The narrative is a familiar one. Seemingly intelligent individual hops onto a general-purpose search engine, with a really specific purpose in mind. Gets impatient, has limited background in doing this particular kind of specialized database research, gives up and blames Google or whoever for being spam-ridden and self-interested.
The “search stinks” meme goes all the way back to when the likes of John Dvorak started their articles with that premise over ten years ago. Unfortunately, since search still does stink and I’m still too lazy to really search (like most of the rest of you), when I try to dig up Original Dvorak, all I’m getting is a more recent 2008 piece by him on how the new Cuil search engine stinks. But I digress.
Anyway, Blekko is fine and working on admirable innovation in the search startup space, but this Techcrunch piece by purported academic Vivek Wadhwa does Gooogle – and search, and searchers – a disservice by relying on sweeping generalizations to complain about failed searches, apparently pinpointing the cause of the problem as Google’s bias towards (or inability to filter out) low-quality sites that profit from advertising links.
The example Wadhwa uses is doing reconnaissance on specific individuals in the technology industry. But how would one structure such a search? Are there ways to structure that task so Google, or for that matter Blekko, would perform better?
When I do similar research, I don’t get the same sense of a spammy cesspool that Wadhwa does. Should he elaborate by comparing specifics, like say how the Google “Timeline” feature works vs. Blekko’s /date feature? Should we be probing more about the technology inherent in dating pages as they are published on the Web? Wadhwa unhelpfully tells us that Blekko “analyzes other information embedded in the page’s HTML.” Hate to break it to you, but if Blekko can do it, then Google can easily do it, and probably will.
It’s no secret that Google fails to dazzle on long tail queries. The solution — in part, the innovation Blekko brings to the table — is to sacrifice breadth in favor of imposed structure. By editing out the vast majority of potential publishers of information (and “editing in” curated “thumbs up” high-quality sites), the members of the blekko community may help each other do research with less spam.
For certain, predetermined tasks — assuming a willing army of volunteers to work on the curation database — that’s no doubt going to produce satisfying results. But is this a Google problem? Did Google claim to be solving that problem?
Where we impose structure, we impose significant hurdles in terms of user and community willingness to play along, to say nothing of sacrificing comprehensiveness.
It’s pretty unfair to Google to claim that certain types of unstructured searches were unsatisfying as compared with sites that are more cut out for those tasks — such as LinkedIn! — especially when Wadhwa admitted that LinkedIn proved unsatisfying in itself, despite all the structure it imposes.
There’s a certain Stone Soup quality to the use of a resource like blekko. Confronted with the idea that there are limited resources and a community of volunteer curators, and a need to pick up your searching game to ensure that the tool works to its maximum potential… darned if you won’t pick up your game and make it work by tapping into its full capabilities. Think that might be possible if you just put a bit more elbow grease into Google searches?
The fact that Paul Kedrosky can’t find credible and helpful reviews about dishwashers is a more significant problem. This does point to a troubling fact of life: Google is creating its own way of curating, sorting, and highlighting user-generated content, reviews, local information, etc., and they don’t stack up too well with how many of us are looking to consume review content. At this juncture, it appears that there is a certain quirkiness to how Google returns results based on guessed intent: more generic but popular consumer queries like “bosch dishwashers” actually do *better* than something more pointed like “bosch dishwashwer reviews.” Although you would think that “bosch dishwasher reviews” would give you more helpful results than the more generic result, chances are what you get instead is less curated, less savvy SERP’s, because even that three-word query with “reviews” in it seems to be a pretty low volume query (read: long tail search query). Long-tail queries right now seem to be quite susceptible to allowing these second-rate Made-for-Adsense, non-best-of-breed sites to rank too high, sometimes first on the page.
It can’t be good for Google’s image to be seemingly so incapable of raising their game on what kind of content users access when they type in an obvious query like “bosch dishwasher reviews.” Surely, it shouldn’t be too hard here to get a handle on the notion that dishwasher-review.com is a non-best-of-breed, scrapey-looking, Made-for-AdSense site should never rank in the top five or six results on the page. And shouldn’t, maybe, Amazon and one or two other sites almost always rank here, as TripAdvisor once did for any travel-related review query?
Perhaps the answer is that Google has lost some of its consumer-advocate’s compass or its focus. It’s also the case, of course, that there are so many challenges with spammy content on the open web that it is hard to keep on top of it all.
Personally, I don’t expect to do every type of custom research using a free, general search engine, hence my rejection of Wadhwa’s specific complaint.
But I do expect Google to serve me some pretty impressive results when I type in a mainstream query about dishwashers. Right now, the main SERP’s aren’t always great, Google keeps shifting how it treats these queries, and Google wants to assume my intent is to compare prices, buy immediately, or find a local vendor to buy it at.
Is Google trying to take tire-kicking, researchey intent and gently massage it into more directly commercial intent? At some point did it lose sight of the fact that it advocates for consumers, rather than representing a Yellow Pages 2.0 listing service for businesses that want to conform to its listing formats? Was a line crossed at some point? Many think it may have been, but all that proves is that Google is a for-profit business.
So yes, this just in, Google can’t be expected to do everything, and yes, it’s flawed in key ways. Charges of a cesspool of spam and junk seem overdone, however. And don’t even get me started on the misguided criticisms of perfectly legitimate AdWords search ads that millions of users happily click on every day.