Everything you need to know about SEO, delivered every Thursday.
Quantifying Brand Bias In Search Results With Rand Fishkin Of Moz
Like many, I enjoyed Rand Fishkin‘s keynote, The SEO Revolution Will Not Be Televised, at SMX West this past March. Rand is a dynamic speaker and a smart marketer; and in my estimation, he deserves most of the attention he gets.
That being said, I was taken aback by his claim during the presentation that “the long-awaited dominance and bias of search results toward brands is here.”
At Resolution, we work with some of the biggest brands in the world, and if Google were really favoring brands, we wouldn’t be growing as quickly as we are, with more brands than ever investing in SEO.
This is also the trend globally with 47% of marketers at companies (and 69% in agencies) planning to increase SEO budgets in 2014, and just 5% planning to decrease.
Why would brands need to continue to invest so heavily in SEO if Google is naturally biased toward brands? Shouldn’t they be investing in branding instead? From my perspective, the answer is obvious. These brands aren’t investing exclusively in branding because it takes more than that to get visibility in Google Search.
Furthermore, two years ago, I presented contrary evidence to this old claim by Aaron Wall at SMX West — and from my standpoint, nothing’s really changed since then.
Rand presented a graph from Mozcast showing an approximately 27% increase over the last two years in the percent of queries containing the top biggest sites on the first page in Google, so I reached out to him to get a better sense of what he’s seeing.
Clearly, having argued against this in the past, I’m very skeptical that this brand bias exists. But I’m approaching this as a scientist and not a zealot — so if brand bias is real, and the data exist to prove it, I would gladly change my earlier position.
Rand’s Advice Is To Be A Brand. But…
First, having talked to Rand and heard his presentation, I think that his overall message is good. Many times when the topic of brand bias is brought up, it’s to point out Google’s hypocrisy, or to bellyache over losing out to brands for competitive keywords; but, Rand is actually telling businesses to play the game and be a brand.
We have to be realistic about the keywords that we target — and competing with a site that has all the advantages in terms of awareness, budget and resources is anything but realistic. Will Reynolds of SEER exemplified this well during a presentation for IRCE Focus: Web Design in Orlando in March, when he asked an attendee to explain why Google should rank his lesser known dress store ahead of Macys.com for the keyword [dresses].
Instead of trying to beat these big brand sites with manipulative tactics that are short-lived, we should be focusing on improving the quality of our sites and our products to build awareness and organic links and shares. We should be building our businesses and our brands, and not be obsessing too much over metrics that are meant to be a proxy for brand strength, popularity and authority.
To be clear, I am in full agreement with this message. But that doesn’t change my stance on brand bias. I still hold that a brand bias probably doesn’t exist — and if it does, is likely more due to correlation than causation. After all, some of the biggest brands in the world are not doing well in Google search despite their recognizable brand names. Let’s not forget, too, that brands can spam — and when they do, they’re penalized in Google search, just like everyone else.
What Is A Brand Bias?
So, what do we mean when we say that Google search is dominated by brands? The definition of branding as “the marketing practice of creating a name, symbol or design that identifies and differentiates a product from other products” is vague enough to include just about anything when it comes to websites. I asked Fishkin to clarify what he meant specifically with his presentation:
My intent here is to say that for a long time, SEOs have been predicting that Google would reward brands over non-brands. Their executives have spoken multiple times about how “brands are how you sort out the cesspool” and I think there’s strong internal belief at Google that sites engaging in proactive SEO shouldn’t have an advantage over those who simply provide great content and great products. Algorithmic changes like Penguin, Hummingbird and Panda (along with many smaller updates) have increased Google’s general bias toward giving brands greater visibility in the search results.
However, alongside that algorithmic bias, there are also lots of second-order impacts that are helping brands. Google being able to use social signals, user and usage-data signals, and a wider variety of inputs in their rankings means that more good things brands do get recognized by Google. It’s also true that in the past five years, many brands have finally invested in SEO more seriously, helping their own causes.
