Google Changes Algorithm to Appease Media ... Or Not
On Black Friday, the New York Times ran one if its oh-so-savvy pieces about Google, demonstrating once again its deep understanding of SEO. </sarcasm> The article, titled "A Bully Finds a Pulpit on the Web," tells the story of DecorMyEyes, an eyeglasses business with an ungrammatical brand and a bad attitude. The owner, whose name, believe it or not, is Vitaly Borker, claims that horrible service – we're talking criminally bad – is his business strategy, because (ex-)customers leave negative reviews on the Internet, driving up his rankings: "I never had the amount of traffic I have now since my 1st complaint. I am in heaven."
Danny Sullivan did a long write-up of the article, calling it "great." But I was immediately suspicious of the NYT blaming DecorMyEyes' success on Google (the paper has a history of casting stones at the #1 search engine). This is really a story about a guy with abominable business practices – not a story about the failure of a search engine. Why is Google being made to be the bad guy, not Borker?
The reasoning seems to be that:
A)Google is supposed to return the "best results," so
B)There must be a problem with the algorithm, if this site has high rankings.
But it feels like people have forgotten what "best results" means – it means that Google returns the sites with the most relevant information. Google isn't necessarily vetting the businesses themselves or the quality of service you'll receive if you choose to become a customer. When I google "pizza," the #1 result is Domino's – I don't think this means Domino's offers the best available pizza, and if I order a pizza from Domino's and it shows up late with the wrong toppings, I don't blame Google, I blame Domino's. It's still my responsibility to research the businesses I'm considering giving business to.
The fact that the SERPs are full of negative reviews of DecorMyEyes and it's still getting business points to a problem with human nature, not search engines – a combination of "all publicity is good publicity" thinking and laziness. People should consider those negative reviews to be as relevant to their search as the site itself.
I'm not the only one who thinks this isn't Google's problem. Eric at SEO Boy writes: "People are blaming Google saying they need to change their algorithm. To me, this is ridiculous. Everyone is forgetting one important fact in this whole subject and it's the customer. From Google's perspective it would seem a no-brainer: 'Don't buy from someone with negative reviews,' this also suggests that we should do our own research before throwing our money at something online."
I enjoyed Byrne Hobart's piece on SEL called "No, You Can't Rank Well Just By Cultivating Terrible Reviews" – Hobart actually does a little investigating and finds that most of the negative reviews aren't passing any link value, if they include links at all. Rather, DecorMyEyes is probably ranking well in Google thanks to its profile of deep links from sites including – you guessed it! – the NYT and a bunch of shady/paid links. So if Google is failing here, it's more because it hadn't noticed the site was using black-hat techniques.
The "discussion of the week" on Sphinn was "What Should Google Do About Negative Activity Being Good for SEO?" First of all, that title begs the question; in other words, Matt McGee here is assuming it's a given that negative activity does help your SEO. But this goes against everything we've ever heard about reputation management – having negative press at the top of the SERPs is a not a good thing! People have paid a lot of money to get negative reviews or other bad press off the first page of Google – did one NYT story make us forget that?
Michelle Robbins was the first to take Google's side:
I'm going to play devil's advocate here (just for fun) and say that Google should be agnostic. The underlying premise here seems to be that Google should only show 'the good guys' and should somehow be able to algorithmically sort out who that is. If the results for 'Tom Ford Sunglasses' returns a business that sells them, isn't it my job as the consumer to investigate that business. If I do my job as a consumer, I'll see those bad reviews (that helped that business to rank), and will pass on giving them my business. I never assume that a top ranked site at Google is any kind of endorsement for that site/business by Google. I'd hate to see Google become the next branch of the nanny state - especially at the behest of this community.
A few others (Jill Whalen, Elysia Brooker) piped up to agree.
