Location-based searching adapts a search to your geographic location. This post analyses Yahoo Mobile Search and China’s booming mobile industry.
Local intent and concept
A large amount of searches have local intent even if they do not have locations included within a search query. “Chinese cuisine [city]”, for instance, suggests that the searcher wants Chinese food places in [city]. “Weather”, conversely, does not contain a city but it is unlikely a searcher wants weather information for North Korea. Search engines must calculate if all queries have local intent.
Yahoo! Search for mobile delivers geo-specific content by the following 3 steps:
- Analyses the concept and the intent of a query;
- Search execution plan is produced which optimises the concept and intent of a query. This query is mixed with various categories, including, for example, web, news, photos and user-generated content, such as, Yahoo Answers;
- Search results are brought together from various categories and reorganised in a manner appropriate to a query. This reorganisation means a blended SERP is constructed.
Why are mobile phones popular in China?
The Chinese government has supported the development of mobile phones by implementing the State Council Policy of 1999 entitled State Affairs Development File No. -5 (MII, 2005). Government legislation and the high number of people and manufacturers means that mobile phones are popular in China — the largest country of people with mobile phones.
Of China’s population of 1,349,585,838 persons there are 1,150,000,000 active subscription service users. This means 85.21% of the population have mobile phones. In terms of mobile devices China is the world’s biggest market.
Mobiles and location
Since mobile devices are packed with various technologies, for example GPS and WiFi, your location is more precise compared to, say, desktop computers. Exact locations are good for local searchers. Search engines, however, need to calculate if a searcher has or has not got local intent within every query. This area of technology is still developing. But marketers need to know about local searches in order to create local SEO campaigns, for instance.
Posted by Gerald Murphy
- Chang, C. Wang, F. and Fu, H. (2009) A strategic analysis of the mobile telephone industry in Mainland China. Journal of Manufacturing Technology Management. 20(4) pp. 489–499
- Lu, Y. Peng, F. Wei, X. and Dumoulin, B. (2010) Personalize web search result with user’s location. SIGIR ’10 July 19–23. pp. 763 — 764
- Reuters. (2013) China’s mobile subscribers up 1.2 pct at 1.15 bln in March. [Online] [Accessed on 03rd October 2013]
- Yi, J. Maghoul, F. and Pedersen, J. (2008) Deciphering mobile search patterns: A study of Yahoo! Mobile search queries. World Wide Web Conference Committee (IW3C2). pp. 257–266
My last post, why do search engines store search history, looked at why search history is offered. This post examines the connection between privacy and search history.
- Right to be left alone
- Limited access to the self
- Control over personal information
Since searches can be a personal experience having control over your search history may invade privacy. If you delete something from your search history is this permanently deleted from all your records? This is a grey area for privacy.
Privacy written on keyboard keys. From g4ll4is
Allmer illustrated that privacy can be easily controlled by not sharing information and keeping it to yourself. Does a search history share your information with a search engine? Where else is this information stored? Is it matched to a personalised user profile?
Search history and cookies, for instance, collectively can build a lot of data, and thus breach privacy, because addresses and clicktrails can be monitored. We are a long way from finding out what information is stored and for what precise purpose. However, it is important to note that some entities are now standing up to large companies (for example Google and the European Union) which means that we are slowly making privacy progress.
Posted by Gerald Murphy
- Allmer, T. (2011) A critical contribution to theoretical foundations of privacy studies. Journal of Information, Communication Ethics in Society. 9(2) 83–101
- Frommkin, M. (2000) The death of privacy? Stanford Law Review. 52(5) pp. 1461–1543
- *Solove, D.J. (2002) Conceptualizing privacy. California Law Review. 90(4) pp. 1087–1155
The number of keywords used varies for a variety of reasons, for instance user intent and what our information needs are, but some studies have found interesting keyword statistics.
How many keywords are typed?
One study looked at the number of keywords used for website search engines, such as Wall Street Journal, and found that keywords are between 2.9 and 3.7 words. Another study found that we use, on average, 4 terms for normal searches but up to 6 terms for advanced searches. The exact number of keywords varies from search engine to search engine, let alone from person to person.
Another user behaviour study found that people generally type 2.7 terms or 13.6 characters. Further analysis also found that 17% have not been able to return to a webpage they once visited. This shows memory and search engines go hand in hand. Striking websites, however, are more likely to be remembered since it triggers an interest. Great content, catchy logos and effective designs make striking websites.
Looking specifically at search logs from file-sharing websites, equivalent to transactional queries on web search engines, one study found that keywords were usually between 5 and 6 words. User intent heavily influences the number of keywords we type.
