Go to your medicine cabinet right now, choose any two medications at random, and then search Google for “MEDICATION-1 and MEDICATION-2″. Now click on any three listings at random. How many of those Websites actually provide you with clear and immediate information regarding the interactions of those two medicines?
Lately I find that I have to search on one medicine or the other and then look for interactions information on the page that is provided. If I search for two meds at a time I get a horde of advertising-laden medical pseudo-articles. They are clearly being generated by scanning public information databases and trawling for keywords.
Google did a pretty good job of cleaning up the single med SERPs but the combination SERPs are clogged with Web spam. I cannot imagine many people really finding good information in these SERPs.
But that is just the tip of the iceberg. I find that if I search for “good hotel and restaurant” I don’t get the selection of hotel+restaurant reviews I would expect to find in forums and consumer social media sites. Instead I am served a selection of optimized hotel and/or restaurant pages. This query is not the same as “good hotel and restaurant reviews” because that query implies (to me) that I am looking for good reviews of hotels and restaurants.
Panda sucks at filtering out clutter from combination queries. Why is that? (Of course, be careful if you search Bing for “good hotel and restaurant reviews” — is this 1996?)
Double-term queries are usually ambiguous. For example, if you search for “pancakes and syrup”, what are you searching for? Recipes? Restaurants that offer pancakes and syrup? Caloric information? The query is too inspecific.
But is that true of a query like “acetaminophen and tagamet”? Both Bing and Google show me a page from the National Institutes of Health as the first listing but after that the spam and the drug addict content kicks in pretty quickly. Neither search engine is doing a good job of deciphering the query and serving reliable, trustworthy content.
So should we all pile on the bandwagon and start hammering Google and Bing for doing such a lousy job? Or should we be asking why the reliable Websites are practicing such limited search engine optimization? If the spammers can influence picky Bing and post-Panda Google then surely the people optimizing search results for the NIH, Mayo Clinic, WebMD, and other generally trusted Websites should be talking to their teams about how to get into the SERPs for drug combinations.
I’m not saying it’s easy to do this — I’m saying that so far the Publishers and the Indexers are not doing a very good job of dominating the SERPs with good information. In the Searchable Web Ecosystem that leaves the Searchers with the burden of finding the right information — and last time I checked I didn’t have any degrees in medicine or pharmaceutical science. Even Wikipedia isn’t able to dominate these SERPs (which I find reassuring but at least Wikipedia could be reviewed and changed over time).
Am I being too fussy about a 2-term combination query pattern? You’re probably right. So let’s step it up and search for, oh, I dunno — maybe “acetaminophen and tagamet and protein diet”. What do you think of THEM search results? People who are concerned about their health (and approximately 1/3 of all American adults are now deemed to be obese or near-obese, meaning they are or soon will be experiencing all sorts of medical complications arising from being overweight) are searching for increasingly complex information. We no longer need to know just about drug interactions, now we need to know about drug interactions and the impact that high protein or low fat or slow carb diets have on these medication regimens.
But it gets worse.
Do you watch cable television in America? Have you been really impressed with those super coupon queens who buy hundreds of dollars’ worth of merchandise for just a few dollars? Do you get all fired up and then go search for things like “coupon savings calculator” or “coupon resources reviews”?
How easy is it to find something that was written in THIS DECADE? I have to click on a lot of Websites when I check out these kinds of queries. I can only imagine what people who really want to save with coupons have to go through in order to find the savings they are looking for.
You either end up with a lot of outdated information or doorway pages or just … stuff. I have no idea of how well monetized the coupon vertical is but it seems to me that coupon-related queries have experienced a sharp increase since late 2008. We can’t blame the Publishers for this — clearly, the search engines are just not paying attention to 3-term queries as much as they are paying attention to question queries.
In other words, the implicit question may be easy to decipher if you can stick “how to”, “where can I find”, “how do I”, etc. in front of the actual query but if the query is so complex it cannot be quickly inferred that way then the statistical signals associated with that query must be pretty complex — too complex for the algorithms.
People naturally increase the complexity of their queries. They do this as they learn how to use search engines, become more familiar with the type of content they can expect to find in SERPs, and switch from thinking in terms of exact match queries (“horse farms in arkansas”) to conceptual queries (“horse stables arkansas”). Do they want to buy a horse, put a horse up in a stable, or are they looking for riding lessons? I’m pretty sure that most people looking for riding lessons would think to use “riding lessons” but the SERPs suggest otherwise.
What do I know? I’m just the dumb hillbilly who doesn’t know how to search for horse riding lessons in arkansas.
These complex conceptual queries may be the next place where the content farms (whatever they may be) turn to for monetizable pages — but I think it more likely that the cottage industries that have found safe havens in these queries will eventually be supplanted by the major highly trusted Websites that are deemed “authorities in their fields”. In other words, Panda Version 2.0 or some other complex learning algorithm will probably hit the SERPs when Google has figured out how to fully automate Panda Version 1.0.
Bing, meanwhile, has an opportunity here to get a headstart on Google by looking at these conceptual queries and developing some strength in those results. After all, now that they have Yahoo!’s search data, they have no more excuses about being able to scale up to compete with Google. Bing will never overtake Google by following in its footsteps — but it just might give Google a serious run for its dollar by addressing these popular conceptual queries that combine unrelated terms.
The cottage industry queries seem to be populated by safe if irrelevant content, but the more consumers a complex query is related to, the more likely it’s being dominated by made-for-advertising spam. The spammers won’t appreciate my saying so, but Panda is so Last Month. Now I as a searcher want to see the search engines do something for me next week.