In light of all this I was really disappointed to see Google bend over in a response on its blog – instead of saying "not our problem," they capitulated with an agreement that "being bad to your customers" should be "bad for business":
We were horrified to read about Ms. Rodriguez's dreadful experience. Even though our initial analysis pointed to this being an edge case and not a widespread problem in our search results, we immediately convened a team that looked carefully at the issue. That team developed an initial algorithmic solution, implemented it, and the solution is already live. I am here to tell you that being bad is, and hopefully will always be, bad for business in Google’s search results.
The post goes on to discuss the options they had in terms of how to address the "problem," including blocking the particular offender, using sentiment analysis, and displaying user reviews. Instead of these options:
In the last few days we developed an algorithmic solution which detects the merchant from the Times article along with hundreds of other merchants that, in our opinion, provide an extremely poor user experience. The algorithm we incorporated into our search rankings represents an initial solution to this issue, and Google users are now getting a better experience as a result.
How vague is this? The implication is that they did not use sentiment analysis to alter the algorithm. So, what did they use? Did they actually change anything? Michael Gray thinks not – in a long rant on Twitter on Wednesday, he said it was obviously just a PR move, as there's no way they could have "fixed" the algorithm that quickly, and why would they want to anyway? He points out examples of other businesses with lots of negative sentiment attached that still rank highly for relevant queries.
Others seemed to take Google's post at face value. What do you think? Did Google really change the algorithm? Should it have?
Here are a few more responses:
- Lisa Barone says "Sorry, 'Crazy Bully' Isn't a Long-Term Business Strategy," pointing out that a temporary boost from a bunch of bad reviews, if it works at all, won't be sustainable for long.
- Bruce Clay's Jessica Lee wonders if "the spammy site slipped through the cracks and Google didn't want to admit it; maybe Google didn't want to draw attention to the spammy practices the business was conducting; maybe Google decided to be the hero and address The Times' perspective on the story." She seems to believe that Google has altered its algorithm, but wonders if it's crossing the line in doing so: "How can algorithms truly assess a business's entire worth?"
- Jill Kocher of Practical Ecommerce writes: "Searchers trust Google to 'recommend' the best stores at which to shop. After all, they wouldn’t rank well if they were bad, right? But Google's algorithm does not pass judgment on the quality and integrity of the stores and sites it ranks, nor should it. Each individual's experiences, values and culture influence decisions about good and bad, right and wrong. Expecting an algorithm to make those determinations for searchers is frightening, if not rare."
- John Battelle focuses on the word "opinion" in Google's response post: "If ever there was an argument that algorithms are subjective, there it is."
What if Google incorporated "sentiment analysis" into its algorithm, but didn't use it to rank results by default? It could be an option for some searches, much like you can now choose to sort results by date. This would be a helpful option for searches that return sites with customer reviews.
Is it Goopon Yet?
In other news, rumors have been swirling all week that Google plans to buy Groupon for several jillion billion dollars – 5 or 6, last I heard (what's another billion on the fire?). The deal hasn't materialized yet, but of course we've got plenty of commentary already:
Back in April, Evan Miller said Groupon has essentially the same business model as Google: "selling the scarce resource of a person's attention to the highest bidder."
David Mihm talks about the implications on local search:
But if the reported deal goes through, it would represent a colossal shake-up in our little world of Local Search…bigger even than Facebook Places (assuming Facebook eventually executes well on that potential). The only thing close to the magnitude of this deal would have been the proposed Google-Yelp deal that didn’t go through about this time last year. And I’m not even sure that was this big, for reasons I try to spell out below.
And here's Trada CEO Niel Robertson on why Groupon would be a great buy for Google.
By next Friday, maybe Goopon will be a reality?
More Web Marketing Highlights
When is a $17 product worth $16 per click? When you know the value of "retargeting" your customer base.
Linda Bustos talks about how vendors of physical goods are profiting from the boom in digital goods.
Smashing Magazine presents five steps for using multivariate testing to increase conversions.
And finally: Nathania Johnson provides an example of when not to abbreviate your signage. (I promise you will LOL.)
Have a great weekend!