I also asked him about Rap Genius’s re-inclusion in Google search after its 10-day penalty, and he introduced one more aspect of Google’s bias toward brands:
Rap Genius is an excellent example of a brand being treated differently than a small site. When they encountered a penalty for manipulative action, it lasted a short time and they came back quickly. Any SEO who’s worked with smaller sites (or less-recognizable brands) can tell you that penalties often last months, years, or even indefinitely when you’ve done something like what Rap Genius did. Other examples include folks like Expedia, BMW, and Google themselves — they may be penalized for manipulative actions, but they’re slaps on the wrist compared to what Google typically does.
So, for Rand, it seems there are really five factors contributing to this brand dominance in Google Search:
- An algorithmic bias that gives brands greater visibility in the search results
- A stated preference at Google for strong brands in search results that users expect to see, which eventually gets realized through updates
- Google’s ability to utilize a wider variety of inputs that will theoretically help brands, and not necessarily non-brands
- The fact that brands are increasing their investment in SEO, making it more difficult for non-brands to compete
- The tendency of Google to be more lenient with brands when it comes to penalties.
This is consistent with what I’ve heard from many others who have made the argument in the past.
So now that we’ve defined what a brand bias is, let’s look at each of the five components closely.
1. Bias In Algorithm Updates
We’re all aware that small business owners have been hit by Penguin, Panda and other updates disproportionately. However, this doesn’t mean that brands haven’t been negatively impacted by Google updates, too.
Remember Demand Media? Their IPO led to a market cap around $1.5 billion in 2011, when interest in the brand was at an all-time high. Panda decimated them, as it has other, smaller brands that created thin content in order to get visibility in Google search. Panda also hurt popular car magazine Motor Trend, and Yahoo-owned Associated Content (now Voices).
Brands got hit by Penguin, as well. Dish Network saw their traffic reduced by 27% when Penguin came out. Salvation Army, REEDS jewelers, and many other well-known brands also lost significant search visibility as a result of the Penguin update.
Certainly, there were more small businesses and affiliates that were wiped out by these major algorithm updates, and this is, of course, unfortunate. But Google’s made no secret of targeting thin affiliate content, and there are proportionately more affiliates and small businesses than there are big brands.
Furthermore, big brands, in my experience, are generally more risk averse when it comes to doing things that can harm their reputation, and may have more to lose and less to gain from the type of short-term tactics used to game Google rankings that Penguin and Panda were meant to clean up. These things alone could explain why more smaller brands than larger brands were affected by these updates, without introducing the complexity of a brand bias.
2. Bias At Google Toward Brands
Yes, Eric Schmidt did say, “Brands are the solution, not the problem. Brands are how you sort out the cesspool,” while addressing a room of publishers and editors in late 2008. What not many people are aware that he said right after he said that, is this:
We don’t actually want you to be successful. The fundamental way to increase your rank is to increase your relevance.
The sensational headline is the one that gets the press, but it’s not the whole truth.
And this is not the only time that Google employees have denied that there’s a brand bias behind the changes that they make to the algorithm, or that they treat brands differently than other sites when it comes to penalties. Matt Cutts, for example, told us in 2009:
But you know inside of Google, at least within the search ranking team, we don’t really think about brands. We think about words like trust, authority, reputation, PageRank, high quality.
Given that I’ve been seeing a lot of presentations like Rand’s, and a lot of articles like this one that go so far as to say that “branding is now a major ranking factor,” I thought Google might like to go on record with some revisions. However, when I gave them the opportunity for this article they said they don’t have anything to add at this time.
This particular issue probably has something to do with the fact that most SEOs don’t always trust Google. But, if you trust them enough to believe them when they say “brands are how you sort out the cesspool,” you can’t legitimately say you don’t trust everything else they’ve said on this topic, which is usually contradictory.
