F is for Facebook

What is Google Suggest? You’re certain have seen it, but may not know it by name. Put in simple terms, Google Suggest is the auto-complete functionality you see when you type a query into the Google search box. As you type, Google attempts to suggest possible query phrases to you. As soon as you begin typing, the system displays up to ten suggestions and, as you continue to type, the system refines the suggestions.

Google Suggest exampleThe suggestions you see are drawn from Google’s knowledge of the most popular search queries. In the screen capture on the right, you can see the results from typing “ap” into Google.com.

For the purposes of this article, Google Suggest is interesting for two reasons: First, with the arrival of Google Instant, predictive search becomes very important to all. Google Suggest gives us some insight into what we can expect when Google Instant makes its appearance here in Asian markets. Second, as Google Suggest is tuned to reflect local search popularity, it provides us with a way to gain some insight into variations between local search audiences. For those of us working in Asia-Pac, this is particularly valuable as many of our clients’ target markets cross national boundaries.


Google Instant asia pacificLooking at Google Instant, we can see several improvements in the Google Suggest functionality. Compare the screenshot at left with the one above. Both show the same two letter search (“ap”), but as you can see at left, Google Instant provides true auto-complete right inside the search box. You will also note that the suggestions display is limited in number to only five. The Google interface has been improved as well, and the suggestions no longer interfere with the first search result. At left, you can clearly see the first search result (Apple), while in the screenshot above you will note the top results are obscured. The predictive auto-complete and the improved interface will have a particularly powerful impact on user behavior, as now the user is both being actively guided in query formulation by Google while at the same time being encouraged to click on the top results associated with that query.


Perhaps more interesting, at least to those of us who work with search marketing, are the insights we can gain into the regional variations in Google query patterns. Google has, over the course of the last year, continued to refine the localization of Google Suggest.

We thought it might be interesting to run a little survey of regional variations, so we went to each of the Google country-specific search engines and tried out Google Suggest. We surveyed Google in six countries in the region and just for kicks, we checked Google Hong Kong in both English and Chinese to see if their were any differences. The survey covers the following:

To run our little experiment, we entered each of the letters A through Z and grabbed the first suggestion for each. We put it all together in a spreadsheet; you can see the results in the table below.

The ABCs of Google Suggest in Asia

Click on the chart to view a larger size.


The results for Google Suggest across Asia show varying degrees of similarity. Of all 26 letters of the alphabet, only one gives the same result across engines. The letter “f” pulls up the same suggestion across the board: Facebook. Solid testament to the strength of the Facebook brand in Asia. Further evidence of the brand’s strength is found in the results for the letter “w,” where “www.facebook.com” is the most popular query in three of the seven markets.

Only two other brands manage to achieve a majority result in any of the other categories: “ebay” is the most popular query for the letter “e” in all markets except Indonesia and Thailand. “youtube” also manages to dominate the “y” query in all markets except Malaysia and Indonesia, where “yahoo” was the top result.

Aside from the three examples cited above, there is a moderate to high degree of variance in the results set. The letters “i,” “l,” “r,” “t,” and “v” share no common results, with country-specific variations often reflecting either local brands, like “longdo” (a dictionary popular in Thailand), “lta” (the Land Transit Authority in Singapore), and “lelong” (an online shopping site in Malaysia). Other variations come from topical news stories that were hot at the time we did the research. The result for the letter “v” in Indonesia being a classic example; the result, “video ariel dan cut tari,” reflects search volume relating to a semi-scandalous sex video featuring two local celebrities.


For search marketers, the disparities in the results sets emphasize the need to be sensitive to the regional variations that exist in Asia. If your clients are concerned about markets outside their immediate country, then it is essential you take a look at how the results vary among the countries.

The second lesson we take away concerns synonyms and homonyms: when selecting related terms for optimization, it is important that you assess the popularity of the variations in order that you select the optimal choices — and you don’t miss any opportunities. You can also see in the chart an interesting result in Thailand, for example, where the most popular query for the letter “u” is “utube.”

Third, looking forward to the arrival of Google Instant in Asia, as we highlighted in a post last week, the predictive query feature is going to be very influential and you will need to do your research to see what users are being shown as they type in your target queries. The Google Instant results are visually more compelling and will have an impact on pushing users into narrower search patterns, focused on popular terms.

We hope you enjoyed this little exercise. We would be remiss not to give credit to FastCompany for the inspiration for this piece, with their article, Google Instant Redefines Your ABCs.