CitabilityFree

Source Attribution

Scores how well you attribute statistics and factual claims to their sources. Properly cited content is more trustworthy to AI engines.

Why It Matters for AI Visibility

AI engines treat sourced claims as more trustworthy than unsourced ones. When ChatGPT or Perplexity encounters a statistic on your page, the presence of attribution language -- "according to McKinsey" or "source: Gartner 2024 report" -- makes the AI more confident in the accuracy of that data and more willing to cite your page. Source attribution creates a trust chain. Your page cites a study, then the AI cites your page. The attribution makes the AI confident that the information has provenance, not just assertion. Without attribution, a claim like "Revenue grew 35%" is indistinguishable from opinion. With attribution -- "Revenue grew 35%, according to the company's 2024 annual report" -- it becomes a citable fact. Google AI Overviews evaluates source quality when selecting content for its summaries. Properly attributed claims signal that content is well-researched and evidence-based rather than opinion-driven. Pages where every statistic traces back to a named source consistently outperform pages full of unsourced numbers. Perplexity's citation system similarly favors content that demonstrates rigorous sourcing.

How We Score It

The analyzer first detects all statistics and data points on your page -- percentages, dollar amounts, multipliers, and specific counts. It then checks a 300-character window around each data point for attribution language. Attribution patterns detected include "according to," "source:," "reported by," "cited in," "per a study/report/survey," academic citation format like "(Smith et al., 2024)," "published in/by," and "data from." Each statistic is labeled as sourced or unsourced. Your score equals the attribution ratio multiplied by 10. If 100% of your statistics are attributed, you score 10. At 70%, you score 7. At 50%, you score 5. If no statistics are attributed, you score 0. If your page has no statistics at all, the score is also 0 -- you need data points to exist before attribution can be assessed.
See how your site scores on this factorAnalyze My Site

How to Improve

  • 1

    Add "according to" before or after every statistic

    "According to McKinsey's 2024 report, 78% of executives prioritize digital transformation." This is the most commonly detected attribution pattern. Place the source name immediately adjacent to the data point so it falls within the analyzer's detection window.

  • 2

    Use academic citation format for formal content

    The parenthetical author-year format -- (Smith et al., 2024) -- is explicitly detected by the analyzer. This works well for research-heavy content, whitepapers, and technical documentation where formal citation style is expected.

  • 3

    Include "source:" labels near data points

    "Source: Bureau of Labor Statistics, Q3 2024" placed immediately after a statistic is a clean, unambiguous attribution the analyzer detects. This format works well in data tables, infographics, and sidebar callouts.

  • 4

    Add "data from" for dataset references

    "Data from the World Bank shows GDP growth of 3.2%" provides clear provenance for each number. The "data from" pattern is explicitly detected and works naturally in content that references organizational datasets or research databases.

  • 5

    Target 80% or higher attribution ratio

    The analyzer recommends improvements when less than 80% of your claims are sourced. Audit every statistic on the page and add attribution to the unsourced ones. Even self-attribution ("according to our 2024 survey of 500 users") counts toward the ratio.

Before & After

Before
SaaS companies see 35% higher retention rates. The average
customer lifetime value is $1,200. Companies that invest in
onboarding see 3x faster time-to-value.

Statistics: 3 detected, 0 attributed. Score: 0.
After
According to Bain & Company, SaaS companies see 35% higher
retention rates. The average customer lifetime value is $1,200
(source: ProfitWell 2024 Benchmark Report). Per a Totango study,
companies that invest in onboarding see 3x faster time-to-value.

Statistics: 3 detected, 3 attributed (100%). Score: 10.

Frequently Asked Questions

What if my statistics are from my own original research?

Use self-attribution language like "our research shows" or "according to our 2024 survey of 500 users." The analyzer detects "according to" and similar patterns regardless of whether the source is internal or external. Self-attributed data counts toward your score.

Does linking to the source count as attribution?

The analyzer checks for text-based attribution patterns, not hyperlinks. A linked statistic without text attribution will not be detected. Having both a text attribution and a link is ideal, but the text pattern is what this factor scores.

What if my page has no statistics at all?

The score will be 0. This factor requires statistics to exist before it can assess attribution. Improve your Statistics and Data Points factor first by adding specific numbers, then add attribution to each data point for this factor.

Related Factors

Authority

External Citations & Sources

Citability

Statistics & Data Points

Citability

Original Research Signals

Check Your GEO Score

Run a free analysis on your website and see how you score across all 52 factors.

Analyze My Site