Google does have a bias toward trust, authority, reputation and high quality, but those things aren’t the exclusive province of brands. It’s fine if you associate building a brand with increasing the quality of your site, but you should understand that they’re not always the same thing.
3. Google’s Diversification Beyond Link Signals Reward Brands
The idea here is that, as Google pays less attention to link data and expands into social signals, user data, entities and the like, brands will benefit the most. I’m not sure this is the case.
When you look at what types of entities are most popular in Facebook and Twitter, musicians by far have done the best in terms of acquiring fans and followers in social media.
And even if you subscribe to the very broadest definition of “brand” and say all of the top sites are brands in some sense because they have a name that differentiates them from others in the space, you still run into non-brands that are doing pretty well without that name.
For example, Texas Holdem Poker, owned by Zynga, is an entity, not a brand, and is the #15 page with more than 70 million Facebook followers. Likewise, basketball is liked by 57 million people on Facebook, football (i.e., soccer) by 44 million people and pizza by 33 million people; yet, none of them are brands. Pizza Hut, which is #94 on the 2013 top brand list, by contrast, has one-third the Facebook followers that the non-brand pizza has.
Certainly brands are entities, and entities are poised to do better in Google with Hummingbird and Knowledge Graph; but, if brands are doing better in search because they’re entities, the fact that they’re brands is almost incidental.
I asked Rand if entities could be at the heart of what he’s seeing, rather than brands. “I’d agree with that,” he replied, “though I’d add the caveat that powerful/influential brands that are well known get additional signals that may help prop up their rankings beyond just the entities features.”
4. Brands Are Investing In SEO
This is the one area that I will not dispute at all. Brands of all shapes and sizes are, in fact, investing heavily in digital marketing, including SEO.
Of the brands identified by branding agency Interbrand as being the top brands in the world, 66% are doing SEO in some form on their website, which anyone can find on LinkedIn by doing a search on the brand name and the term SEO. A lot of them have dedicated employees and agencies doing SEO full-time. It’s worth noting that this number could actually be higher than 66%, as many companies had digital marketing experts with SEO experience, but didn’t specify that they do or have done SEO on the site.
That’s not to say that the SEO is working just because they’re investing in it to some degree. Of the brands that identified themselves as doing SEO, 63% lost visibility in search engines since 2012. I’m sure all of these people have been hired to turn it around, but it doesn’t happen just because a company starts making an investment.
But yes, they’re investing. SEMPO didn’t release this level of detail in its State of Search report this year; but last year, they reported that 5% of respondents are spending more than 3 million on SEO, up from 1% the year before.
We can’t tell from this graph how many of these sites making this type of investment are big brands as opposed to big sites, or even small brands with access to capital; but, it’s likely that many of them are larger than those brands represented by the average Search Engine Land reader.
5. Google Leniency Toward Big Brands
I’m not in total disagreement with this one. On one occasion in my 14 years doing SEO for large brands, I have seen a larger brand get access to engineering that smaller brands don’t necessarily get. But more often, I’ve seen large brands get told to post in the webmaster help forums, just like everybody else. It’s inconsistent.
Like a lot of agency SEOs, I’ve done work for some of the biggest brands on the planet, and I have never had a meeting with Matt Cutts nor anyone at Google engineering to discuss how they can rank better. When I’ve seen it happen for clients, it’s an anomaly — as in I’ve seen it happen once.
With Rap Genius, which is not so much a brand as much as a venture capital-backed startup, Google may have been more lenient than they have with others that have committed similar violations, and that’s a shame. However, Rap Genius is still getting crushed by AZLyrics in the category, and not because of branding as no one really looks for either of them when it comes to searching for lyrics.
The point about penalties being a slap on the wrist is not entirely true. When large brands get deindexed, even for two weeks, they can lose more money than many small businesses will make in a lifetime.
