Hugo has quietly become one of the most trusted platforms for fast, lightweight websites. It is especially popular with developers and teams who care about speed, control, and clean builds.

At the time of publishing, it is estimated that 14,435 websites use Hugo, which makes it a meaningful part of the modern web stack.

This benchmarking study, by SALT.agency, looks at how Hugo websites perform from a technical SEO point of view, based on a sample of 5,000 Hugo site URLs. The aim is to give you a realistic baseline for what “normal” for SEO looks like in 2026, and what good looks like if you want to be ahead of the curve.

In this study, we focused on three areas that often shape how well a site performs in search and AI driven discovery:

  • Canonical tags on Hugo
  • Hreflang on Hugo
  • PageSpeed Insights benchmark KPIs for Hugo.

Canonical tags on Hugo

Canonical implementation is mixed across the sample. The main pattern is simple. A lot of Hugo URLs do not include a canonical tag at all, which in itself isn’t an issue, but can be if there are issues with duplicate content messaging and value propositions across various pages.

Metric Count % of Sample
Self referencing canonical 1,837 36.74%
Canonical missing 2,675 53.50%
Canonical present and points elsewhere 417 8.34%
Canonicalised in indexability checks 405 8.10%
Crawlable baseline for canonical checks 4,929 98.58%
Not crawlable and excluded 71 1.42%

What this tells us about Hugo websites in 2026:

  • Missing canonicals are very common across Hugo builds
  • This does not always lead to ranking issues, but it can make URL selection less predictable
  • Self-referencing canonicals remain a strong default for most sites, especially across blog and content templates.

Hreflang on Hugo

Most URLs in the study do not use hreflang.

That is completely fine if the site targets a single country and a single language. But for multilingual sites, this benchmark suggests there is a lot of incomplete coverage.

Metric Count % of Sample
Missing hreflang 4,548 90.96%
Missing x default 410 8.20%
Missing self reference 31 0.62%
Invalid language or region code 5 0.10%
Not crawlable and excluded 114 2.28%
Missing hreflang 4,548 90.96%

What this tells us about Hugo sites targeting international audiences in 2026:

  • Most Hugo sites are not running international targeting or are not implementing hreflang
  • The most common incomplete setup is missing x default
  • Errors like invalid language or region codes exist, but they are rare in this sample.

PageSpeed Insights benchmark KPIs for Hugo

If there is one area where Hugo tends to shine, it is performance, and the the data backs this up. Most scored URLs sit in the top performance range on mobile, which is where real users often feel problems first.

Performance band breakdown

Performance band % of scored URLs
0 to 49 1.10%
50 to 89 34.93%
90 to 100 63.97%

PSI KPI benchmarks for mobile

KPI Count
Total scored URLs 4,727
Average performance score 88.97
Median performance score 94

SEO priorities if you run a Hugo site

If you are running a Hugo site and want to improve technical SEO results without overcomplicating things, the best approach is to focus on consistency. Hugo makes it easy to generate pages at scale, but that also means small template decisions can affect hundreds or thousands of URLs.

Hugo often produces fast pages, but performance issues still creep in over time. The usual reasons are not Hugo itself.

They come from what is added on top. Marketing scripts, tag managers, embedded media, cookie consent tools, third-party marketing tools, analytics plugins, and large uncompressed images can all pull scores down, especially on mobile.

The sites that stay fast are normally the ones that monitor performance across multiple key templates and keep an eye on real user experience, not only lab scores. A small amount of regular performance upkeep will keep your Hugo build feeling lightweight, which is one of the main reasons people choose it in the first place.

See how your site compares. Get a technical SEO audit based on the same benchmarks used in this study. Get in touch.