The Hitwise blog post reported that Twitter accounted for only 0.14% of “upstream” visits to the “News & Media” sites category last week, far beyond similar stats for Facebook and Google. Here are all of them compared, based on that blog post:

• Facebook: 3.64%, the 3rd biggest source of visits to News & Media sites
• Google News: 1.27%, the 11th biggest source of visits
• Twitter: 0.14%, the 39th source of visits

Google News Versus Google

My first issue with these stats is that Google News — only one small slice of Google — is being compared to all of Facebook. That doesn’t seem a fair comparison. Unless Facebook has a Facebook News area (it doesn’t), it seems like you need to compare “whole” Facebook to “whole” Google when discussing who drives traffic.

As it turns out, traffic from “whole” Google leaves Facebook in the dust, as a driver of traffic to news sites. So doYahoo and MSN. If you count traffic from any Google domain (or any Yahoo domain or MSN domain), the stats work out like this:

• Google sites, 20.16%
• Yahoo sites, 18.92%
• MSN sites: 8.76
• Facebook: 3.64%

Specifically, Hitwise looked at the top 100 sites sending traffic to News & Media sites last week, then added up all those that were run by Google or Yahoo or MSN to give the “whole” figures shown above. Each of the first three was listed as “Google properties” or “Yahoo properties” or “MSN properties” but Facebook was not listed that way. However, as Facebook run virtually everything I know of within Facebook.com, my assumption is that the Facebook figure is for “whole” Facebook as well. It’s not clear if MSN includes all Mircrosoft properties such as Bing or not.

Hitwise also sent figures — and rankings — based on the main domain of each company. These were:

• Google.com, 16.50%, ranked 1st
• Yahoo.com, 9.40%, ranked 2nd
• Facebook.com: 3.64%, ranked 3rd

NOTE: Earlier I had stats of 11.03% and 10.98% for Google and Yahoo respectively. These weren’t for the traffic that Google and Yahoo sent to news sites — which is what I had asked for — but instead either the amount of visits Google and Yahoo themselves receive from all sites or that they send to all sites. That caused me to changed the headline of this article from “Stat Rant: Google Actually Trumps Facebook For News & Can’t We Measure Twitter Correctly?” to “Stat Rant: Does Facebook Trump Google For News & Can’t We Measure Twitter Correctly?” until I could get better clarity from Hitwise. Now that they’ve sent me the correct numbers, Google indeed trumps Facebook even more than I originally thought, as a news driver.

Google News Isn’t Google Reader

If the earlier stats got you thinking that Facebook was the new killer news “app” versus Google, there’s more of that today with another post from Hitwise. This one talks about how Facebook users are more “loyal” to news sites than those coming from Google News. We’re told:

• 78% of Facebook users were returning visitors to the top 5 print media sites for the week ending March 6
• 67% of Google News users were returning visitors
• 77% of Facebook users were returning visitors to the top 5 broadcast media sites
• 64% of Google News users were

The headline — “Facebook Visitors Come Back Again And Again” — is meant to follow up on what author Heather Hopkins encountered in feedback when she posted a few weeks ago about Facebook as potentially being the web’s biggest feed reader (and this despite Facebook actually making it kind of hard to use it as a feed reader). Some said “big deal” if Facebook drives traffic. You want loyal readers, people who will come again and again.

So see — Facebook does have loyal visitors! But then again, you kind of expect that from Facebook versus Google News. In Facebook, you can be a fan of a news organization, effectively subscribing to it for updates. That’s all designed to keep you coming back. Google News has over 20,000 sources, and you can’t subscribe to any of them. If you want to be a subscriber through Google, you have to use Google Reader. So what’s the loyalty rate for that?

No idea. That’s probably because the last time Heather looked, when doing her post about Facebook as a feedreader, Google Reader drove only a tiny amount of visits to News & Media sites: 0.1%. Google News sends more traffic, so it has becomes the comparison choice, even if it’s the wrong one, in my opinion.

