One original survey = Perplexity citations in 3 weeks. Here's the exact format that worked.
Background: i've done content marketing for other people's SaaS for years, finally building my own tool, so i know the playbook. Decided to run a tiny original-research experiment.
Survey: 110 respondents, solo operators and small agency owners. Topic: how they're handling AI-generated content in their client deliverables. One hero stat that was genuinely surprising.
Distribution: own blog + LinkedIn carousel + one industry newsletter mention.
What happened:
- Perplexity started citing the survey page within 18 days. Not just mentioning my brand — actually sourcing the stat.
- Two industry blog roundups linked it. Those links seem to be what triggered the Perplexity pickup.
- ChatGPT: still nothing after 5 weeks (training data lag, expected).
The format that matters, specifically:
- Hero stat in the H1, not buried in paragraph 2.
- Methodology in a visible callout box (sample size, dates, screener criteria). AI engines seem to treat visible methodology as a credibility signal.
- Downloadable CSV linked from the post. Got 3 of those industry links when people re-cited with credit.
one thing i'd do differently: send the data to 2-3 larger publications as an exclusive before going broad. second-degree citations last way longer than first-party.
Has anyone had a survey flop? Wondering what separates the sticky data from the data that just dies.
4 replies
- Marcus B.
The downloadable CSV as a link-generation mechanism is clever. It turns 'citing the data' into something publishers can do without paraphrasing. They just link to the CSV and credit you. Lower friction than asking them to rewrite your finding.
- Dave A.
How do you handle surveys with small sample sizes credibly? 110 respondents is on the thin side for most topics. Did any publications push back on that, or is 'N=110 + visible methodology' enough in practice?
- Inês Pereira
The 'exclusive to 2-3 pubs first' advice is right. We did the opposite — broadcast it everywhere simultaneously — and got zero second-degree citations because nobody had anything differentiated to say. The publication that gets an exclusive has an incentive to make more of the data. Learned that the hard way.
The methodology callout box detail is gold. I've been putting methodology in a footer note and wondering why it doesn't get picked up. Makes sense that visibility of the methodology is itself a credibility signal — AI engines can't cite what they can't find easily.