Ask.com’s patent application published this week, Methods and systems for generating query and result-based relevance indexes, notes that information from users’ queries and interactions with search results is limited in use to provide more relevant results to searchers.
That kind of information includes time of searches, the geographic sources of queries, and demographic variables about searchers. This information might be useful to businesses willing to pay for it.
Should a search engine offer that kind of information to businesses?
Some examples from the patent application:
On a query for Nascar:
Most stock car races are located in Southeastern region. However, in a case where a plot of the geographic profile for users asking about “NASCAR” indicates that the number of “NASCAR” queries originating from the Northwestern region was unexpectedly high, external entities, such as NASCAR, may use this business information on the public interest to by moving some races to tracks in the interested region.
On queries for specific movie names:
Foreign and art house movies are often released in a limited number of major cities. Query volume may indicate regions where interest in a particular movie is unusually high, perhaps due to concentrations of an ethnic or language group in that location. The same could be said for other products only available in limited release, such as designer fashions and specialty foods.
On searches for specific musicians’ names:
A musical artist is planning a tour. A factor that might affect which cities are on the tour could be the level of interest in the artist as measured by queries. One artist might be popular in college towns, another in urbanized areas, another only on the West Coast.
On queries for advertised products:
An external entity, such as an advertising company is trying multiple ad campaigns in different cities. The relative increase in queries about the product in each city could be one measure of the effectiveness of each campaign.
How much might a search engine know about its users, and what kinds of information might it share with others?
In another embodiment, demographic variables of a user may be logged. Exemplary demographic variables include age, gender, income, occupation, education, socioeconomic status, religion, nationality, race, family life cycle, family size, and sexual orientation. Any number and combination of demographic data may be used to define a demographic metric. Further, psychographic characteristics (adventuresome, techno-phobic, conservative, etc.) of users may also be logged. Various methods of determining a likely demographic variable or psychographic characteristics may be used.
All of the major search engines collect a considerable amount of information about the queries that people use, and their interactions with the search engine and other Web sites.
Indexing information found on the Web, and providing information to searchers is one thing. Collecting and indexing information about searchers and offering it for sale is something that many searchers may not consider in their use of search engines.
My oh my.
I just hope that Ask won’t pursue this trend too much.
I’m sure that it is a temptation to any organization that accumulates data on that kind of level to see how it might be used in other ways.
I guess in some ways, the recent controversies around Google and privacy might be concerned with this issue, too. It is a lot of data – how else might it be used.
It seems like an extended version of Google Trends. You can already see where geographical interests are on a keyword, so in the NASCAR example this data is already available – and for free to all.
Also, Microsoft has tried to use similar data to help its Adcenter advertisers allowing them to target by demographic data.
So I don’t see it too much of a stretch for them to sell the data although I figure it would be more likely that they will just give it away in a fashion similar to how Google and Microsoft have done so far.
Hi Mark,
It’s going to be interesting seeing where Ask may take this, and how tempted the other search engines are in sharing information. I think that there is the potential for them to take it too far.
It’d be a huge advantage to have info like this. While it would make things more efficient in a way, it would also make them uneven if the price were too high and only the top companies could afford to buy the stats. I think it would be great if they sold it for a nominal cost.
That’s a tough one. I suppose it would depend on how much and what data they chose to release.
Hi Christina,
Good points. The patent application doesn’t go into detail on how such information would be sold, but it does refer to the people who might use it as “clients.” The described uses tend to go beyond where one would advertise, so they aren’t talking about advertisers.
Chris (business blogger),
I agree that it would probably depend upon where lines were drawn regarding what kind of information was released, and how much of it. We’re just not given enough information on that front, from this document.
I don’t see that selling some very general data on keyword searches– e.g., how many people in X region are using X keyword– would be a privacy issue. That kind of stuff can’t be connected to individual people. The danger is if search engine companies collect data that can relate searches to specific IP addresses, or people (for example– by crossreferencing search terms). As we’ve all seen, at some point the government might be tempted to subpoena such data, which could put a damper on the freedom with which people use the technology.
Hi Robin,
The ability to associate searches with specific IP addresses probably would be a problem.
If detailed demographics are available, we start getting closer to that level of granularity.
If small geographic areas such as Silicon Valley are targeted, might there be issues involving information gathering amongst high tech companies that might be an issue?
This kind of business intelligence information is very valuable. And used in the right way it will make shopping and other activities easier for everybody.
Used in the wrong way – I guess I don’t have to explain that. It all comes down to company ethics.
My only concern is that users of such services must be made aware that their anonymized data can be sold/used and they must be able to turn tracking of their activities off.
Hi Nikolaj,
All of the search engines collect an incredible amount of information, both about the Web, and about how people use it.
I suspect that all of them would be pretty hesitant to provide information to anyone that includes specific personal information about specific people, but we are seeing some information from them shared about things like what topics seem important to searchers through things like Google Trends and Google Insights for Search, and I suspect that we will see more.
I agree completely with you that the providers of these services should make their users aware of how their information might be used.
It will improve the end users experience so i say yes.
Hi Dan,
I’m not sure that this kind of sale of business intelligence would benefit searchers – the idea isn’t to provide information that might benefit websites, but rather to sell information about people’s searches to benefit businesses. The problem I see with it is that it may present a conflict of interest on behalf of the search engines – instead of focusing upon providing information to searchers, it takes information about searches and sells it to the highest bidder. It does come down to making sure that searchers are fully informed about how information related to their searches might be used by the search engine – especially when it isn’t being used to improve the quality of the searches themselves.