A few years ago, a company came to us because their large agency had been buying paid links and got their site listed on page six for navigational terms for a month. They estimated it cost them a million dollars in sales. It wasn’t well publicized, because people who work for large brands sign NDAs (and have no incentive to announce it to the SEO community, anyway), but it happened. And if you think a million dollars is a slap on the wrist, you’re making more doing SEO than I am.
But yes, Google is somewhat more lenient toward brands when it comes to penalties. And this is completely understandable because of what Rand said in his presentation. If Google looks bad taking you out of the search results — if people who are searching on Google can’t find what they’re looking for and search quality suffers — they would be foolish to take you out of the index for longer than they have to.
Think of it this way: there’s a new hamburger joint in my Northwest Chicago suburb that I like called The Sandlot. They are a brand, but not a big brand. Most of you have never heard of them. This is their search volume in Google Trends this year:
There’s not enough to register (which is a shame; try their Two-n-two).
And this is the same chart along with the search volume for Five Guys, which a lot of people say their burgers taste like:
If Google makes The Sandlot invisible in search results (which is not hard because they have no website), a few people like me would wonder where they were and might walk over and tell them to hire an SEO. Most people would probably search for another burger place.
If Google makes Five Guys invisible in search results, a lot more people would care. Some of them would wonder why Google’s not returning the information they want when they search for the brand, and some of them might even use Yahoo! or Bing to search instead.
If you think that would hurt Google’s search quality, what do you think would happen if they took out McDonald’s? All these people would be penalized for using Google, in addition to the brand:
So yes, there is a brand bias, in a sense, because Google needs to serve users the results they’re looking for, and users search for brands. It’s the same scenario with other entities, but brands are unique in that they can’t easily be replaced with another site.
What’s Behind The 27% Increase In Two Years?
So, if not brand bias in an algorithmic sense, what could be behind the 27% increase in percent of queries containing the “Big 10” on the first page of Google.com over two years? Any number of things, actually.
First, let’s make sure we understand what we’re looking at. The chart from Mozcast that Rand presented shows what percent of the thousand queries include one of the following “Big 10″ domains on the first page in Google for that query.
Here’s the big ten for today, May 8, 2014, although the list is dynamic based on which sites have the most visibility in Google US:
So, for the 1,000 keywords that Mozcast tracks, at its peak, these domains could be found on the first page in 16% of these queries, and at the lowest point they were closer to 12%, which when Moz used actual figures was an increase of about 27% over two years.
Does this increase really indicate brand dominance of search results? There are a couple of reasons I think it doesn’t:
1. These are big sites and not necessarily big brands. Only three of the top ten sites were mentioned in Interbrand’s list of top 100 brands for 2013. The others are clearly big sites — but, if we’re trying to demonstrate that the bigger a brand is, the better chance it has at visibility in Google search, then the fact that 70% of the “big 10″ sites aren’t big brands by the standards of people who measure brand size doesn’t fare well for this theory.
And, the fact that none of the top ten brands (again, according to the Interbrand list) appear anywhere at all in the big ten makes it even less likely that this growth has anything to do with the size or power of brands.
2. These sites still don’t appear in search results on the first page for 84% of queries tracked. Dominance in this context means most common, and to be most common you should command at least 50%, no? Sixteen percent, up from 12% two years ago, hardly sounds dominant in any sense.
3. We don’t know what percentage of these queries is branded to begin with. We’re only talking about a thousand queries, and if all of them were variations of Facebook then we would expect the percentage of the search results with Facebook.com on the first page to be higher than 16%. If none of them are branded or navigational, that would be another matter; but, there’s nothing to indicate this in the description of Mozcast.
Even if you do think the numbers indicate brand dominance of search results, it could be happening for a number of other, more likely reasons:
A. There’s a high correlation between awareness of big brands and quality, authority and links. That doesn’t necessarily imply causation.
