Global SEO, by one definition, is doing SEO in markets other than English. There are several advantages and also some serious disadvantages. First of all most of the world doesn’t have English as a first language. Whilst their economies are smaller than the US/British/Canada/Australia block they are growing and will in time be larger. So there is a lot of money on the table.
On the other hand these people don’t primarily search in English and you don’t read or write their language, so whilst you may be able to get compelling copy written or translated it may well be useless for SEO.
That makes it a lot harder when you try to reach customers in Bangladesh and Bulgaria. However you have one large advantage compared to doing SEO in the English world.
Google has 20,000 engineers and they speak English. Hummingbird, the google update, that brought in sentiment analysis and hurt a lot of old fashioned SEO doesn’t appear to work in other languages. The resources required to replicate Hummingbird into 200+ other languages aren’t being devoted to it probably because of the lower ROI than alternative algorithm improvements for the English market.
What does this mean? This means that you can be working the traditional approach of identifying high volume keywords with low competition and using them to drive your business.
In the rest of this post I’ll describe how to find those keywords.
Don’t Use Machine Translation for Global SEO
First off; machine translation sucks. There are lots of reasons why this is so but for SEO purposes one of the key ones is that you’ll probably use Google Translate. And so does everyone else. If you use it much of your time doing Global SEO keyword research will end up improving Google translate and make Global Hummingbird a bit closer. The other reason of course is that many of the suggestions that Google offers are plain wrong. Google Translate, as much as anything, captures how a language is written, not how it is spoken. Search terms are a verbal expression of intent!
Word Context is Critical
As important as the translation of the words is the context in which they are used. When on business trips in Indonesia and Thailand I have often been forced to rely on Google Translate . The results of sentence translation have been comical when there has not been a total failure. Far better to rely on the frequently used words and phrases where there is a large corpus of evidence to support machine translation.
So you lose the slang, the nuance, how people thing and speak everything that helps you find great keywords. Ands that means at early stages that when looking at a foreign market that the Google keyword tool is crap and the Global Market Finder tool is even worse.
So what do you do?
Start with One Word
You start with one word. Look it up in an online English-“Foreign Language” dictionary and then go to the google search engine in your target country. This might be Google.pl for poland or Google.com.my in Malaysia.
Then you look at the results. The obvious problem is that they are all in a foreign language…. Oh dear! Don’t reach for Google Translate. Have a look at the images, either on page, or the image version of the search. Does this capture what you are looking for?
Often I will do a search for the term on my own google search page and also in the locale based search page. That way I can see how the search term is interpreted differently.
Find Synonyms and Validate
That may give you a result or it may not. Then you look for synonyms in the local language. A local language thesaurus is great for this task. Then you repeat the process for the synonyms.
What you do is to work through the synonyms slowly building up a corpus of words that you have validated as being appropriate to the target market.
Then you start adding modifiers onto your validated keywords. And you validate these exactly the same. Here though you have a slightly different problem. When you test the keyword + modifier you can end up with a page of results. Are they any good?
Does Search Volume Support it?
The best quick indicator here is the number of search results. So for example when I search “cheap car hire” I get 40 million results. That looks good. When I search for “miserly car hire” I get 90,000. So if you get no results or a low number the keyword is probably no good. A rule of thumb is that if there are less than a million results your keyword phrase is not regularly used.
Start Keyword Mining
Now you have a firm foundation of good keywords for Global SEO it’s time to be more traditional and start mining wikipedia and competitors websites for additional keywords. Whilst you may not be able to read or understand anything on the page a useful heuristic is to look for
- Bold
- Highlights
- Links
- Metadata
Suck them all up and add them to your list of language specific keywords. Make sure that as you do this you are also picking up slang usage. So for users may refer to a can of ‘pop’ ‘soda’ or ‘soda-pop’ include all this into your list of keywords
Then it’s getting back to the tried and tested tools of keyword research
- Keywordtool.io
- Semrush.com
- Advanced Google Queries
Once you’ve done all this dump the data into your excel file (machine translate into English so that you are able to keep track of them all) and put it into the Google keyword tool looking for local search volume in your target country.
Bang!
You should have a number of good keywords that you can use.
Translate using a Native
That’s the research done. Now you need to get your content translated by a NATIVE. not a fluent speaker, by a NATIVE making sure that they are focusing on the keywords that you give them for each piece.
And soon traffic will be flowing through to you from Google search.
The first time you do this it is time consuming, but that is as much due to learning and refining a new process. Once you have got it working smoothly it will be another effective tool in your arsenal.
Thoughts on process
First of all thanks to Eli Schwartz at Survey Monkey who inspired this post at a lecture at MAGiC
Whilst I’ve done a fair amount of SEO over the years in APAC as much as anything the key thing is to build yourself a nice, easily repeatable process. This can significantly increase the time cost the first time you do it but in future you get a lot of speed and quality improvements. Most critically you can scale. You can pass it to someone else and say you do it like this. Then you can check to see if the process is followed. That is where you get the big wins.
We help companies improve their sales and marketing processes by squeezing out speed and cost to make more sales with less time and money. If you are interested to find out more please use the contact firm below