Heather did cite in today’s report how Google overall is, in terms of loyalty:

I’ve been encouraged by some readers to include Google.com in this series. In most cases, Google.com is the #1 source of traffic to these sites. Interestingly, visitors from Google are less likely to be returning visitors than average for either Google News or Facebook.

No specific figure, but that’s not surprising. As with Google News, people can’t “subscribe” to a source in regular Google. They discover sources that way. And often when discovering them, they have little need to go back to Google.

Revisiting The Search Gap

This phenomenon is something I called the “search gap” back in 2001. Pick your survey, and you’ll discover that searching is among the top internet activity out there. And yet, look at web site stats, and search engine traffic often is not the top traffic driver. Why is there a big gap between what people do and the traffic search engines send? Because once you’ve found a trusted site (including a news site), you may not need to search for it again. As I wrote:

Let’s say you want to buy a particular book. You do a search at your favorite search engine and find a page from Amazon about the book. You visit the Amazon site, like the price and information you are shown, so purchase the book from them. Thanks to search engines, Amazon has gained a customer.

A month later, you need another book. Remembering your positive experience at Amazon, you go directly to the web site rather than using a search engine to find it. Thus, your second visit isn’t credited to search engines. However, it would have never occurred if you hadn’t found Amazon via search engines the first time AND had a favorable impression of the site.

So, once people find trusted sites, they return to them directly for particular needs …. However, because our needs are wide-ranging, we are constantly searching for new things — which accounts for the overall high usage of search engines that other studies find.

It would be a mistake to interpret the search gap as meaning that search engines are not important. They remain a top way users will locate web sites initially and so cannot be ignored. Instead, the real lesson of the search gap is the age-old adage that first impressions count. Make a good impression when people first come to your site via search engines, and they may come back directly to you in the future.

So the Facebook versus Google loyalty stats? They’re interesting, but they’re not necessarily comparable, given that users in both places may be doing radically different things.

Twitter Is Not Twitter.com

Another big issue I had with the traffic to news sites post was that it tried to compare Twitter as a traffic driver by using stats that measure only Twitter.com. That’s a big flaw to me, because so much Twitter activity happens off the site itself. People interact with Twitter through third party applications, or by seeing tweets in things like Google Real Time result or posted on individual blogs.

If you don’t measure these things, you aren’t properly measuring the Twitter ecosystem as a traffic driver — perhaps grossly undercounting it. The posts below go into more depth about this:

• How Twitter Might Send Far More Traffic Than You Think
• Is Twitter Sending You 500% To 1600% More Traffic Than You Might Think?

I asked Hitwise about this and got back:

We are definitely looking into expanding our data and measuring traffic from apps/mobile.

That’s good, but it doesn’t help a company like Twitter now, which gets painted as some type of weak player compared to Facebook. If the stats don’t allow a fair comparison, then don’t make a comparison. Say that you can’t measure the two.

Note that Hitwise isn’t the only one with this issue. I’ve seen plenty of stats from various places that suggest that Twitter isn’t being measured correctly. It’s not a Hitwise-only problem.

As a sidenote, Chitika has new data out that doesn’t have traffic figures but does try to measure the intent of what people from various social networks are after, from when they leave those networks. News is tops to those from Twitter, though exactly what “Twitter” is and how this is measured isn’t said. Nor is it clear whether something like “Tech” might include tech news:

• Twitter: 47% head to news sites
• Facebook: 28% head to news
• Digg: 18% head to news
• MySpace: News not in top five categories

Referrer Vs Upstream Traffic

One proxy might be to look at the amount of traffic that Bit.ly sends to News & Media sites. Bit.ly is the default URL shortener for tweeted URLs. Regardless of where you encounter a link in the Twitter ecosystem — on Twitter.com itself, in a third party app and so on — many of these will route you through Bit.ly. That means measuring Bit.ly as a driver to News & Media sites might give you a better idea of how Twitter is doing.

I asked for those stats, but I didn’t get the right ones from Hitwise. Instead, they sent me a chart showing how many people go to Bit.ly after Twitter. I’ll try again on this. But that also brings up a key difference to how Hitwise measures what drives traffic and what analytics tools do.

