7 Tips to optimize content for multilanguage search
One Web site does not serve all needs. Although the online business community does understand the value of creating and maintaining multiple brands (which nonetheless remains a controversial practice for reasons I cannot into here), our vision of branding is still pretty limited. It’s not like you can walk into a mall and ask for the latest Xenite.Org feature article. You have to go to Xenite to get that content. Some online marketers do syndicate their content to vetted markets. I’m not talking about link spammers who submit articles to free ezine directories. I’m talking about major sites that distribute their feeds across the Web. With targeted distribution, you place your feed on partner sites that address the audience your demographic research tells you to contact.
Latino writer Reuben Navarrette, Jr. drove the point home in a commentary about how John McCain reaches out to Hispanic Americans better than Baracak Obama. As a Hispanic who has been raised like an Anglo I am intensely aware of the cultural differences between other ethnic groups and the Hispanic communities in our country (and note that I used the plural form because is not just one Hispanic culture among us).
The oldest European families in this country came from Spain, France, and Portugal in the 1500s — decades before the Pilgrims set foot in New England. The oldest city in the United States, St. Augustine, FL, was founded by the Spanish on August 28, 1565. Hispanic Americans are not just people who wandered across the border from Mexico looking for better jobs; we are a diverse people of many ancestries, languages, and customs.
Just as the Hispanic cultures that have helped shape the United States are becoming increasingly important to our political landscape, they have also arrived on the economic scene in a big way. And that economic impact is being felt on the Internet. But here’s the problem for the SEO community: most of us (including me) don’t speak or write in Spanish.
We’re treading water in more than one seas because all the seas are connected by one huge ocean. And the multilingual seas include more than just Spanish. This blog, for example, receives a fair amount of traffic from people whose browsers are set for Spanish, Deutsch (German), Dutch, French, Italian, Russian, and Chinese. My personal Web sites receive traffic from more than 100 nations around the world.
Through the years I’ve received requests from people to translate my work into other languages: Polish, French, German, Italian, Portuguese, Czech, Hebrew, and more. We recently approved a translation request for some articles from SEO Theory. And we have also experimented with providing automatic translation for blogs.
The translation always works best when you work with someone who is bilingual enough to understand idiom in both languages and regions. If you don’t speak the other language yourself, you’re left at the mercy of the translator. I’ve had two Web sites translated into Polish. A friend of mine who reads Polish told me the first translation was well done but the second translation seemed flat. The translator struggled with many of the expressions I personally use.
When I worked with a Hebrew translator a few years ago, he pointed out that much of the humor in my articles was lost in translation because there were no comparable expressions in Hebrew. I wrote an article for Tolkien fandom called “Celeborn Unplugged”, which addressed many of the popular ideas about a relatively minor character in The Lord of the Rings. I modeled the title on the popular MTV show format where featured artists performed for live audiences in a very small, intimate setting (e.g., “Eric Clapton Unplugged”, “Mariah Carey Unplugged”, etc.).
Fortunately for that essay, the MTV Unplugged shows were popular in Israel and so the translator had a common reference to use. But for another essay, which I titled “Is your canon on the loose?” (in which I was discussing how canonical discussions in Tolkien circles quickly break down because of differences in cited sources), there was no similar expression.
In English we have an expression, “That person is a loose cannon on the deck”, meaning that someone is so out of control they will cause damage to a situation. The metaphor clearly comes down to us from the days when ships carried cannon that occasionally broke loose during storms or battles. In American English pronunciation, both “cannon” (a weapon) and “canon” (the authoritative body of works for an author or group) are pronounced exactly the same way. The pun plays on the similarity between the unintentional harm that can be caused by cannon rolling across decks and canonical discussions that do not use the same acknowledged sources.
The translator suggested we use a title that, translated back to English, meant “Choir of a thousand voices”. Although the humorous aspect of the title was lost, the translator was able to preserve the title’s metaphorical function. That is, he found a suitable metaphor to substitute for my metaphor, thus preserving the undertone of ambiguity that I had wanted to convey with the title.
If your Web site is being visited by people whose native language is different from your own, you have to ask how they found your site. Some of those referrals come from non-English search tools. Those search tools may be Google.de and de.yahoo.com or they may be completely native search engines (e.g., in Spanish a search engine is called a “buscador”).
I can find myself on a Spanish search engine in part because of the Spanish-language content that I have created (with help) or that has been created about me. Even SEO theory appears in Spanish search.
People can search for brand names across search engines, so brand-based search works very much in other languages the way it does in English. The better-known the brand, the more likely the brand will be searched for (and the more likely that relevant copy will be found for the brand). But conceptual search works differently and that is where the content we create falls apart.
You can use any online translation tool to convert your pages to other languages, but what you end up with may look like complete gibberish to native readers in those languages. They won’t search for full expressions the way you think they might. They certainly won’t perform grammatical searches the way we do in English. Automation like Google’s Universal Search, which claims to direct non-English searchers to English-documents if there are too few native language documents, has to make some assumptions. Hence, searchers won’t necessarily find what they are looking for.
But we as optimizers can find ways to help provide foreign language visitors with appropriate cues. And these techniques work for any language, not just English. Hence, Web sites written in Bulgarian can make themselves available to English-language search or German-language search. Since most Web documents are written in English, however, it may not be as easy to cross-lingually optimize non-English documents for English search as for other non-English searches.
Cross-language optimization tip 1: Create language-specific pages on your sites
When one of my books was translated into Spanish, the translation team collaborated with me on writing a press release in Spanish. So you can find my site if you search for Parma Endorion en Espaol, nota de prensa.
How does that press release look to a Spanish-language reader? Not so good, given that the document structure doesn’t specific a language or doc set (the Web was so much more forgiving of my ignorance in those days).
