Teardown No. 29 · llms.txt

How Anastasiia Kyslenko got ChatGPT and Gemini to cite her blog with zero ad budget

Anastasiia Kyslenko structured every article on her Ad Adastra blog so an AI could pull a ready answer out of it. She did not buy links or pay for SEO promotion. A few months later, GA4 showed visitors arriving from chatgpt.com and then gemini.google.com. She describes the traffic as "a handful of sessions a month" and treats it as a signal, not a traffic channel.

Anastasiia KyslenkoVerified Oct 16, 2026Markdown copy
Part of playbook: Make your site easy for AI assistants to cite, and measure whether it works

The problem

She wanted her content to show up in AI answers without spending on ads, link building or an SEO agency. Her blog is “not a corporate site,” and she did the work “without a dedicated team and without a separate GEO budget.”

The blueprint (as she describes it)

She lists “six things I do systematically, in every piece, with no exceptions,” plus a monthly check.

  1. Open with a direct answer in the first 40-80 words. The first sentence answers the question in the headline. In her words, “AI pulls exactly the opening sentences.”
  2. End every article with an FAQ block. It has 5-6 questions and answers, marked up with FAQPage schema.
  3. Use tables and numbered lists instead of walls of text. Her examples are pricing model comparisons, a cost breakdown by layer and a nine-step checklist.
  4. Cite a source for every external fact. The link to the original source sits right next to the number, with a specific name and year.
  5. Keep one description of herself everywhere. The author byline, the About page and service descriptions all carry the same core: “Anastasiia Kyslenko, 8 years of experience, 105 clients, Google Partner status.” Her checklist adds social profiles.
  6. Go deep on niche topics. She writes about GEO optimization, AI Max vs. Performance Max and product schema for AI, which she calls “topics most solo specialists in my niche only skim.”
  7. Measure two ways. She checks GA4 for referral traffic from chatgpt.com and gemini.google.com. Once a month she also asks ChatGPT and Gemini questions from her niche and checks whether she shows up. Her checklist suggests 10-15 questions. She calls this “essentially measuring Share of Model.”

Stated results

  • In spring she noticed chatgpt.com as a new referrer in GA4, and gemini.google.com a bit later. Her status line is dated “As of September 2026.”
  • Traffic is “a handful of sessions a month.” No exact count is stated.
  • First mentions showed up “after roughly a couple of months of systematically restructuring content.”
  • Spend: no ad budget, no separate SEO agency budget and no link buying. “The resource I spent was my own time restructuring content”
  • Her own read on what worked fastest: the direct answer at the top of the article, then FAQPage schema. She thinks she “was cited for specific questions, not the general topic of the article.”
  • No conversion, lead or revenue figures are stated.

Why this matters for postby.ai

  • Our teardowns already open with a short answer paragraph. Keep that first paragraph a self-contained answer to the title.
  • An FAQ block with FAQPage schema is a cheap test to add to teardowns.
  • Use one consistent description of Postby across the byline, About page and social profiles.
  • Pair her GA4 and monthly-question checks with server logs, as in the Umar Gora teardown.

Caveats

  • First-person and honest about scale, but the post promotes her services. It offers a review of “how citation-ready your content actually is,” a structural audit via Telegram and a one-page PDF checklist delivered via Telegram. Whether the review is paid is not stated.
  • She says plainly: “I do not have a controlled A/B test that proves cause and effect.” She also cannot confirm the role of external mentions or site speed.
  • She gives no visit numbers, only “a handful.”
  • The post cites outside figures we have not checked. A Victorious study covered 177 brands across 8 AI platforms in Q1 2026 and found only 18 had any AI mention. AirOps found only 16% of Fortune 500 companies track their AI search visibility. Her “2-6 weeks” timeline for first mentions is attributed to “GEO research” with no named source.
  • X: Unverified not checked.

Sources

Want agents like these working your distribution? See the Agent Stack Teardown. Get new teardowns weekly: subscribe (double opt-in).

Get one teardown a week.

Same format, a new operator every issue. Double opt-in.