Since it took me a few hours to put together my SMX presentation I figured it was worth sharing that information on the blog as well. This post will discuss examples of how Google has dialed up their brand bias over time & points to where Google may be headed in the future.
Note that I don’t have anything against them promoting brands, I just think it is dishonest to claim they are not.
Against All Odds
When analyzing Google’s big-brand bias the question is not “do some small sites manage to succeed against all odds” but…
- What are the trends?
- What are the biases?
Eric Schmidt once stated that “Brands are the solution, not the problem. Brands are how you sort out the cesspool. Brand affinity is clearly hard wired.”
We have a fear of the unknown. Thus that which we have already experienced is seen as less risky than something new & different. This is a big part of why & how cumulative advantage works – it lowers perceived risk.
A significant portion of brand-related searches are driven by offline advertising. When a story becomes popular in the news people look online to learn more. The same sort of impact can be seen with ads – from infomercials to Superbowl ads. Geico alone spends nearly a billion Dollars per year on advertising, & Warren Buffet mentioned that 3/4 of their quotes come from the internet.
Some of the most profitable business models are built off of questionable means.
Many big brands are owned by conglomorates with many horses in the race. When one gets caught doing something illegal they close it down or sell off the assets & move to promote their parallel projects more aggressively.
If things aligned with brands become relevancy signals then to some degree those measure longevity & size of a company (and their ad budget) rather than the quality of their offering.
Even before the Panda update Google’s Amit Singhal suggested the problem with this:
Companies with a high page rank are in a strong position to move into new markets. By “pointing” to this new information from their existing sites they can pass on some of their existing search engine aura, guaranteeing them more prominence.
Google’s Mr Singhal calls this the problem of “brand recognition”: where companies whose standing is based on their success in one area use this to “venture out into another class of information which they may not be as rich at”. Google uses human raters to assess the quality of individual sites in order to counter this effect, he adds.
Since Panda Overstock has moved into offering ebooks & insurance quotes while companies like Barnes & Noble run affiliate listings for rugs.
As an example of the above trend gone astray, my wonderful wife recently purchased me a new computer. I was trying to figure out how to move over some user databases (like our Rank Checker & Advanced Web Ranking) and in the search results were pages like this one:
The problems with the above are:
- actual legitimate reviews get pushed down by such filler
- the business model behind doing such actual reviews gets eroded by the automated syndicated reviews
- outside of branding & navigation the content is fully syndicated
- that particular page is referencing the 2005 version of the software, so the listed price is wrong & the feature set has changed a lot in the last 7 years
Such scrape-n-mash content strategies by large brands are not uncommon. Sites like Answers.com can quickly add a coupons section, sites like FindTheBest can create 10s of millions of automated cross-referencing pages that load a massive keyword net of related keywords below the fold, news sites can create auto-generated subdomains of scraped content, etc.
Eric Schmidt highlighted FindTheBest publicly as an example of a successful vertical search play. That site was launched by an ex-Googler, but if I did the same thing you can be certain that the only way Google would highlight it publicly would be as a “type of spam.”
The issue with broadly measuring user experience is that I am still going to visit Yahoo! Sports repeatedly even if my experience on Yahoo! Downloads is pretty crappy. A site which is a market leader in one niche can take thoe signals to launch a “me too” service in other parallel markets & quickly dominate the market.
Potential Brand Signals
When attempting to debunk the concept of “brand bias” some people claim that it would be rediculous for Google to have a list of brands that get an across-the-board boost. Of course that debunking is debunking a straw man that was never stated publicly (outside of the irrelevant debunking).
However, some of Google’s old rater documents *did* have certain sites whitelisted & Google’s Scott Huffman once wrote the following:
At a [search] quality level, we have something similar. On a continuous basis in every one of our data centers, a large set of queries are being run in the background, and we’re looking at the results, looking up our evaluations of them and making sure that all of our quality metrics are within tolerance.
These are queries that we have used as ongoing tests, sort of a sample of queries that we have scored results for; our evaluators have given scores to them. So we’re constantly running these across dozens of locales. Both broad query sets and navigational query sets, like “San Francisco bike shop” to the more mundane, like: Here’s every U.S. state and they have a home page and we better get that home page in the top results, and if we don’t … then literally somebody’s pager goes off.
(Outside of some fraternal Google properties) the algorithm isn’t hardcoded to rank sites x & y at #1, but if some sites don’t rank for certain queries it does cause an alert to be sent out.