Photo showing fewer keywords are more popular; thus competitive. Source.
There is less competition, from an SEO perspective, for longer search terms. Some scholarly articles have argued that long tail queries were associated with poor search success. It is not a good idea to only target long tail queries.
Importance of long tail searches*
- Easy to rank
- More conversational and natural
- Long tail searches work well with content marketing
- Especially if you are creating an informal content marketing campaign or using actual questions.
*Modified from (The Marketing People no date: online).
User behaviour and algorithms
One thorough study of Yahoo log files found that 88% of repeat clicks occurred if search engine rankings did not change whereas 47% clicked on new websites if rankings changed. Searchers’ like new websites by exploring a range of results on a search engine results page.
How many keywords do you type into a search engine? Tweet Gerald.
Posted by Gerald Murphy
- Graphic, Visualization, and Usability Center. GVU’s Tenth WWW User Survey. October 1998.
- Lau, E.P. and Goh, D.H.L. (2006) In search of query patterns: A case study of of university OPAC. Information Processing and Management. 42[Issue number missing] pp. 1316–1329
- Ruthven, I. (2003) Re-examining the potential effectiveness of interactive query expansion. Proceedings of the 26th Annual ACM International Conference on Research and Develpoment in Information Retrieval. New York: ACM Press pp. 213–220
- Teevan, J. Adar, E. Jones, R. and Potts, M. (2006) History repeats itself: Repeat queries in Yahoo’s logs. SIGIR. pp. 703 — 704
- The Marketing People. (no date) 6 reasons why long tail searches are important too [Online] [Accessed on 09th September 2013]
Posted in search
Tagged ranking, seo, yahoo
Different results for Bing and Google
Numerous factors explain why Bing gives different results than Google. Two of the main reasons are covered on this post.
- Search engine specific bots and indexes
- Bing and Google, two popular, crawler-based search engines have unique bots which ‘crawl’ and store keywords within their own index. Each index is therefore a search engine’s representation of how they see the web. So search engines “see” the web differently.
- Unique algorithms place different weight on different things. Most search engines are different because they have unique bots and indexes which rank websites in their own way.
- Bing powers Yahoo but their results are different. This is because each search engine uses different algorithms, ranking queries uniquely, so a search for “belfast” on Yahoo will be different from “belfast” using Bing.
- Personalisation adapts a system to a specific computer. Calculations predict likes and dislikes to contextualise search results. So a political searcher who searches for “brown” are more likely to get results on Gordon Brown whereas a landscape artist who searches for “brown” are more likely to get information on art products and paint colours.
- Although you might be using a global search engine, for example Google, your results are tailored to your geographic location. IP addresses play the largest role in doing this. Each IP address is similar to a house address. Search engines can find out where these IP addresses are located and personalise results based on this location. Results are therefore different for people who are in different countries.
Since personalised search can lead to relevant information, a feature of quality information, personalisation is not necessarily a negative feature.
SERP of different search engines. Source.
Search engines give different results because bots crawl and have different representations of the web. Search engine specific algorithms rank keywords differently so the same search is presented differently.
Posted by Gerald Murphy
- Ostrow, A. (2010) Bing Now Powers Yahoo Search. [Online] [Accessed on 18th August 2013]
Most popular search engines personalise, whereby past searches influence present and future, results. What are the main ways a search engine personalises results and why do they do this?
Personalisation: How and why?
Google decided to personalise search as a trail on 29th March 2004. So personalisation is not a new feature.
From an information retrieval viewpoint personalisation improves retrieval effectiveness by adjusting search results based on a searcher’s interests. These interests can be observed by mining short-term or long-term behaviours.
- Short term
- Search engine mines current search session.
- Long term
- Searches from many previous sessions are mined.
Personalisation ultimately occurs to everyone but it can be reduced by making smart use of technologies, for example, incognito mode or using a cookie removing facility.
Browsing & search history
Browsing history, search history and user’s explicitly declaring their interests to a search engine all help build interest profiles. Some studies do not take a broad overview of the various technologies. This means that smaller, more obscure, technologies are simply ignored, such as, cookies, for instance. It is possible for cookies to store keywords which can help search engines to personalise results.
Read more about other personalisation technologies.
