Source Attribution Quality: Showing Where Your Information Came From

The previous post dealt with whether your claims can be supported. This one looks at something slightly different: how clearly the page shows the support. You can have solid sources and still attribute them so loosely that nobody can follow the trail.

Weak versus good attribution

Weak	Better
"Studies show..."	"A 2023 survey of 1,200 UK small businesses by the Federation of Small Businesses found..."
"Experts agree..."	"Google's documentation on canonical URLs says..."
"According to research, customers prefer..."	"In our 2025 customer survey (n=340), 62 percent said they preferred..."
"Source: internet"	A named publisher with a link
A bare link with the word "here"	A link with the publisher and title as anchor text

The pattern is simple: who, what, when, and where to check.

The four parts of a good attribution

1. Who originated it. The organisation, author or dataset, not the blog that repeated it.
2. What exactly they found or said. Reported accurately, with the key number and its unit.
3. When. Year at least. Old data should say it's old.
4. Where to verify. A working link or a reference to the document.

Step 1: find vague attributions

Search your pages for phrases like "studies show", "research suggests", "experts say", "it's well known", "statistics prove". Each one marks a place where the trail is missing.

Step 2: trace to the original

Follow the claim back. Many pieces of data travel through five blog posts before reaching a source. Find the first one: the government dataset, the academic paper, the company report. If you can't find an original, consider whether the claim deserves a place on your page.

Step 3: rewrite the attribution

Place the credit where the claim is, usually in the same sentence, and keep it short. If you quote someone, use their name and role. For longer material, add a "Sources" list at the end with full titles, publishers, dates and links.

Step 4: attribute your own work properly

Original data is one of your best assets. Make it easy to cite. Describe how you collected it, how many data points there were, the time period, and any limitations. Give the study its own page with a stable URL, and put the date on it. Other writers, and AI systems, will find that easier to refer to than a number buried in a blog post.

Step 5: keep attribution consistent

Pick a style and use it throughout: inline credit for key claims, a sources list for the rest. Don't leave some pages carefully referenced and others bare.

Step 6: check for ethical and legal issues

• Don't copy long passages from other sites. Paraphrase and credit, or quote briefly with proper marks.
• Respect licences on images and datasets.
• Don't imply an endorsement that doesn't exist.

Step 7: monitor your outbound links

Broken source links defeat the point. Check them periodically with your link checker, and see the guide on broken links for how.

Common mistakes

• Attributing to a famous name when the actual data came from someone smaller.
• Citing an old figure as if it were current.
• Using a source's headline number while ignoring its caveats.
• Linking to the homepage of a publisher rather than the specific page.

Good attribution costs a few words per claim and gives the page a firm footing.

Further reading: how Google treats outbound links and Google's guidance on creating helpful, reliable content are good places to check the details straight from the source.