A while back we wrote a post on potential brand signals, but a short list of examples would be:
- Classical relevancy signals
- domain name
- website age
- anchor text
- link diversity
- keyword co-citation
- inclusion in trusted databases
- Search behavior
- keyword search volume trends
- CTR of users on search results (including how users respond to changes in rank)
- URL-based searches & other branded searches (the most popular keyword on Google is Facebook)
- back button clicks (did the user find what they were looking for? or did they look somewhere else?)
- repeat visitors (if someone repeatedly visits a website that is generally a pretty strong indication they had a positive user experience)
- search query chains (Google suggested this was a big driver in the Vince update)
- Passive user monitoring
- search has become the primary mode of navigation online
- Google has long offered a search toolbar & paid to have it installed in new computers
- Google paid Mozilla about a billion Dollars for default search placement in Firefox
- Google owns Chrome & Android
- Google offers the most widely used analytics program
- Google can also use AdSense ads and YouTube data to track users
- Google was recently caught in privacy-related snafus with tracking Safari & Internet Explorer users
In 2008 Rhea Drysdale created the following image, which highted how the same activity could be viewed as a legitimate marketing strategy or spam based on nothing other than who was doing it.
The Vince Update
In 2009 Google rolled some of their brand bias directly into the relevancy algorithms. A bunch of branded sites all jumped up in rankings out of nowhere for core industry keywords.
Around that time Microsoft offered a search funnels tool, which showed what people searched for after searching for a particular keyword.
The above screenshots (from Rankpulse and the Microsoft Search Funnels) are both from now defunct tools, but Yahoo! has since launched a tool called Yahoo! Clues which shows similar relationships.
Amit Singhal told the Telegraph that Google is “the biggest kingmaker on this Earth.”
A Google engineer admitted that the Vince update was largely driven by search funnels. Google then rolled out a search results interface change which promoted brands & stores directly in the search results.
If you search for “fishing gear” and then click their Bass Shop refinement link in the search results, you are thus directly creating that search funnels relevancy “signal.” Even if you don’t click on that link the exposure to the term may make you remember it and search for it later.
Are paid links evil?
Once again, it depends on who is doing it.
When the largest flower websites were caught buying massive quantities of links, a Google spokesperson told the New York Times: “None of the links … had a significant impact on our rankings, due to automated systems we have in place to assess the relevance of links.”
When some small bloggers were selling paid links to K-Mart as part of a “sponsored conversations” outreach, Matt Cutts equated the practice to selling bogus solutions to brain cancer & stated: “Those blogs are not trusted in Google’s algorithms any more.”
Google also started sending webmasters automated messages for bad links pointing at their sites:
Dear site owner or webmaster of domain.com, We’ve detected that some of your site’s pages may be using techniques that are outside Google’s Webmaster Guidelines.
We encourage you to make changes to your site so that it meets our quality guidelines. Once you’ve made these changes, please submit your site for reconsideration in Google’s search results.
So if you run a big site & they automatically detect paid links they generally just ignore those links and leave your site alone. If you are a small site & they automatically detect paid links they may decide to automatically penalize your site.
Same offence, entirely different outcome.
Is cloaking evil?
Once again, it depends on who is doing it.
I have a Vistaprint Visa card (so I could get a credit card with our dog’s picture on it) and one of the pages that was ranking for Vistaprint Visa was the Singapore Groupon website.
The page forces a pop up and you can’t do anything on that page (view the content, scroll around the site, etc.) other than filling in the lead generation form or logging into an existing account. I would never try that because I know I would get smoked for it.
Groupon has also ran AdWords accounts where the only option was to fill in the lead generation form or click into the TOS which are in another language!
After the first iteration of the Google Panda update Google allowed users to vote to block websites. Experts Exchange was hated among some programmers in part because they used scroll cloaking. That in turn got their site hit by the second Panda update.
Google then later rolled out a new ad unit where you pay for viewing content by taking a Google survey & some YouTube videos use preroll ads.
Are doorway pages evil?
Once again, it depends on who is doing it.
After the Panda update Ikea’s thin content-free pages started ranking page 1 for some pretty competitive keywords.
Huffington Post later wrapped 3rd party Tweets in their site’s template & ranked those in Google.
Smaller webmasters who ran network of sites in some cases got hit with “doorway page” penalties for owning networks of sites registered in Google Webmaster Tools, even if each site was a full fledged ecommerce website.
