From Ranking to Cited: The Practice-Page Conversion Worksheet

Turn a page that ranks #1 into a page AI actually names in the answer.

Clientory — clientory.org

How to Use This Worksheet

At Google I/O 2026, the company confirmed what the data had been signaling for months: rankings and citations have decoupled. The share of AI citations drawn from top-ten results fell from roughly 76% in mid-2025 to about 38% by early 2026. Being on page one is no longer enough to be named inside the answer. Since Gemini 3 became the default for AI Overviews in late January and AI Mode crossed a billion users, Google's query fan-out splits one client question into many sub-queries — and pulls the clearest, best-structured passages, not just the highest-ranked URLs.

For an immigration firm that means a practice page can hold its #1 spot in traditional search and still go uncited when a prospective client asks an AI assistant "who can help me with an H-1B transfer" or "do I qualify for adjustment of status."

This worksheet walks one of your strong-ranking practice pages through the conversion from rank-ready to citation-ready: a direct answer up top, dated proof, and clean structure an AI can lift verbatim. Budget under 60 minutes. Have the page open in one tab and your CMS in another.

Pick your page: choose one practice-area page that ranks well but you suspect is not being cited. Write it here.

• Page URL / title: __________________
• Case type: __________________
• Current rank (if known): __________________

Part 1 — The Decoupling Check

First, confirm you actually have the problem this worksheet solves: strong ranking, weak citation. Run your page's primary query through two or three AI assistants (Google AI Mode, ChatGPT, Perplexity) and record what you see.

Reading your result: if you rank on page one but were cited in zero or one of the three, you are living the decoupling — a page-one page that the answer layer skips over. That is the page worth converting. If a competitor keeps appearing, open their page in Part 4 and note what their answer block does that yours does not.

Part 2 — The Direct-Answer Test

AI assistants lift the passage that answers the sub-question most directly and earliest. Most firm pages bury the answer under a warm-up paragraph about "navigating the complex immigration system." Check how your page opens.

• The first 2–3 sentences directly answer the page's core question (e.g., "An H-1B transfer lets you change employers without re-entering the lottery. Here's who qualifies and how long it takes.")
• The answer names the specific case type, agency, and form (USCIS, Form I-129, PERM, adjustment of status) rather than speaking in generalities.
• A reader — or a model — gets a complete, quotable answer before any call-to-action or firm bio.
• The opening avoids throat-clearing ("In today's ever-changing landscape...") and gets to substance in sentence one.
• There is at least one clean, self-contained sentence that could be quoted on its own and still be true and useful.

Score: ___ / 5 checked. What's missing: __________________

Part 3 — The Dated-Proof Test

Models favor content that signals it is current and authoritative. Vague, undated claims read as stale — and in immigration, where rules move quarterly, stale is disqualifying.

• The page carries a visible, honest "Last updated" date.
• It references at least one 2026 development by name and date (e.g., the weighted H-1B selection rule, a recent BIA ruling, updated USCIS processing times or filing fees).
• Processing times, fees, and form versions are current — not 2023 figures.
• Claims of experience are specific ("In the past year we've filed 40+ adjustment-of-status cases") rather than generic ("decades of experience").
• No stale tells: dead links, superseded policy described as current, outdated form editions.

Score: ___ / 5 checked. Notes: __________________

Part 4 — The Structure Test

Query fan-out rewards pages a model can parse and extract cleanly. Tangled structure means your good answer never gets lifted.

• One clear H1, with logical H2/H3 subheads that each pose or answer a real client question.
• An FAQ section using the actual words clients type ("Can I change jobs on an H-1B?" not "Portability considerations").
• FAQ or how-to schema markup is present where possible.
• Key facts sit in short paragraphs or lists near the top — not buried in paragraph six.
• Compare against the competitor cited in Part 1: note one structural thing their page does better.

Competitor's edge to copy: __________________

Score: ___ / 5 checked. Notes: __________________

Part 5 — Summary & Next Steps

Tally your checks: Part 2 ___/5 · Part 3 ___/5 · Part 4 ___/5 · Total ___/15

• 11–15: Citation-ready. Re-run Part 1 monthly and protect this page when policy shifts.
• 6–10: Half-converted. The answer or the proof is there but not both — finish the weakest part first.
• 0–5: Ranking on borrowed time. This is your highest-leverage page to rebuild this week.

Based on your answers, your 2–3 highest-priority actions:

• __________________
• __________________
• __________________

Refresh by: __________ Owner: __________

The pattern that wins citations is consistent: answer the question in the first three sentences, prove the page is current with a date and a 2026 fact, and structure it so a model can lift the answer whole. A #1 ranking earns the crawl. These three moves earn the citation.

Built by Clientory — helping immigration law firms become visible to the families and individuals who need them most.

Request a GEO visibility audit