Lately I find that I have to search on one medicine or the other and then look for interactions information on the page that is provided. If I search for two meds at a time I get a horde of advertising-laden medical pseudo-articles. They are clearly being generated by scanning public information databases and trawling for keywords.
Google did a pretty good job of cleaning up the single med SERPs but the combination SERPs are clogged with Web spam. I cannot imagine many people really finding good information in these SERPs.
But that is just the tip of the iceberg. I find that if I search for “good hotel and restaurant” I don’t get the selection of hotel+restaurant reviews I would expect to find in forums and consumer social media sites. Instead I am served a selection of optimized hotel and/or restaurant pages. This query is not the same as “good hotel and restaurant reviews” because that query implies (to me) that I am looking for good reviews of hotels and restaurants.
Panda sucks at filtering out clutter from combination queries. Why is that? (Of course, be careful if you search Bing for “good hotel and restaurant reviews” — is this 1996?)
Double-term queries are usually ambiguous. For example, if you search for “pancakes and syrup”, what are you searching for? Recipes? Restaurants that offer pancakes and syrup? Caloric information? The query is too inspecific.
But is that true of a query like “acetaminophen and tagamet”? Both Bing and Google show me a page from the National Institutes of Health as the first listing but after that the spam and the drug addict content kicks in pretty quickly. Neither search engine is doing a good job of deciphering the query and serving reliable, trustworthy content.
So should we all pile on the bandwagon and start hammering Google and Bing for doing such a lousy job? Or should we be asking why the reliable Websites are practicing such limited search engine optimization? If the spammers can influence picky Bing and post-Panda Google then surely the people optimizing search results for the NIH, Mayo Clinic, WebMD, and other generally trusted Websites should be talking to their teams about how to get into the SERPs for drug combinations.
I’m not saying it’s easy to do this — I’m saying that so far the Publishers and the Indexers are not doing a very good job of dominating the SERPs with good information. In the Searchable Web Ecosystem that leaves the Searchers with the burden of finding the right information — and last time I checked I didn’t have any degrees in medicine or pharmaceutical science. Even Wikipedia isn’t able to dominate these SERPs (which I find reassuring but at least Wikipedia could be reviewed and changed over time).
Am I being too fussy about a 2-term combination query pattern? You’re probably right. So let’s step it up and search for, oh, I dunno — maybe “acetaminophen and tagamet and protein diet”. What do you think of THEM search results? People who are concerned about their health (and approximately 1/3 of all American adults are now deemed to be obese or near-obese, meaning they are or soon will be experiencing all sorts of medical complications arising from being overweight) are searching for increasingly complex information. We no longer need to know just about drug interactions, now we need to know about drug interactions and the impact that high protein or low fat or slow carb diets have on these medication regimens.
But it gets worse.
Do you watch cable television in America? Have you been really impressed with those super coupon queens who buy hundreds of dollars’ worth of merchandise for just a few dollars? Do you get all fired up and then go search for things like “coupon savings calculator” or “coupon resources reviews”?
How easy is it to find something that was written in THIS DECADE? I have to click on a lot of Websites when I check out these kinds of queries. I can only imagine what people who really want to save with coupons have to go through in order to find the savings they are looking for.
You either end up with a lot of outdated information or doorway pages or just … stuff. I have no idea of how well monetized the coupon vertical is but it seems to me that coupon-related queries have experienced a sharp increase since late 2008. We can’t blame the Publishers for this — clearly, the search engines are just not paying attention to 3-term queries as much as they are paying attention to question queries.
In other words, the implicit question may be easy to decipher if you can stick “how to”, “where can I find”, “how do I”, etc. in front of the actual query but if the query is so complex it cannot be quickly inferred that way then the statistical signals associated with that query must be pretty complex — too complex for the algorithms.
People naturally increase the complexity of their queries. They do this as they learn how to use search engines, become more familiar with the type of content they can expect to find in SERPs, and switch from thinking in terms of exact match queries (“horse farms in arkansas”) to conceptual queries (“horse stables arkansas”). Do they want to buy a horse, put a horse up in a stable, or are they looking for riding lessons? I’m pretty sure that most people looking for riding lessons would think to use “riding lessons” but the SERPs suggest otherwise.
What do I know? I’m just the dumb hillbilly who doesn’t know how to search for horse riding lessons in arkansas.
These complex conceptual queries may be the next place where the content farms (whatever they may be) turn to for monetizable pages — but I think it more likely that the cottage industries that have found safe havens in these queries will eventually be supplanted by the major highly trusted Websites that are deemed “authorities in their fields”. In other words, Panda Version 2.0 or some other complex learning algorithm will probably hit the SERPs when Google has figured out how to fully automate Panda Version 1.0.
Bing, meanwhile, has an opportunity here to get a headstart on Google by looking at these conceptual queries and developing some strength in those results. After all, now that they have Yahoo!’s search data, they have no more excuses about being able to scale up to compete with Google. Bing will never overtake Google by following in its footsteps — but it just might give Google a serious run for its dollar by addressing these popular conceptual queries that combine unrelated terms.
The cottage industry queries seem to be populated by safe if irrelevant content, but the more consumers a complex query is related to, the more likely it’s being dominated by made-for-advertising spam. The spammers won’t appreciate my saying so, but Panda is so Last Month. Now I as a searcher want to see the search engines do something for me next week.
No comments:
Post a Comment
Get 1000 Friends Instant