Is content farming evil?
Once again, it depends on who is doing it (and where it is hosted).
Long before the Panda update I highlighted some of the informationless videos Demand Media was uploading to Youtube.
In spite of Google’s Panda hitting eHow, Google still decided to pre-pay Demand Media to keep uploading YouTube videos.
Another thing that is interesting about the content farms and the alleged need for the Panda algorithm was that in spite of flagrant editorial violations by both eHow and Mahalo, Google didn’t smoke them until it could be done “algorithmically.”
On the flip side of the above, in some cases Google has chose to keep smaller webmasters penalized because content that was at one point on their site months in the past!
When Google+ launched I highlighted how it was acting as a scraper site by outranking original publisher content. About a half-year later some tech blogger noticed that issue & caused a big stink over it. A Google engineer then suggested that it was childish to place any of the blame on Google. Shortly after that Google integrated Google+ in the search results far more aggressively.
A couple weeks after that aggressive promotional integration Amit Singal stated: “The overall takeaway that I have in my mind is that people are judging a product and an overall direction that we have in the first two weeks of a launch, where we are producing a product for the long term.”
The problem with build preferential rankings first & increase quality later is that is the exact opposite of what Google is asking publishers to do with algorithms like Panda. Worse yet, Google not only does this integration when you are logged in, but also shows it on obscure longtail advanced queries when you are not logged in.
When Google’s ad ecosystem was young they loved affiliates, but that changed over time.
In Google’s remote rater documents they suggested that hotel affiliate sites be marked as spam, even if they are helpful.
On Google’s reconsideration request form they also stated: “In general, sites that directly profit from traffic (e.g. search engine optimizers, affiliate programs, etc.) may need to provide more evidence of good faith before a site will be reconsidered.”
And while Google has biased their editorial philosophies away from affiliates, some of the trusted brands like Barnes & Noble added affiliate listings to their websites, selling things like rugs.
The Business Cycle
Most businesses tend to grow in a cycle…
- Boostrap / self-funded
- Raise funds / take out a loan
- Build exposure
- Monetize attention
- Re-invest in increased quality
- Build a brand
- Build further exposure
- Monetize more attention
- Re-invest in increased quality
The broken piggy bank in the above cycle highlights the break that exists in the process to building a big brand. It is quite hard to have any level of certainty in the search ecosystem with an alogorithm like Panda. Without that level of certainty companies must build from low cost structures, but that very constraint makes them more likely to get hit by an algorithm or search engineer.
Being an entrepreneur is all about taking smart calculated bets & managing risk. However as search engines become closed off portals that compete with (& exclude) publishers, there are so many unknowns that estimating risk is exceptionally challenging.
- When the New York Times bought About.com Google was one of the leading competing bidders. But after Panda, About.com’s profit declined by 2/3.
- TeachStreet was a fast growing start up that was shut down due to Panda.
- CustomMade is a Google-funded start up launched by an SEO who purchased an old website & created a vertical directory out of it (just like TeachStreet was trying to do, but in a different vertical). Googler’s helped with the project & the article highlighting that shared this quote: “Having Google as an investor gives you a branding piece that you can’t ignore.” – Christopher Ahlberg.
Penalties: How Hard Were They Hit?
- Years ago when BMW or WordPress.org got caught spamming aggressively they were back in good graces in a mater of days.
- About the only times well known (non-affiliate) sites have been penalized for a significant duration was when JC Penney & Overstock.com were hit. But that happened around the time of the Panda fiasco & Google had incentive to show who was boss. When the flower sites were outed for massive link buying that was ignored because Google had already rolled out Panda & reasserted the perception of their brand.
- When Google was caught buying links (again) to promote Google’s Chrome browser & that story spread widely throughout the mainstream press, Googlers lied & claimed there was only 1 paid link in 1 single page & penalized a single page of their site. Small website owners that have been caught in similar link buying (or selling) campaigns have been hit much harder. Remember the above story about the bloggers blogging about K-Mart? So far this year Google has sent webmasters over 700,000 messages in Google Webmaster Central.
1 Strike – You’re Out
In 2009 Google banned over 30,000 affiliates from the AdWords auction. In some cases the problem was not with a current ad (or even a landing page the advertiser controlled), but rather ads that ran years ago promoting 3rd party products. In some cases Google changed their AdWords TOS after the fact in an ex-post facto style. Google won’t allow some of these advertisers to advertise unless they fix the landing page, but if they don’t control the landing page they can’t ever fix the problem. Making things worse, to this day Google still suggests affiliates do direct linking. But if the company they promote gets bought out by someone too aggressive then that affiliate could be waiting for a lifetime ban through no fault of their own.
