized stock platform. Same morning, a Solana venue told the SEC it's coming in 2027. Wall Street is picking chains.
LinkedIn's VP of Engineering says every post shown beyond your connections now goes through classifiers trained on 1M+ member reports. AI still helps with the draft, but the post has to say something only you could say. Translation: don't post AI slop.
Tokenized asset volume on Solana went from under $100M a month to nearly $5B. Actual trading volume.
ethereum:native
I’m Building an Army of Marketing Agents. Meet the First 7. Every marketing decision should come from the last round of data. What worked, what didn’t, and what to do next. Most teams already have the data. It’s just scattered. Leads live in the CRM. Ad
OpenAI appears to be building a dashboard showing businesses their ChatGPT mentions and website clicks!! Still in development, not officially launched. Another metric for my agent to track, and one I’m genuinely excited about.
Anthropic is researching how Claude could fix its own bad memories between conversations. → Check old notes against the original conversation → Correct something it got wrong → Update information that’s changed → Combine scattered notes so useful details are easier to find This isn’t a new feature in every Claude chat yet. But Claude’s memory has been legit good the past couple months. If it can start catching and fixing its own mistakes too, even better.
Visa tested 22 AI models on shopping scenarios. They passed 57.2% of the time. The tests required them to ask questions and figure out what the shopper actually needed before deciding whether to buy. There’s room to improve, but this gives developers something concrete to work on.
Follow along on X: @shannonsheriff
Coverage
Three beats, one question underneath all of them: what decides which people, brands, and assets get surfaced, and which ones get skipped.
How ChatGPT, Claude, Perplexity, Gemini, Grok, and Google AI Overviews choose which sources to cite, and what changes when they update.
Ranking and distribution changes across X, LinkedIn, Instagram, TikTok, YouTube, and Facebook, translated into what to actually do differently.
Tokenized stocks, tokenized real estate, real world assets, AI agent wallets, and the crypto keywords people are searching at any given moment. Member of the MetaMask Agent Wallet Traders beta group.
Building with AI
I have been building with AI since January 2023 and have built more than ten AI reporting agents. I am not a developer. I learn by building, then I write up what worked and what broke so anyone can follow along.
Track record
Built a payments infrastructure company's X presence in under six months while holding a 5 percent engagement rate, well above the platform average for technology accounts.
Used entity mapping, structured content, and custom GA4 channel groups to make AI engines a measurable, recurring traffic source.
Walmart's longest standing content creator through an exclusive GigPig contract, producing roughly 150 social ads.
Audits sites for how AI engines and search crawlers read them: crawler access, structured data, page structure, entity consistency across profiles, and whether key pages answer the questions people actually ask.
Reviews content before it ships: fact and source checks, brand voice, structure for AI answer engines, and the patterns that make copy read as unedited AI output.
A 2024 book on NFTs and on chain ownership, written for people coming to blockchain from outside the industry.
Writing
Public work turned into revenue and job offers at OpenAI, Vercel, and xAI this year. The people it happened to started before they felt ready.
"A stranger who finds you in month six doesn't meet a pitch. They meet half a year of proof."
FAQ
Shannon Sheriff is an AI content strategist focused on content strategy, content management, and content reviews. She also runs website audits for AI search, writes about answer engine optimization, social media algorithms, and blockchain tokenization, and is the author of NFTs: The Future is On Chain (2024).
SEO (search engine optimization) works to rank pages in a list of search results. AEO (answer engine optimization) works to get a person or brand cited inside the answer an AI system writes. SEO leans on keywords and backlinks. AEO leans on clear entity signals, structured data, and consistent facts across the web. Short version: SEO gets you ranked, AEO gets you cited.
Shares, especially copy link shares. In the For You ranking weights X published in August 2026, a predicted copy link share carries 20.0, the highest positive weight. A reply, quote, or DM share carries 5.0, and a like carries 0.5. Write posts people want to send to someone.