Moderated by Christine Churchill of Key Relevance. She welcomes people to the second day and introduces the speakers.
Dixon Jones from Receptional Internet Marketing. Will speak from the POV that people are considering going to EU market. Starts with some facts about EU and the world.. USA 300 M pop, 200M Internet users versus EU 800 pop and 300 M Internet users. 90 million more Internet users now than in US, with an anticipated additional half billion coming online fast. Just in UK there are 40M online. Over 10% of all ad spend in both UK and Sweden was online in 2006. This makes Internet advertising more developed as an industry tan the US. Poland is the 5th fastest growing country on the Internet. Worldwide 1.2B people. Choose your message, either my market is 5X bigger than I thought, or “OMG, third world countries are going to sell to my clients.” He suggests looking at it the first way. G
Goes over some GDP per capita, per CIA figures. Next chart lists countries by GDP growth rate. Starts with Azerbaijan, Angola,….China at 10,,,Usa is #139. The American strategy: attack is the best form of defense. “Take the battle to the enemy to protect the homeland.” Biggest problem is language, which may be a barrier mentally, but is a boon for us in the industry. Discusses the advantage of SEO in targeting other-than-English words for SEO. IE there are only 4 M pages for the word “Londres” versus 50M for London at G. The Greek word is less than 1M. If it wasn’t for the language barrier, SEO is considerably easier in other languages.
He says there is no real easy way to do international SEO. Recommends employing language-specific operators to take calls from overseas. Either do SEO yourself in foreign language, outsource locally, or use affiliate tactics. If you do try yourself, buy a TLD for the right country. Or Host the site in the tgt country. Track link building in a way that regionally themes the site. Use local press release services. Use a native translator. Minimize legal issues and deploying ground troops until you are confident of success in the market.
Jessica Bowman from Business.com. is on second day on job for them, used to be with Enterprise Car rentals. Will speak about multi-lingual SEO. She will not deny, “you are in for a ride.” KW research: need to brainstorm with a native speaker. Go to the competition. Tools are a challenge, but says Trellian’s Keyword Discovery is good, for example for breaking out UK English. Also says that Hitwise has a few languages, but probably not enough, in her opinion. Since there aren’t many tools out there, there has been a study that showed that the volume of search results in English tends to correspond with other languages. Andy will speak more to this.
It is best to start the copy in the foreign language, versus translating, although unfortunately that isn’t always and option. Not many translation companies know seo. Recommends learning the translation process inside and out, because the process breakdown is inevitable. Found that they used automated translation tools, which are fine for non-SEO focused. Also found they had 2 different sets of translation data, one sort of a glossary of terms, and the other a collection of sentences (“Memory Bank”). She found the problem was that these were out of sync. They needed to separate them. Once the translation come back: check it. She has seen error messages that said you “are not allowed to log on.” This was bad because they were telling the customer they had no right to login to their store, even though it was actually and logon password error.
Using wrong words. Found that sometimes more than one translator worked on the same doc, and they used different words/phrases referencing the same thing on the same page. Sometimes content comes back very long and doesn’t fit in the space. Factor this iteration into the cost, and make sure that the translator knows that this will occasionally get “kicked back.” Also translators sometimes add or embellish to content, which can lead to serious legal ramifications (I would guess especially in an industry like Pharma). Bottom line: do not assume it is good-to-go when it comes back from translators.
Final thoughts: Train the translator on SEO, using their own work as examples if possible. Send them useful information such as screenshots to see how much space there is. There may be character limitations for spacing and/or database fields, as another example. Remember to send corrections back to the translation company. Also, remember that translations confuse people, and painted a scenario where they wanted to change “car rental” to “car hire,” in UK English, and then to French. Found issues existed.
What she has learned: German text comes back 3x the length of English text. German grammar is a killer – sometimes merges words depending on the context, and there are two forms of the German language, one more formal and one less. UK English: treat it like a different language! British tend to be more verbose, and again you need a native speaker to review the copy. Spanish: consider which region it will be translated to. Use “North American” Spanish for US Spanish speakers. Spain Spanish is extremely different than any American Spanishes.
Michael Bonfils of SEMInternational. “10 Steps to Cracking South East Asia” Starts with SE Asia stats. 300 M Internet users, majority in China at about 118M. Japan 86M, South Korea at 34M. Indonesia 18M, Vietnam 10M, and more… Shows an Internet Population rate chart that is very interesting: SK 67% while China is only 9%. If China had the same penetration rate as Taiwan, they would be looking at 784M users!
Ten steps are: Understand audience; understands Asian domain names, hosting, search engines, translation, keywords, paid search, organic search, reporting and analytics, red tape. Step 1: Understanding Asian customer: often what is funny and creative over there, is not in other places. For example they have little pictures that “follow you around” during a search in Korea. Remember to think about that, what works in Western EU and US may not work there. Internet is more often used for researching, while buying is often done “downstairs.” In China approx 50% of all Internet usage comes from cafes. Often no coffee served, and not really a “social environment.” Shows a funny picture of a floating Internet café.
