Writesonic

@aaliya_va It is

@WriteSonic That’s a big difference.

We analyzed 1,161 ChatGPT citations across both new models. GPT-5.4: 56% go to brand websites. GPT-5.3: 8%. Same question. 7% source overlap.

GPT-5.4 uses site: operators to visit brands directly, checks pricing pages, validates on G2. No previous model did this. 75% of its cited domains don't appear in Google results for the same query. SEO doesn't get you in. Your brand does. Full study + extraction script ↓ writesonic.com/blog/chatgpt-c…

The world has moved on from traditional search. Brands are asking "what resources do you need for AI visibility?" And too many SEOs are responding with "it's the same thing we've always done." @iPullRank's Mike King on why “AI search is a completely different surface.” Watch the full webinar 👇youtube.com/watch?v=6BdJgz…

We're tracking how GPT-5.4 responses differ from GPT-5.2 across brands. The shifts are already visible. Full breakdown of what changed and what to do about it: writesonic.com/blog/gpt-5-4-r…

This changes GEO. Before: optimize for what models learn during training. Now: also optimize for what GPT-5.4 finds when it visits your site right now. It reads your pages, evaluates the claims, cross-references against competitors, and decides whether to cite you or skip you.

GPT-5.4 can now browse your website in real time before answering questions about your brand. It scores above the human benchmark on autonomous computer use (75% vs 72.4%). Your site isn't just for customers anymore. It's a live input to every ChatGPT answer about your category.

You're tracking hundreds of AI prompts. But when someone asks "which ones are we losing to competitors?" you're exporting to spreadsheets and manually hunting for answers. Not anymore. Advanced Filters is now live in Writesonic's GEO Platform. Build multi-condition queries with AND/OR logic directly on your All Prompts and All Citations pages. No exports needed. 9 filters on All Prompts. 3 on All Citations. Filter by brand mentions, citing domains, visibility score, rank, prompt volume, and more. Want to find every prompt where a competitor shows up and you don't? One filter. High-volume prompts citing your domain? One filter. AI visibility data is only useful if you can act on it fast. Start filtering: writesonic.com/blog/introduci… #GEO #AIVisibility #Writesonic #AISearch #MarTech

Your brand ranks #1 on Google. But when a CMO asks ChatGPT "what's the best platform for [your category]"... you don't exist. That's the new reality. And most enterprise marketing teams don't even know it's happening. We mapped where brands actually sit in AI search. It comes down to two questions: → Is your content optimized for AI to cite? → Are you showing up for category queries, not just your own name? Here's what we see across 500M+ tracked AI conversations: 🔴 Sitting Duck AI gets asked about your brand. It gives a vague, outdated answer. Or recommends a competitor. Your own story is being told by someone else. This is where most enterprise brands are today. 🟡 Home Base AI answers your brand queries accurately. That's good. But it's defense. You're only winning searches you already own. No new pipeline is coming from here. 🟢 Growth Engine This is the shift. AI recommends you when someone asks "best tool for X" — and they've never heard of you before. You're cited, sourced, positioned as the answer. This is where pipeline lives. The path from Sitting Duck to Growth Engine isn't complicated. But it is intentional: 1. Fix your brand layer first. Make sure AI knows who you are, what you do, and can answer correctly. 2. Build content that answers category questions directly. Not blog posts about your features. Content that answers the exact prompts buyers type into ChatGPT and Perplexity. 3. Track it. Weekly. AI results shift fast. What worked last month might not work today. The brands making this move now are building a moat. The ones ignoring it are ceding territory to competitors who showed up first. SEO got you rankings. AEO gets you recommended. #AEO #AISearch #MarketingStrategy #CMO #GrowthMarketing

@aaliya_va wild but true. 82% of AI citations come from outside your owned properties. the game changed

@WriteSonic This is wild but makes sense. AI seems to trust real user content more than polished websites.

AI search is walking away from your website 🐧 82% of AI citations = third-party sources. Google and OpenAI pay Reddit $203M/year. Your perfectly optimized website? AI walks right past it toward a Reddit comment with 34 upvotes. Free webinar TODAY 11am ET — why AI trusts UGC more than your website + live Reddit playbook walkthrough. 🔗👇

@Seozilla_ai exactly. the brands winning in AI search aren't the ones with perfect SEO — they're the ones people actually talk about

@WriteSonic Yep-AI trusts real user content more than polished pages. Quick tip: seed genuine Reddit discussions that spotlight your best UGC

Your relationship status with AI search: it's complicated ✅ Google ranks you #1 ❌ ChatGPT doesn't mention you ❌ Perplexity cites your competitor 👻 AI Overviews? Ghost. 11,000+ brands made it official. Track your AI visibility → writesonic.com

AI recommends brands based on 3 signals: ✅ Citations from trusted sources ✅ Content that directly answers questions ✅ Consistent brand mentions Most companies do zero of these intentionally. They're optimizing for Google page 1 while AI is already picking winners without them.

Wait for it... The average video takes 20+ hours to edit. What if it took one click? Studio sound. Smart cuts. Captions. B-roll. Zooms. All automatic. The editing tax is almost over. 👀

Register: us06web.zoom.us/webinar/regist…

Tripled AI search traffic in 3 months & landed multiple $100K+ clients just by tracking & optimizing AI visibility with @WriteSonic. From Yahoo engineer days to dominating AI search—don't get left behind in the old SEO world. Read the full case study: writesonic.com/case-study/bli…

AI search is changing how brands get discovered. Reddit, UGC platforms, and community content now influence what AI recommends. If you're only optimizing for Google, you're missing a big piece. Leigha Henderson on why Reddit matters for AI visibility. Samanyou Garg on agentic workflows reshaping marketing. Free webinar. Feb 11. 11 AM ET. Register → us06web.zoom.us/webinar/regist…

I set up an @openclaw bot yesterday to test this out for some marketing use cases. One of the most important use cases as we are releasing Bansi AI is influencer marketing. So, I set up this Telegram bot called Influencer connected to an OpenClaw instance hosted on a VPS. It did take a good amount of time to set it up as compared to what people are claiming, as for our use case, I needed to give it access to Youtube API, SERP API, and various other things. But once set up, it has been working since yesterday finding relevant influencers on Youtube, adding them to a Google sheet with all the details like subscriber count, engagement rate, etc and even qualifying them based on if they will be a fit for us. Then, finding emails for each of them and updating on the sheet. This if done manually takes days of work to do. It has been sending me hourly updates and has already found so many influencers. Next, want to test if it can also do email reach outs. But yeah, really interesting stuff that's been developing right now. Happy to answer any questions if anyone has about setting up their own OpenClaw bots.

Hiring an AI Growth Marketer Watch our GEO webinar. Turn it into 2 video clips + 1 carousel + engagement workflow. Screen record everything. 3-4 hours. Details: Notion link below Email to sam [at] writesonic [dot] com with subject: AI Growth Marketer – [Your Name]