B. Google is rewarding sites that have content that’s relevant to a lot of things, and some of those sites are brands. When I looked at what the top sites in SEMRush and SearchMetrics have in common, what stood out the most was not the fact that some of them were big brands, but the fact that they are relevant to many topics.
Whether you consider them a big brand or not, Wikipedia does not have more than 8x the organic search traffic that the number-one brand on the Interbrand list, Apple.com, has because they’re a bigger brand. They have more traffic because they’re relevant to more things than computers and iPhones.
In fact, according to SEMRush, Wikipedia appears organically on 8x the number of keywords that Apple.com does. And when you look at where Apple.com gets their traffic in SEMRush, much of it is from navigational keywords that aren’t related to Apple. Because of the App Store, Apple has first page ranking on some of the biggest brands on the internet, and it brings them traffic above and beyond what their own brand strength would:
C. Google is rewarding entities, of which brands are a type.
D. Any of the possible reasons I mentioned that Google would be rewarding big brands that aren’t algorithmic in nature (i.e., punishing brands long term would degrade search quality, brands are doing SEO).
I do want to thank Rand for talking to me about what he’s seen that has caused him to think that brands have an advantage over others when it comes to Google search. As I told him, I’ve been reading his stuff since before most people knew who he was, and I’ve always respected his opinion, even when I didn’t agree with it.
Like Rand, I am a skeptic, and skeptics have Occam’s Razor to help them eliminate theories that aren’t necessary for an understanding of the issue at hand. After looking at this issue in depth, I’m afraid I’m still skeptical that a brand bias, in most of the forms that it’s thought to take, exists. The data doesn’t support it. The phenomenon that’s been observed could (and should) be explained without it.
If you started building your brand because of Rand’s SMX West presentation, the last thing I want to do is cause you to change course. As I said, I agree that it’s true that the “find and abuse exploits” era of SEO is ending, and I say good riddance. That has never been relevant to the consumer and has always been spam more than SEO, to me, anyway.
But while you do build your brand, don’t think that will be all that’s necessary to do well in Google search. Unfortunately while there are advantages to building a brand both for your business and for SEO, ranking in Google search is a bit more complex.
In the spirit of collaboration on big ideas, I asked Rand to contribute throughout this article and then gave him a final draft to read before publication. This is his feedback on the draft, with my responses below.
In your second paragraph you made the argument that your company wouldn’t be growing quickly if brands were having success in SEO — that logic seems extremely sketchy and unconnected (far more so than any of the points of mine or Aaron’s you’ve argued against). I think it actually weakens your case.
Good to know. The point isn’t that my company wouldn’t be growing quickly if brands were having success in SEO. We help brands of all sizes have success in SEO. The point is that it takes more than building a brand to have success in SEO. Certainly, as I say, it can’t hurt, and it will probably help to some extent, indirectly. Especially if zero time has been spent building a brand to date.
But the fact that there are many companies that have invested massively in branding and built large, trusted brands and yet still hire SEOs to improve their visibility in search, I think undermines the point that branding is at the heart of visibility in Google.
It seems that you’re saying this is a part of SEO, and that brands that are already brand building should also invest in SEO; but, I can see that point being lost on many executives I’ve worked with in my career. Not necessarily my logic, but many of them might see SEO as unnecessary given that they already have a trusted brand and would shift budget to branding instead. Glad to see you’re not saying this, but I wouldn’t be surprised at all if that’s how it’s interpreted in certain circles.
You said “brands can spam — and when they do, they’re penalized in Google search, just like everyone else.” No one who follows the space would agree with that statement. While big brands do get penalized, the severity of those penalties is massively different from how non-big-brand sites are treated.
I do address that later in the article. The penalties are different, but the duration of time isn’t the only indication of severity, as brands suffer deeply financially and search quality suffers more when searchers can’t find popular brands they’re looking for en masse. I also think most of those in the space who talk about things like this don’t work for big brands, so the fact that what I’m saying might be unpopular is completely understandable.
