I spoke at Crawl Error 2021, a virtual SEO conference organised by Dan Taylor and Salt Agency, about how to make your international SEO strategy a success. If you want to check out my talk, you can see it on YouTube. I also wanted to summarise the main points here in a blog post. So here goes!
Why does it matter?
If you’re selling your products or promoting your services in multiple international markets, investing time in developing your international SEO strategy is essential. From a technical perspective, you need to make sure the search engines are able to understand which version of your content is relevant for each target country / language. And from a user perspective, your content has to match the search behaviour of the target market.
Localisation > translation
A central pillar of any successful international SEO strategy is the principle that localisation is greater than translation. Localisation is more than just simply translating the content from your primary market. When you do that, you might run in to some embarrassing mis-translations like this example on Charles Tyrwhitt. On their German website, “Menu” was translated as “Speisekarte”, which in German is the food menu you would get in a restaurant.
This was rectified pretty quickly, but it shows the potentially awkward pitfalls of one to one translation. Another, slightly more practical, comes from an eCommerce client specialising in warehouse and industry equipment. Rather than just directly translating the keyword strategy from their primary market to Polish, they worked with local experts to fully understand local search behaviour. For their pallet truck category page, rather than just focusing on the exact, formal translation of the keyword pallet truck, ‘wózek paletowy’, the localised keyword strategy identified other highly relevant keywords, including the more informal translation ‘paleciak’, which has three times the search volume of ‘wózek paletowy’.
So you need to create a truly local keyword strategy, based on the local search behaviour of your target market, and not just a translation of the keyword strategy from your primary market. This should then be used to create high quality content, using the keywords and topics relevant for the local market, matching the search intent of your target audience. A typical example would be a fashion shop creating content about ‘sneakers’ for their US website and trainers for their UK website.
International targeting and hreflang
The hreflang is one of the most important elements of a successful international SEO strategy. Your hreflang implementation has to be spot on if you want search engines to properly understand the relationship between your different international versions of content.
Some basic principles are of hreflang are:
- It can be implemented in the head of the page or in sitemap. I would usually recommend going with the head implementation. If you go for the sitemap implementation, it means your whole hreflang strategy has one single point of failure if something goes wrong with your sitemap.
- All existing language versions of the current URL, including the current URL, should be referenced in your hreflang. You can use Screaming Frog to find any missing return links – just make sure crawling of external URLs is allowed in your crawl settings.
- Language must be specified, while country is optional. For example you could use hreflang=”de” to target all German speakers, regardless of location or you could use hreflang=”de-ch” to target German speakers in Switzerland, but you couldn’t use hreflang=”ch” to target the location Switzerland, regardless of language (hreflang=”ch” would target the Chamorro language).
- You can specify a default version to be shown to users for whom their specific language version isn’t available using hreflang=”x-default”.
- Never specify non-indexable URLs in your hreflang. The hreflang attribute is a tool to help Google index and serve the correct content to your users based on their language preferences. So if your hreflang contains URLs that Google can’t show in the SERPs (i.e. non-200 status code URLs, URLs with a “noindex” tag, URLs that canonicalise to another URL), you are sending conflicting messages. Again, Screaming Frog can help you identify any non-indexable hreflang URLs.
Common hreflang errors: missing return tags
One of the most common hreflang errors is missing return links. With the hreflang attribute, if URL X links to URL Y, then URL Y must link back to URL X.
Otherwise the implementation is incorrect and even the information provided by the correctly implemented part of the hreflang may be disregarded by search engines. So, in the example below, the Austrian page is missing the hreflang attribute linking to the German page. Theoretically, this could lead to search engines disregarding the instructions given by the business for this page and its international equivalent on all three domains.
Common hreflang errors: incorrect language or country codes
Another common error is the use of the incorrect language or country codes. The hreflang uses the ISO 639-1 format for language and the ISO 3166-1 Alpha 2 for countries. A couple of other things to note is that regions are not supported, like the EU or an entire continent, and that sometimes the language code doesn’t match up to the country code, even in cases where the language is mainly only spoken in that one country. For example the language code for Danish is dk but the country code for Denmark is da.
One of the mix-ups I see most frequently is the use of “br-br” in the hreflang attribute by a business wanting to target Brazilian Portuguese speakers in Brazil. The correct code for this target market would actually be “pt-br” (“br-br” is Breton speakers in Brazil).
The other really common one is the use of “en-uk” to try and target English speakers in the United Kingdom. The correct code for this would be “en-gb” (“en-uk” is English speakers in Ukraine”).
But as we can see from the above examples, this is still something even big businesses like Nestle and Outfittery have to work on.
International targeting – IP based redirects
One important point about international targeting is that IP based redirects are not a good solution for sending users to the page for their language/location. Generally speaking, it’s not a good idea to take the user (or bot) somewhere that they haven’t asked to be taken. From the user’s perspective, this can be really frustrating. And for search engines, this can cause indexing issues, as Googlebot usually crawls from a US IP address, so if they can’t see your international content due to your IP based redirects, that content isn’t going to get indexed.
Which domain strategy fits my international strategy?
If you’re in a position where you can have influence over the domain strategy that will be selected, for example if your company is already planning a migration, or if you’re setting up a new site from scratch, it is worth considering which of the different domain strategies would be best for your international SEO efforts. The three main options are country code top level domains (ccTLD), a generic top level domain (gTLD) with subfolders or a generic top level domain with subdomains:
The advantages and disadvantages of each are as follows:
Country code top level domains
Generic top level domain with subfolders
Generic top level domain with subdomains
Generally, I wouldn’t recommend going with subdomains. There are no unique benefits to subdomains that you wouldn’t get with either ccTLDs or a gTLD subfolders, but all of the same disadvantages. So if it’s possible, I would go for the gTLD with subfolders or the subdomains.