When stockbrokers who spend their day searching for financial information about different businesses type the word “Starbucks” into Google’s search box, chances are that they are more likely to be looking for stock price information than the closest place that they can get a mint mocha chip frappuccino.
When a city-dwelling college student, who likes to meet up with friends at new places all the time, using his cell phone to find and map out those places, types the word “Starbucks” into his phone’s browser, the first thing he wants to see is probably a map to the nearest Starbucks.
Can a search engine be smart enough to serve a stock price quote at the top of search results to the stockbroker, and a map to the college student, even if both are using handheld devices to connect to the Web?
Local One Box: Relevance and Informational Needs
Sometimes the most relevant and most popular page on the Web isn’t the best response to a query submitted at a search engine. Sometimes the best page to return to a searcher is the one most appropriate to their informational need.
How well can a search engine determine what the informational need of a searcher might be?
A new patent application from Google explores that topic and introduces such things as Local One Box results to show to searchers, as well as changing the order of different types of results that it might present to those search engine users
Can information from a search engine be presented to searchers in a way that is most likely relevant to them, by presenting search results ordered in groups, such as “web, image, blog, scholar, desktop, maps, news, video, and other sets”?
What does this mean for searchers, and for site owners?
The search engine might decide the groups of results to present based upon the language used in a query, whether there is some geographic relevance in a query and the quality of search results on the Web that exists for a query.
If the searcher is using a desktop computer or a handheld device to connect to the Web might play a role in results show, as well as individual and aggregated user data about the searcher, the device used to search with, and the query terms involved in the search.
Different Users, Different Experiences
The patent presents a few examples, which are interesting. I started my post with a simplified version of one of them, about the stockbroker and the college student.
Might a search engine be smart enough to give the student the results that he may want to see first, a map, and to give the stockbroker the result that she may want to see first, financial information about the corporation? There are a number of factors that a search engine might consider in determining which results to show to someone, and which order to show them in.
The search engine may also show news results, images, web search information and more, but the order in which those categories of information are presented may be different for different users. Which set is initially shown to a searcher, may be based upon predictions of which sets might be most helpful.
The Google local one box patent application is:
Presentation of Local Results
Invented by Gabriel Wolosin and Charity Yueh-Chwen Lu
Assigned to Google
US Patent Application 20080172374
Published July 17, 2008
Filed: January 17, 2007
Abstract
A computer-implemented method is disclosed. The method includes receiving from a remote device a search query, generating a local result set and one or more non-local result sets for the search query, determining a display location for the local result set relative to the non-local result set based on a position of the search query in a local relevance indicium.
A search for “pizza places,” may be seen as a local search, and a “local” set of search results may be initially displayed.
If the device someone is searching upon is a mobile computing device, like a cell phone or smartphone, that may call for ordering search results in a different order than if the search is done on a computer desktop.
A search for “Marilyn Monroe” or “James Dean” might result in images shown first, then shopping, then web, then blogs, then news.
A search for “polydicyclopentadiene,” might trigger a scholar or scientific articles-directed group as most relevant, and then groups sorted as web results, then blogs, then news, then shopping, then images.
Ordering Results Sets
There are a number of ways that different groups of result sets might be ordered.
1) They could be based upon looking at a particular query, and aggregated observations of user behavior in response to receiving query results for that query.
If most users searching for “Marilyn Monroe” are observed clicking on an “images” control even if the initial results are provided as web results, that user behavior may indicate that searchers associate the query closely with images and thus prefer to have images displayed first.
2) Results may be tied to the kind of device someone is using to connect to the web – the user of a handheld looking for “pizza” may be trying to find directions to a place where they can get a slice or two.
3) Results may be tied to a mix of user behavior information in response to a particular query, and type of device used in the search.
4) Results may be based upon result quality – if there is little relevant news for a query term, or few images, or little in the way of local search results, then returning those type of results first may not make much sense.
5) Results may be tied to a profile about a particular searcher, keeping track of the kinds of results that searcher might prefer to see. If 53% of Joe’s search queries are for map information and 61% of Jane’s queries relate to stock information, then showing maps first to Joe, and stock information first to Jane makes sense.
6) Results may be based upon how likely a query might be tied to “local search” results. The patent tells us that “terms or queries such as ‘restaurant,’ ‘weather,’ ‘movie,’ ‘directions,’ and ‘McDonalds’ may be highly correlated with local search requests. In contrast, terms or queries like ‘Google,’ ‘David Hasselhoff,’ ‘podcast,’ or ‘Monty Python’ would not.”
Local One Box – Single Search Results Sets When Appropriate
While some result sets may include lists of search results, such as image or web search results, others may include a single result such as local results or weather results when the user’s location is known, or a non-local result such as stock information.
