yara-hreflang-hero-1600x900.jpg

Yara International

Prioritising local sites in search results

yara-colour-logo.png

Hreflang Implementation

As a leader in the agricultural and chemicals sectors, Yara serves a wide range of markets across the world. With revenue reaching US$13.9bn in 2015, Yara is one of Norway’s biggest companies with a truly global reach.

Making Search Local

Building Blocks has worked with Yara to improve its digital offering for quite some time now, so we were enthusiastic when they asked us about putting ‘Hreflang’ in place across all of their  downstream websites.

Hreflang is a technical solution for the prioritisation of local sites in search results, where a site owner has many sites of similar content across different markets/regions. Users naturally want search engines to deliver the most useful and appropriate site for them. So if you are a Dutch user in the Netherlands and you search for Yara, both you as a user and Yara as a company are best served by the Yara Netherlands site being delivered prominently in the results. However, it is often the case that the English versions of pages rank higher.

Another problem mitigated by Hreflang is duplicate content harming search results and/or ranking. For instance a global enterprise like Yara has several markets in English; these sites have similar content apart from changes in addresses, currency, spelling of specific words etc. You understand that these sites are aimed at different markets but Google does not.

Despite you having gone to all the effort of creating content optimised for language and location, search engines may still present versions of pages that don’t match up. Hreflang is a way to indicate to the search engine that this is almost the same content but optimised for different users.

Custom Sitemap

To make Hreflang work for Yara and improve the user experience for their customers, our team assessed a variety of solutions before settling on a combined approach.

We built a custom Tridion storage extension to index pages into (the open source enterprise search platform) SOLR and a console application that runs intermittently, fetching live page data and building up-to-date Google Sitemaps for each country publication.

In order to get around sitemap size restrictions and avoid excessive publisher load we developed a way to make use of multiple sitemaps. The end result was a flexible, scalable, lightweight solution.