NOTE: Since I wrote this, I’ve received the stats. In Feb. 2010, Bit.ly sent 0.0043% of traffic to News & Media sites. I suspect that means it’s not a very good proxy for Twitter’s traffic overall.

Most browsers (Firefox, Internet Explorer, Safari, Chrome and so on) are configured to tell a web server the last page they viewed before they arrived at a web site. This is called the “referrer,” though often it’s called the referral source as well. Referrer information allows site owners to know the search terms used to reach their web sites. They allow them to tell if Facebook is a big traffic driver, or Google or other sites with a good degree of accuracy in many cases.

Hitwise isn’t reporting referrer information. It doesn’t have this. Instead, it depends primarily on data through deals it has with internet service providers. It can see from this data what people do before and after they go to a particular web site. The before is called “upstream” traffic and the after is called “downstream.”

This isn’t as precise as referrer data. Consider this. You have a browser window that opens automatically to Facebook, when you start. You scan your Facebook feed, don’t click on anything, then decide to do a search for the iPad. You type in Google into your browser’s address bar, do a search from the Google home page when it loads, see a link to Apple and click on that.

Hitwise records things this way:

Facebook –> Google –> Apple

In reality, here’s what happened:

Facebook

Google –> Apple

Facebook drove no traffic to Google. But in Hitwise’s system, since it was the last site someone saw before Google, it would be counted as an “upstream” driver.

By the way, if you want to play with this more yourself, try this site. If you click on that link, you should see that the “referer” box will get filled with the URL of this article you’re reading. If you then copy and past the URL and go to it directly, you’ll get no “referer” shown.

Looking Beyond The Stats

In general, I’ve felt there’s been an increasing push from various places — not just with Hitwise — to punch out stats. These get quickly reported often without much analysis. From this, we end up with facts that often aren’t really true.

For example, we all know from stats earlier this year that 44% of Google News readers simply scan headlines without going to sites, right? Right?

That’s not actually true, as my 44% Of Google News Readers Only Scan Headlines? Maybe Not! article gets into. But despite this, I keep hearing that stat repeated over and over again.

Meanwhile, Facebook is once again reported to be the most visited US web site, beating Google. But then again, that’s not measuring all of Google, such as Gmail — so is Facebook really bigger? (See Hitwise: Facebook (Sort Of) More Visited Than Google On Christmas for more on this).

In addition, Facebook auto-refreshes its pages. If I leave a Facebook page up, I’m probably generating “visits” all day long even though I’m not always looking at that page. Does that skew comparing to sites that don’t auto-refresh?

There’s no easy solution to much of this. Stats have been glossed over, twisted or had comparison issues as long as we’ve had stats. I can be as guilty as anyone of pushing a stat quickly without delving into it more. Maybe the trend is no better or worse than ever. But I’ll be trying to dig even more doing forward, and I hope others will do the same.

For related discussion, see here on Mediagazer.

“The clickthrough rate (CTR) was 2% on Friday, but 3% on Saturday. That’s a 50% increase. 50% is a lot, right?,” they ask. Well, it certainly is for income tax rates, but not necessarily for differences in clickthroughs. What if Friday saw 2 clicks from 100 impressions and Saturday saw 3 clicks from 100 impressions? Doesn’t sound so impressive anymore, does it?

Part of the problem is that it’s simply impossible to tell from that few impressions whether both have an inherent CTR of 2.5% (and you just happened to see 2 clicks for one and 3 clicks for the other) or whether they legitimately have different underlying CTRs.

Imagine a more extreme case: one ad has a CTR of 2% and one has a CTR of 100%. We see four impressions and all get clicks. How likely is this to be the data for the 2%-CTR ad?

Well, if it is the 2%-CTR ad’s data, then there’s a 2% chance that the first impression would generate a click. That’s about the same as the chance of randomly drawing the ace of spades from a well-shuffled deck of cards. And there’s a 2% chance that the next impression will generate a click, which is about the same as reshuffling that deck and then randomly drawing the ace of spades again (without any sleight-of-hand trickery).

