Skip to content

Insights

Turning your expertise into content that gets cited

Torn Studio3 min read

Short answer

Content that gets cited comes from people who already know the subject. The method is a 40-minute interview, a draft the same day, a fact-check by the expert, and publication within the week. Freshness carries weight: about half of what AI systems cite is under 13 weeks old, per Profound’s citation-decay analysis.

68% of US Google searches now end on the results page, per SparkToro and Similarweb data for January to April 2026. Text quoted inside the answer therefore does more work than text that gets visited, and that raises the question of where the text comes from.

Why generic writing does not get cited

Text written by someone with no subject knowledge comes out generic, because the generic version is the only one available to a writer who knows nothing. The risk is stated policy too: Google’s March 2026 spam update classified scaled content abuse as spam, and the rule turns on whether a page gives the reader something, whatever produced it.

The interview is the production step

40 minutes with the person who knows the subject yields more usable text than a week of research. The questions are simple: what do customers ask most often, what do they usually get wrong, which decision do most of them regret, and which numbers do you have that others lack.

From conversation to published article

  • A 40-minute interview, recorded and transcribed
  • A draft the same day, every claim traceable to something that was said
  • A fact-check by the expert: figures, conditions and exceptions
  • An edit against a fixed shape, direct answer first and the questions last
  • Publication, internal linking and a visible date

The rhythm is part of the method

One article a week beats four in a single day each month, because the rhythm is what survives a busy period. One interview session a month sustains that pace if each conversation is broken into three to five pieces. In our experience that breakdown is the step most often skipped.

What makes the text worth quoting

A self-contained 40–60 word answer at the top, headings phrased as questions, specific numbers with a stated source, and a visible updated date. Mentions off the site carry more weight than links: Ahrefs’ study of 75,000 brands puts branded mentions at 0.66 correlation with visibility in Google’s AI Overviews, against 0.22 for backlinks.

Common questions

How much of the expert’s time does this take?
One 40-minute interview a month, and roughly half an hour per piece for the fact-check. The writing, editing and publishing sit with us.
Can AI write good content?
Yes, when the model has something real to work from. A transcript from an expert, your own numbers and a fixed structure produce text that holds up. An empty prompt produces generic text, which is exactly the kind of page Google’s March 2026 update hit.
How do we know what to write about?
Start with the questions you already answer in sales calls and support. They have demand, you know them by heart, and they carry buying intent, which is where the clicks remain once informational searches are answered in the results page.
How long should an article be?
Long enough to answer the question. Ours usually land between 700 and 1,200 words. Length ranks nothing on its own; what counts is that the answer appears early and the rest earns the reading.
Should we publish in several languages?
Only in the languages you actually sell in, and only as real text written in that language. A machine-translated copy of everything costs more to maintain than it returns, and it looks like volume content to whatever classifies the site.
How do we measure whether the content works?
Track three things: organic sessions to the articles, how often they get named when you put your own questions to ChatGPT, Perplexity and Google, and the share of enquiries that read at least one article first.

Sources

  1. In 2026, Less than One Third of Google Searches Still Send a Click — SparkToro

    Backs the 68% figure: SparkToro’s analysis of Similarweb clickstream data for US searches, January to April 2026.

  2. Introducing Citation Decay in Profound

    Backs the finding that about half of what AI systems cite is under 13 weeks old.

  3. An Analysis of AI Overview Brand Visibility Factors (75K Brands Studied) — Ahrefs

    Backs the 0.66 correlation for mentions against 0.22 for links, across 75,000 brands.

Tell us what you want to build

Thirty minutes, free of charge, and a straight answer on whether we are the right studio for it.