A popular programmer who has been an AdSense publisher for 8 years had their AdSense account arbitrarily suspended without warning. After an ex-Googler expressed outrage over the issue he was able to get his AdSense account reactivated. A publisher without those friendships would have been done.
In Australia a small travel site had a similar issue with AdSense. The only way they were able to get a reconsideration was to lodge a formal complaint with regulators. If that is how Google treats their business partners, it colors how they view non-business partners who monetize traffic without giving Google a taste of the revenues.
Why Does Google Lean Into Brand?
- Minimize legal risks: if they hit small businesses almost nobody will see/notice/care, but big businesses are flush with cash and political connections. When Google hits big businesses they create organizations & movements like Fair Search & Search Neutrality.
- Minimize duplication: some small businesses & affiliates simply repeat offers that exist on larger merchant sites. That said, many big businesses buy out a 2nd, 3rd, 4th, or even 5th site in a verticle to have multiple placements in the search results.
- Better user experience: the theory is that the larger sites have more data and capital to improve user experience, but they don’t always do it.
- Business partnerships: if Google wants to strike up closed door business partnerships with big business then some of those negotiations will have specific terms attached to them. It costs Google nothing to give away part of the organic results as part of some custom deals. If Google wants to sell TV ads & run a media streaming device they need to play well with brands.
- CPA-based product ads: on some searches Google provides CPA-based product ads above the search results. It makes sense for Google to promote those who are buying their ads to get the best relationships possible.
- Fewer people tasting the revenues: the fewer organizations an ecosystem needs to support the more of the profits from that ecosystem that can be kept by the manager.
- More complete ad cycle: if Google caters to direct response advertisers they get to monetize the demand fulfillment of demand, however that is only a small slice of the complete ad cycle. If Google caters to brands they get to monetize (directly or indirectly) every piece of the ad cycle. For example, buying display ads helps build brand searches which helps create brand signals. In such a way, improved rankings in the organic results subsidize ad buying.
- Brands buying their equity: Google has create exceptionally large ad units & has convinced many brands to buy their own pre-existing brand equity.
Lack of Diversity
The big issue with brand bias is that a lot of the same *types* of companies rank with roughly similar consumer experiences. If there is a mix of large and small businesses that rank then many of those small businesses will be able to differentiate their offering by adding services to their products, doing in-depth reviews, and so on.
Sure Zappos is a big company known for customer service, but how different is the consumer-facing experience if I click on Target.com or Walmart.com? Sure the text on the page may be slightly different, but is there any real difference beyond aesthetic? Further, a lot of the business models built around strong in-depth editorial reviews & comparisons are eroded by the current algorithms. If the consumer reviews are not good enough, then tough luck!
Do Brands Always Provide a Better User Experience?
Some larger retailers track people in ways that are creepy:
For decades, Target has collected vast amounts of data on every person who regularly walks into one of its stores. Whenever possible, Target assigns each shopper a unique code — known internally as the Guest ID number — that keeps tabs on everything they buy. “If you use a credit card or a coupon, or fill out a survey, or mail in a refund, or call the customer help line, or open an e-mail we’ve sent you or visit our Web site, we’ll record it and link it to your Guest ID,” Pole said. “We want to know everything we can.”
Many big media companies provided watered down versions of their content online because they don’t want to cannibalize their offline channels. Likewise some large stores may consider their website an afterthought. When I wanted to order my wife a specific shoe directly from the brand they didn’t have customer support open for extended hours during the holidays and their shopping cart kept kicking an error. Since they *are* the brand, that brand strength allows them to get away with other issues that need fixed.
Some of those same sites carry huge AdSense ad blocks on the category pages & have funky technical issues which act like doorway pages & force users who are using any browser to go through their homepage if they land on a deep page.
Missing the Target indeed.
That above “screw you” redirect error has been going on literally for weeks now, with Target’s webmaster asleep at the wheel. Perhaps they want you to navigate their site by internal search so they can track every character you type.
Riding The Waves
With SEO many aggressive techniques work for a period of time & then suddenly stop working. Every so often there are major changes like the Florida update & the Panda update, but in between these there are other smaller algorithmic updates that aim to fill in the holes until a big change comes about.