You point out Demand Media, but I (and many others) would argue that the company did one of the worst jobs in the media world of creating a strong brand for its sites or the content. On that spectrum I talked about, they’d be one of the largest companies with one of the worst brands, and a great example of Google seeking to reward brands by pushing Demand’s crap out.
That’s interesting, actually. I would agree to some extent, but I think that this goes way beyond the basic definition of brand for marketing and underscores why I think you need to use a different word when explaining this.
Demand Media clearly identified and differentiated their product from other products, or they wouldn’t have been able to get as big as they did, and yet doing so didn’t help them at all in Google search. The definition of branding you’re using is focused more on quality, authority and trust; I understand, but from my perspective it’s less confusing if you just focus on quality, authority and trust and leave the brand aspect, which is more broad, out.
I’ll also point out that using individual examples vs. aggregated stats is a very poor way to make these arguments. Just because Google penalizes some brands or because some big companies struggle doesn’t mean that brands, in aggregate, are losing visibility.
That’s why I supplemented it with the Interbrand data. Those 100 brands that branding experts have identified as the biggest brands in the world have, in aggregate, lost visibility in Google search. But the individual examples do illustrate the point that observations of reality conflict with this theory that Google favors brands in all cases.
The Matt Cutts quote: “we don’t really think about brands. We think about words like trust, authority, reputation, PageRank, high quality.” is the epitome of Google favoring brands, since a great definition of a brand vs. a non-brand is one that has consumer trust, authority and reputation (PageRank & quality notwithstanding).
I disagree, and so does Merriam-Webster. The definition I quoted from Entrepreneur.com is simply “The marketing practice of creating a name, symbol or design that identifies and differentiates a product from other products.”
I understand that you’re thinking of brand-building as building consumer trust, authority and reputation, but Google isn’t looking at brand as a ranking factor at all, as some people seem to think. They’re just looking at the byproducts of a good brand, which should be consumer trust, authority and reputation. And they always have been looking at these things without bringing the concept of brand into it.
When you do introduce the concept of brand you complicate things unnecessarily, and run into problems, as building a brand that has consumer trust, authority and reputation, as all the brands on the Interbrand list have, is not enough to do well in Google search.
Sounds like you’re not saying this, exactly, but I’m sure there are plenty of others who represent big brands who might rejoice at the news that brand dominance of Google search is finally here, if only so they could confirm that SEO is dead and shift their budget elsewhere. I’m speaking out against this because I think it would be a shame if they did.
You’re using that quote to prove your point, but it actually works exactly against it. You later say that these aren’t the exclusive province of brands, but they are most certainly elements of a brand and of branding, and they’re a big reason I advocate working towards brand-building as part of your marketing.
And there’s nothing wrong with advocating working toward that, provided your audience understands that Google doesn’t actually look at brands, but looks at things that you consider to be a byproduct of brand building. And provided that they know they could spend a lot of time building their brand and it would still have little impact on their visibility in Google search. From the reaction I’ve heard to your presentation, I don’t think people understand this at all.
I disagree that using Interbrand’s “big brands” list is reasonable, because I’m not just talking about “big brands” but those who invest in branding and have built the elements of a brand — uniqueness, trust, reputation, authority, consumer recognition, etc. vs. those who have not.
Yes, but these are the brands that have invested the most and have built the biggest brands on the planet. The fact that they’re largely decreasing when it comes to visibility in Google search doesn’t fare well for your theory that “the closer you are to worldwide brand, the more benefits (directly and indirectly) you accrue in Google’s system.” These are all not just worldwide brands, but the biggest, most trusted worldwide brands, and should therefore have the most benefits in Google’s system, no?
In this article, Bryson Meunier and Rand Fishkin discuss what a brand bias in Google search means, specifically. Both agree that building a strong brand is important for many reasons, including SEO; but, Bryson doesn’t believe there is sufficient data to show that building a brand will increase visibility in Google search.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.