An unscientific assessment of Baidu

Google has finally left the Chinese market for search, so I thought I would try the alternative, hence an unscientific assessment of Baidu. My trial is unscientific in nature and not particularly rigorous. I did what most consumers would have done and searched for myself.

I was quite open-minded about this, on the one hand Google has been killing the search market in Europe, nothing can touch it in the EU and they have made moderately successful forays into other sectors as well. I also know that Google is not all conquering. In fact the wheels start to come off the Google search wagon when you venture into areas with non-Roman languages such as Russian, Korean, Japanese and Chinese.

On the other hand, Robin Li over at Baidu is no slouch. Baidu is famous for is huge index and its continued appetite to crawl content whenever and wherever it can find it.

Baidu like its Korean counterpart Naver has also managed a successful social search product running a question-and-answer service like a better version of Yahoo! Answers – largely free of spam and a more middle-class range of participants provide highly relevant quality content.

It is also blatantly obvious that Baidu doesn’t care whether it attracts a potential English-speaking audience as the entire site apart from investor relations is in Chinese.

Methodology

Outcomes

I was expecting some divergence between Google and Baidu search engine results pages for a number of reasons. Google crawls an estimated 15 per cent of the total web, and Baidu is likely to crawl a slightly larger amount. That means that their search indexes are likely to be slightly different. Secondly, both will have started with slightly different algorithms and these will change over time with a experience of what users want. Finally, the results are usually ‘flavoured’ according to local market preferences such as language and local content.

I was a bit surprised at the level of divergence between Google and Baidu, which was great than I had seen between Google and Yahoo! in the past.

First of all flavouring. A comparison between the Japanese and Chinese versions of Baidu show a high degree of variance between the two versions of the Baidu search engine.

Baidu CN vs Baidu JP

Part of the reason for the difference may be due to Chinese regulations around permitted services, for instance an educational video of me by Econsultancy on YouTube is the top result on the Japanese site and a couple of twitter related hits come in at six and seven. The Japanese site skews much more toward video services than the Chinese site which picked up profile services Plaxo and Naymz.

Interestingly, the Chinese site picked up the re-direct URI for my blog (renaissancechambara.com), whereas neither the Japanese or the Chinese versions picked up my proper domain (renaissancechambara.jp) at all. Even when I clicked a few pages down.

Plotting Baidu China against Google Hong Kong produced an interesting diversity of the results.

Baidu CN vs Google HK

Their one point of correlation, my profile on Naymz. Again part of this may be because of my presence on services that don’t do business in China for instance YouTube and Twitter. Google rightly puts more weight and a consequently higher ranking on my Crunchbase and LinkedIn profiles than Plaxo which appears a couple of pages down on Google.

Baidu obviously puts much more emphasis on a historic redirect URI I have for my blog than the ‘real’ one and doesn’t seem to crawl the site in any great depth. I am guessing that this is because of its largely English language content.

Baidu JP vs Google JP

In Japan, the Baidu | Google comparison told a similar story. The Google flavouring between Hong Kong and Japanese versions wasn’t that great only showing differences at position five and lower on the page. Baidu Japan managed to pick up my last.fm profile and twitter profile, but didn’t pick up my blog or any professional information on the first page.

In conclusion, my unscientific assessment of Baidu has shown provides a great search experience for consumers. But I am uncertain how valuable it would be for people in a professional context, for instance researching foreigners with whom they may be doing business or finding foreign presentations. I can understand why Chinese scientific audiences would be concerned by the departure of Google.

I also suspect that optimising content to make it searchable on Baidu is different to the process that I would go through for Google or Yahoo!, but that would merit far more investigation before I could blog with any confidence about it. More Baidu related posts here.


Comments

2 responses to “An unscientific assessment of Baidu”