The search engine may use a specific format for single-entry results, producing “more pleasing and easier-to-understand results that can be viewed in less screen space and with less need for the user to navigate through results.”
These could be a single map with one location upon it, or a dropdown for travelers to book a flight, or a weather report for a specific location, or a stock quote, or a number of other One Box results that Google shows.
The One Box is primarily aimed at improving a searchers experience when searching. Does it improve yours when you see a One Box Result?
Local One Box Conclusion
The patent filing delves into a lot more details about how a search engine might choose to order and present particular groupings of results for different searchers in different situations, based upon their chosen search terms.
Notice that when you search for some queries at Google, that the top results that you see might be mostly local, or mostly news, or scholarly, or images, or financial.
Is Google sorting search results based upon what kind of task that it thinks that you might be trying to perform, whether it’s to gather a specific kind of information, or to perform some kind of task such as finding the closest coffee spot or travel information, or performing some kind of navigational query to find a specific Web site?
What does that mean for searchers? What does it mean for site owners?
Yeah I think personalized search results is going to become a big thing soon and this will be good for some websites and bad for others.
Hi Chris,
Search isn’t getting any simpler, and finding ways to provide results that are helpful to searchers, even if those results differ for the same query, from one searcher to another, is definitely a direction that search engines are moving towards.
It is challenging, and will affect sites differently. It may mean not only anticipating what words someone searching for a site might use in a search, but also what they might expect to be able to do on a site once they arrive.
For instance, if those query terms are ones that searchers associate with pictures, then a site owner might want to add more images to their site – and present them in a way that a search engine might use to rank them well.
Simple premise of search is this – If a user is searching for X, I show results that get him X.
As a website owner who has X, you want people to come to you who want X.
From that perspective, website receives more targeted customers. I believe search engines have a lot of ground to cover to make this a reality.
At the same time, it makes it harder for new entrants to break into a new industry which is pretty close to how offline industries are 🙁
Rajat
Hi Rajat,
I think that you’re exactly right on the premise behind the approach described in the patent application, and it’s not surprising that Google is trying to show pages to searchers that may be the best match for what they want to see.
If a site owner works to understand who their audiences are, and what those audiences want to see on the pages of the site, then they should be off to a good start. What this document does tell us is that we not only need to pay attention to who the people are who might want what we offer on our websites, but also what words they might use to try to find what we offer, what kinds of devices those searchers might use to look for our sites, and what type of results (images, videos, web pages, scholarly articles, etc.) might be the most appropriate to focus upon in getting a message out to that audience.
There are a lot of challenges that any new site owner faces when they enter into a market – hopefully having a document like this patent filing from Google might give them some ideas on things that they can do that older, and more established sites aren’t.
For instance, focusing upon making web pages that show up well on a mobile phone, in an industry where people might use mobile phones to make reservations, or buy something while traveling, and where others in that market haven’t designed their sites to show up well on a smart phone or PDA, can be an opportunity that others may be missing.
The other interesting side to this argument is human-edited search directories, like Mahalo. In fact, I believe Mahalo results started showing up in Google’s OneBox results as well.
For SEO professionals, it all comes down to strong content. Search engines will become tougher to game, and human-edited directories will look for interesting, dynamic content. Searches will become more personal – you might not be able to dominate a term as before – but strong content will still help your site shine.
Hi Dan,
Thanks. Interesting point about Mahalo showing up in One Box results. I wonder how people would react if wikipedia articles started showing up in One Boxes.
Personalization will likely mean that businesses will have to know their intended audiences better, and create pages that not only address those people, but engage them, and keep them around.
Strong content that can compel someone to bookmark pages from a site, and keep them returning may be a start to getting pages to rank well in a world of personalized search results. That isn’t a bad approach regardless of personalization.
I think that ultimately this personalization will create some big challenges and opportunities for SEOs. It will be harder to target a single search phrase, since that phrase may deliver different results depending on many different factors. To me, that means focusing on quality content and hoping you get ranked, rather than trying to optimize for all of the myriad possible SERPs.
Another challenge is that your client might be seeing a different SERP than you are, meaning you have to spend more time educating them about how search engines work.
Overall, it will be harder to dominate, but also easier to show up somewhere, as long as your site is relevant. If the overall quality of visitors is improved, then I see this as a wonderful advance. The traffic metric is a pretty clumsy tool because it doesn’t tell you much about the quality of the visitors.
Hi Robert,
Excellent points on challenges involving both indexing and education of clients. If search engine personalization works the way that it is supposed to work, the people who arrive on your pages should be the audience that is interested in what your pages provide.
Of course, for a search engine to pull that off, the people behind the site need to know more than a few things about their audience and what they might want to see, and find a way to include that on the pages of their sites.