No matter what Google promotes, they will always have some gaps & relevancy issues. Some businesses that “ignore the algorithms and focus on the user” are likely to run on thinner margins than those who understand where they algorithms are headed. Those thin margins can quickly turn negative if either Google enters your niche or top competitors keep reinvesting in growth to buy more marketshare.
Given the above pattern – where trends spread until they get hit hard – those who quickly figure out where the algorithms are going & where there are opportunities have plenty of time to monetize their efforts. Whereas if you have to wait until things are widely spread on SEO blogs as common “tricks of the trade” or wait until a Google engineer explicitly confirms something then you are likely only going to be adopting techniques and strategies after most of the profit potential is sucked out of them, just before the goal posts move yet again.
People who cloned some of the most profitable eHow articles years ago had plenty of time to profit before the content farm business model got hit. Those who waited until Demand Media spelled their business model out in a Wired article had about 1.5 years until the hammer. Those who waited until the content farm controversy started creating a public relations issue to clone the model may have only had a couple months of enhanced revenues before their site got hit & was worse off than before they chased the algorithm late in the game.
Ride The Brand
If Google does over-represent established branded websites in their algorithms then in many cases it will be far easier to rank a Facebook notes page or a YouTube video than to try to rank a new site from scratch. There are a ton of web 2.0 sites driven by user generated content.
In addition to those sorts of sites, also consider participating in industry websites in your niche & buying presell pages on sites that rank especially well.
Collecting (& Abusing) User Data
Google has been repeatedly branded as being a bit creepy for their interest in user tracking.
Collecting that data & using it for ad targeting can have profound personal implications (think of serving a girl with anorexia ads about losing weight everywhere she goes online, simply because she clicks the ad, in such a case Google reinforces a warped worldview). Then when the person needs counseling Google can recommend a service provider there as well.
Trust in Google’s ability to do the right thing would be greater if they were not caught in that drug sting selling ads to fake Mexican pharmacies selling illicit products, a practice they were involved in before going public.
They also take aggregate collected data and sell it off to banksters.
Google as Content Host (& Merchant)
Throughout the history of the web there will be many cycles between open and closed ecosystems. Currently we are cyling toward closed silos (Apple, Amazon, Google, Facebook). As these silos become more closed off they will end up leaving gaps that create new opportunities.
Goolge has been pushing aggressively for years to host content & crowd out the organic search results.
- As publishers mark up their content Google can aggregate & displace publishers higher up the value chain. Google has done research on measuring non-click user satisfaction (no need to bother leaving Google & visiting the publisher site). If your business model was one of the examples on that list I would bet on Google entering your niche aggressively sooner rather than later.
- When many other content farms were hit YouTube gained traffic. YouTube is investing in original content & partnering with NBC for Olympics coverage and Disney for kid shows.
- Google+ is inserted directly in the organic search results & often outranks the original content source.
- Google Books has seen aggressive integration in the search results, even before the legal issues have been resolved.
- Google Places has long pushed down the organic search results & in some cases (like with hotels) Google also adds price ads to these listings.
- Google Flights has pushed down the organic search results & Google is also powering flight booking options on some smaller regional carriers.
While on one front Google keeps making it easier for brands to compete against non-brands, Google also keeps clawing back a bigger slice of that branded traffic through larger AdWords ad units & integration of listings from services like Google+, which can in some cases outrank the actual brand.
Google has ran multiple platforms (Android Marketplace, Chrome Marketplace, Enterprise Marketplace) competing against iTunes. Google recently decided to merge some of their offerings into Google Play. In addition to games, music & books, Play will soon include audiobooks, magazines & other content formats.
Google also wants to compete against Amazon.com to launch an Amazon Prime-like delivery service.
Having a brand & following will still be important for allowing preimium rates, fatter margins, building non-search distribution (which can be used to influence the “relevancy” signals), and to help overturn manual editorial interventions. But algorithmically brand emphasis will peak in the next year or two as Google comes to appreciate that they have excesssively consolidated some markets and made it too hard for themselves to break into those markets. (Recall how Google came up with their QDF algorithms only *after* Google Finance wasn’t able to rank). At that point in time Google will push their own verticals more aggressively & launch some aggressive public relations campaigns about helping small businesses succeed online.
Once Google is the merchant of record, almost everyone is just an affiliate, especially in digital marketplaces with digital delivery.