Corollary to Tip 1: Use correct HTML markup to serve the content in the right language, including fonts and document specifications. You cannot afford to be sloppy.
Cross-language optimization tip 2: Build language-specific blurbs into your copy
Keeping in mind the Corollary for Tip 1, you may nonetheless be able to create indexable blurbs to embed in your content. Remember the SEO Method?. You experiment, evaluate, and adjust.
For cross-language optimization, you might want to add one step: Research, Experiment, Evaluate, and Adjust.
Now, it’s true that you should be doing keyword research before you start optimizing a document, but when you’re dabbling with languages you know nothing about you cannot take those intuitive short cuts you become used to in your native language. You should search the Web for documents that refer to your site in other languages. Try to identify the expressions people use to describe your site. Then look at your server logs to determine if people use similar foreign language expressions to find your content.
If you have access to keyword research tools in foreign languages, use those, too. But remember that you’re building a foundation of content that will expand outward. Your intuitive knowledge of English keywords won’t help you as much as you might think in non-English languages (and the reverse is true as well).
Not only do people search in many different ways in English, they also search in different ways in other languages. You won’t easily find one-to-one correlations between your native language assumptions and other people’s native language assumptions. They will use different metaphors, different grammars, and sometimes very oblique concepts to identify content like yours.
Use your blurbs to introduce people to your site and suggest translation tools that may help them. Understand that the translation tools could do more harm than good to your relationship with people using other languages.
Cross-language optimization tip 3: Build links to your content in other languages
How easily is this done? It takes effort. There are some non-English languages with robust search technologies (Russian and Chinese come to mind). There are some non-English languages for which search tools are virtually non-existant. Some non-English languages have to rely on Google, Live, and Yahoo! for their search. Some non-English languages are still relying on a multitude of Web directories.
Your ability to develop linking resources in other languages will depend greatly on your resourcefulness and ingenuity. It won’t hurt to have friends who optimize in those languages. And though I don’t recommend you participate in interlingual link farms, you might be able to collaborate with some people to build multi-lingual resources that act as switching stations between languages.
These interlingual link hubs would have to host articles in multiple languages, provide automated translation, and use unified site structures (navigation, HTML sitemaps, cross-promotional links, and XML sitemaps) to bind everything together, and attract natural links in all the various languages. You can provide translated resource pages.
I would not want to participate in a free-for-all community where everyone gets their own sub-domain or Web page; rather, I would want to participate in a team-managed project where admission was limited. That kind of Web site will be easier to protect against unscrupulous manipulation. So even if a few spammers get together to build an interlingual hub, that would be better than someone just throwing up an interlingual social media site and letting the world forces brutalize search engines in all languages.
Cross-language optimization tip 4: Create content for other languages
If you can work with someone who is able to translate at least a few articles for you, do it. Offer to help them improve their English-specific optimization. Provide some value in exchange for value.
Get your articles posted on the non-spammiest sites you can find in the other language. Build your visibility. If people see you are contributing to the document base in their own language, they will be more inclined to look at your native language documents — and to help you promote your content in their own language.
The English-language Web dominates the Internet today but that won’t always be the case. Several other languages (including Chinese, Russian, Spanish, French, and Deutsch) have the potential to become major Web influencers because of their large populations and because of the extensive literatures they provide. A great deal of original scientific and technical literature is published in all of these languages, which are also all used in more than one country.
We may eventually find the Web is not dominated by one language but by an oligarchy of several very distinct languages, each with its own grammar, lexicon, and idiom — each further sub-divided by region, country, and even generation.
If you have the resources to develop whole sites in other languages, do it. Make sure you know which character sets work for those languages. Make sure you don’t end up looking like a fool because your translation stinks.
Cross-language optimization tip 5: Use simple sentence structures wherever possible
I have found that short sentences work better than long sentences. I have found that the fewer metaphors and similes I use, the easier it is for automated tools to translate what I write. The more complex and convoluted a sentence is (the more punctuation you use), the more likely the translation tools create gibberish.
You are not patronizing your readers if you write short, simple sentence that reach out to people in other languages. You can use abbreviated writing styles to help make your copy seem less rigid. For example, you can practice using bullet points. You can use lists. You can write “news copy”, which only uses 1 or 2 sentences for paragraphs. You can use more graphics with short captions.
Wherever your native language audience will welcome short sentences, use them. Then find ways to shape your content so that you are able to use short sentences in more places.
For someone like me, that is very hard to do. I like to write long sentences. I like long paragraphs. But I have found I reach international audiences better with short sentences in my copy.
Cross-language optimization tip 6: Include translations in your copy
I you write an article that is intended for two specific audiences, such as an English and French audience, then you can embed copy in both languages in the article. I often did this when corresponding with people who worked with foreign languages.
I wrote my copy in English, I used a translation tool to convert the message to the other language, and then I embedded both versions in the message. At least that way the other person — who might know more about English than I know about their language — could see my point in my original language. Maybe they could work with a better translation tool.
Cross-language optimization tip 7: Translate each sentence twice
I normally prefer to translate one sentence at a time. I translate it from English to the other language. Then I translate from the other language back to English.
When my retranslated English sentence makes sense, I move on to the next.
This method does not guarantee that someone will understand you correctly in another language. What it does do, however, is teach you to use more precise language that the translation tool can handle better. You may find yourself using more sentences to explain a simple concept.
Nonetheless, taking small steps together can help you cover a lot of ground. Learning to use those expressions you find associated with your site in other languages will help you optimize for search in multiple languages. And once you have mastered that skill, you can begin to build brand value in those languages — making it easier for you, because people will focus their searches on your brand.
www.seo-theory.com
published @ September 23, 2008