Step 2 know the major search engines. In China, for example Baidu has 62% of the market versus 25.3% Google and 10% Yahoo!. In Taiwan, 90% Yahoo!, 5% G. Japan, 55% Y Japan, 35% G, 5% MSN. In South Korea: “Naver” 63%, 14% “Daum,” 11% Yahoo! Korea. Shows a funny relationship chart taking off from Bruce Clay’s that he calls the “Bruce Lee Chinese search engine relationship chart. There “Sogou” provides most paid results to other engines.
Ste 3 Asian Domain name. make sure it is pronounceable. For example, Google not pronounceable in China, where there call it “Goo-Goo,” due to inability to do the L. Look for proper domain extensions. There are a variety: China: .cn or .com.cn. Korea: .kr or .co.kr. Japan: .jp or .co.jp. Taiwan: .tw com.tw. Hong Kong: .hk or .com.hk
Ste 4: Get hosting: Us hosting option gives slow access due to a gateway. It is more economical than hosting in Asia. If using Chinese hosting option, keep in mind that the larger the hosting company, the more politically regulated it is. Gives a list of china’s top web hosting companies.
Step 5: translate well. Gives a couple of examples, and agrees that you have to double/triple check. Even recommends using one trans.. Uses “Pepsi brings you back to life” example of translation that went to “Pepsi brings your ancestors back from the grave.” Ad:tech Shanghai brochure even had a mistake that was “cost per lead” translated as “cost per leadership.”
Step 6: Develop keywords. Same expression in English can be translated up to 15 different expressions across china, HK, and Taiwan. Local translation and research matters!
Step 7: Implement paid search. Would reco starting with this over SEO which is much tougher. G and Y! Works pretty much the same in Asia as in US in terms of paid results placement. Baidu works a little differently, including paid listings within “organic listings.” Sometimes will be in the center of the result, and you may get more traffic and conversions from this.
Step 8: develop organic search. Similar issues as us such as content, structure, Meta and kw placement. Link pop also a factor. Think local! There is a strong local company favoritism in China…try to use local sites and TLDs.
Step 9: understanding reporting. Shows some Baidu reports and how the daily spending reporting and keyword Report differ slightly. Also shows the much cleaner looking (IMO) Yahoo report. Step 9.5: Implement analytics. They happen to use Google Analytics since they were in the process of translating Urchin at the time of G’s purchase.
Step 10: Understand the red tape and unusual business practices. People impatient on the phone or expect payment. There is no API, no CPM reporting, and it is commonplace to find kickbacks and discounts. Quite normal to have someone knock on your door in China and say (as an example) “we are from Baidu and could use some extra $ to give more favorable results.” Poor payment methods…probably will need to wire to companies that you are looking to run campaigns with. Remember Political favoritism, and that they are partnered with Chine Government, which is a formidable competition. “Be not afraid of growing slowly, be afraid only of standing still” (Chinese proverb).
Barry Lloyd was unavailable due to being stuck in China. Andy Atkins-Kruger of WebCertain will replace him. Will be brief. Mentions blog: “Multilingual Search.” He starts by reminding people that SEO, Paid and other methods are closely related. Sows a chart of amazing growth that would happen if you applied NA adoption rates to the rest of the world. Prob of ranking+ internet audience+ market size must be measured. Uses example of “football boots” research (equivalent to football cleats in US) actually being most desirable as a target term in Spain, not the UK as one might have thought. Talks about different ways to operate overseas, similar to other speakers.
Recommended process: First: kw research then creation of glossary then translation then optimization. Feels that you do need a local domain, ideally. Things like local hosting and local links will come much easier. Beware of things such as duplication issues like same German content targeting Germany or Austria. Sometimes the wrong page can appear at the wrong time. Showed another study that showed that the Google results (number of pages), when compared to Yahoo suggestion tool (number of searches), showed an actual correlation in pattern. Looks at another example where there can be plurals and singular searches, as well as the use of prepositions in romance languages.
These do have an impact on the way results appear. People search with or without these prepositions, and you have to plan for both. Accents can also be a problem, but the search engines seem to have come up with a way to handle this. They do better with accents that have a distinct effect on meaning versus those that do not really affect the meaning. Also alternative characters. Also aggregation or not? Sometimes long words can be split into two, depending on how people are thinking when they search. Also a possible thing to look for are “declensions,” which many languages have. Ie: 16 different cases in Icelandic. Shows an example in Russian. SE’s don’t necessarily do well with non-Latin/romance languages, since algorithms are not written to cope with those issues. Do not use free online translation tools! Also gives a chart that shows a few of the top SE’s in other countries. Thanks everyone.