Posted by Gerald Murphy
- Hines, M. (2004) Google takes searching personally. [Online] [Accessed on 13th August 2013]
- Shen, X. Tan, B. Zhai, C. (2005) Context-sensitive information retrieval using implicit feedback. SIGIR. pp. 43 — 50
- Shen, X. Tan, B. Zhai, C. (2006) Mining long-term search history to improve search accuracy. KDD. pp. 718 — 723
- Sontag, D. Collins-Thompson, K. Bennett, P.N. White, R.W. Dumais, S.T. and von Billerback, B. (2012) Probabilistic models for personalizing web search. WSDM. pp. 433 — 442
- White, R.W. Bennett, P.N. and Dumais, S.T. (2010) Predicting short-term interests using activity-based search context. CIKM. pp. 1009 — 1018
Several eye tracking studies found how we search through keywords. Higher positioned results are also the only results likely to be reviewed by most users. Along with the design and user interface the position of results heavily influences what we click and do not click.
Ranking of results matters
Some eye tracking studies seem to find that searchers concentrate on the top results and, mostly, ignore lower ranked websites. So the position of your website / result on search engines, for example being in the top four position on Bing or Google, does matter because the top results are the only results likely to be reviewed by the searcher.
Person searching for information. Source.
Search engines find information
One study found that users of search engines were likely to use Yahoo!, Bing and Google first for medical information rather than searching on specialised search engines. General and popular search engines are worth concentrating on because your highly ranked website on Bing, Google and Yahoo! is more likely to get more visitors. But it is also worth noting that less popular search engines, in particular Bing, get higher conversation rates for PPC compared to Google. So search engines themselves heavily influence user interactions because the design and user interface is different for every search engine.
Other eye tracking studies found that relevant results are read from top to bottom (i.e. position one followed by position two and three etc). If the top results were irrelevant searchers were likely to review more hits. Bad searches means bad results which results in the searcher reviewing more results until they find a relevant hit.
A good search experience goes hand-in-hand with an appropriate choice of keywords. So it is not just about getting your website to rank high but to also allow people to find your website for a range of keywords. It is also worth noting that eye tracking studies are not frequently employed because they are tedious, requiring extensive data, and they can be too narrow to generalise. However, eye tracking studies allow user behaviour to be better understood.
What way do we search for information?
We tend to focus on the top, usually the first four, results. If these results are irrelevant then we are likely to review more results. Ranked results means that we read top-down rather than bottom-up. Furthermore, choosing a website at a random position is unlikely to occur due to this user behaviour.
Posted by Gerald Murphy
- Bing. (no date) Bing Ads Drives Higher Conversion Rates and Lower CPAs for Agency Clients. [Online] [Accessed on 09th August 2013]
- Goldberg, J.H. Stimson, M.J. Lewenstein, M. Scott, N. and Wichansky, A.M. (2002) Eye Tracking in Web Search Tasks: Design Implications.
- Granka, L. Joachims, T. Gay, G. (2004) Eye-tracking analysis of user behaviour in WWW search. In Proceedings of the 27th Annual ACM SIGIR Conference on Research and Development in Information Retrieval. SIGIR ’04. ACM. New York: NY. pp. 478 — 479
- Guan, Z. and Cutrell, E. (2007) An eye tracking study of the effect of target rank of Web search. In CIII 2007, ACM. New York: NY. pp. 407 — 416
- Lorigo, L. et al. (2008) Eye tracking and online search: Lessons learned and challenges ahead. Journal of Americian Society of Information Science Technology. 59(7) pp. 1041 — 1052
Search history, provided by a search engine and produced by the searcher, has become a well-used feature that is easy to store because it does not use a lot of memory (Leung et al 2012:3065); however, this memory is used to personalise the searcher’s services (Anonymous 2005:8). Google, Bing and Yahoo! all collect search histories alongside paid services.
A Google user, theoretically, has control over their search history because they may delete it from their user account. But search history may be stored by other technologies, for instance, cookies.
Does Google personalise searches?
In short, yes! Fox (2007:24) identified what technologies are used with Google’s personalised search; they include:
- Previous click behaviour
- Search history
- Web history
- Use of other Google services
- Country restriction
All Google users, from 2009, automatically opt-in to Google’s Personalised Search allowing a users’ web history to be monitored (Horling and Kulick 2009: online).
So it is clear that there are several technologies which might influence one single search.
- Fox, V. (2007) ‘The anatomy of personalized Google results. Information Today. 24(11) pp. 24 — 24
- Horling, B. and Kulick, M. (2009) Personalized search for everyone. [Online] [Accessed on 11th Feb 2013] http://is.gd/BVwG11
- Leung, S.W. Yuen, S.Y. and Chow, C.K. (2012) ‘Parameter control system of evolutionary algorithm that is aided by the entire search history.’ Applied Soft Computing. 12(9) pp. 3063 — 3078