So, the chance of seeing 4 clicks from 4 impressions for a 2%-CTR word must be very, very small, but (please take a minute to convince yourself of this, if you need to) it’s not absolutely zero. Even an ad with only a 2% CTR still might possibly generate 4 clicks from 4 impressions. It’s improbable, but not impossible.

That is why statisticians rarely seem to give a straight answer to whether two ad’s CTRs are different or not. “Statistical significance” is not really a Yes or No situation, it’s just the probability of seeing a certain sequence of events (like four ace-of-spades in a row) not happen purely by chance. Every new impression increases the certainty in our answer, but there is no specific amount of information that seals the deal.

By convention, statisticians often set an arbitrary cut-off of “5% chance of being explained purely by randomness” for classifying whether or not a difference is “statistically significant” or not. That’s why when a magician declares that he’ll pull a certain card from a deck, and then actually does so, the average geek in your life will joyously exclaim, “That’s statistically improbable!” We know that there’s less than the 5% cut-off chance that that card appeared purely by luck.

Imagine now that we have two ads, Ad A, for which we have observed a 2% CTR, and Ad B, whose observed CTR is shown on the x-axis of the graph below. The graph shows the number of impressions (per ad) we must see to be 95% certain that the two ads have different CTRs.

If Ad A has seen 2 clicks from 100 impressions (2% CTR) and B has seen 14 clicks from 100 impressions (14% CTR), then we can be more than 95% certain that Ad B’s CTR is higher than A’s. If the observed CTR of Ad B is only 3%, then we actually need nearly 4000 impressions each to be 95% certain that Ad B performs better. That’s why the difference in observed CTRs between the Friday and Saturday ad performance wouldn’t look so impressive if they only had 200 impressions between them.

As the CTR of Ad B approaches 2%, it takes staggeringly more and more data to differentiate the two ads. Trying to tell a 2.00% CTR ad from one with a CTR of 1.95% (or 2.05%) takes more than a million impressions each. And, if the two ads perform identically, with exactly a 2% CTR, obviously even an infinite amount of data couldn’t tell them apart.

Though the concepts I’ve described above are (hopefully) now very clear, unfortunately some of the web-based tools for differentiating CTRs seem to have disregarded them completely.

For example, if one ad got 1 click with a 25% CTR (that is, 4 impressions) and a second ad got 2 clicks with a 100% CTR (that is, 2 impressions), Splittester.com by Brian Teasley and Perry Marshall says: “You are approximately 99% confident that the ads will have different long term response rates.” 99% confident from just 6 impressions?! No, I’m not. If I flip a coin 4 times and get 1 “heads” and another coin 2 times and get 2 “heads,” I wouldn’t be 99% certain that either one of their per-flip chances deviate from 50% at all.

Supersplittester.com, a similar site by Dr. Glenn Livingston (I presume), has similar deficiencies. For the case of Ad A, with 4 impressions, 1 click (25% CTR) and 1 conversion (100% CR), and Ad B with 2 impressions, 2 clicks (100% CTR) and 1 conversion (50% CR), the site tells me both that “Ad B has a higher CTR than ad A (99% Confidence Level)” and that “Ad A has a higher conversion rate than ad B (80% Confidence Level).”

Frankly, the only thing I have 99% confidence about is that Teasley, Marshall and Livingston should have a second look at their computer code to see what’s going wrong.

In the McKinsey Quarterly, Google’s chief economist Dr. Hal Varian said: “I keep saying the sexy job in the next ten years will be statisticians. People think I’m joking, but who would’ve guessed that computer engineers would’ve been the sexy job of the 1990s?”.

He’s absolutely right. In 1990, only a handful of geeks knew what a “homepage” or an “email” were. Ten years later, few people didn’t know. Likewise for search marketing, even basic concepts like determining a confidence interval to identify statistical significance can still seem esoteric. But the industry is quickly realizing that being able to do these calculations is not just for geeks anymore.

The Foundem complaint/brief with the FCC argues that Google favors its own products and thus “search neutrality” is required to prevent Google from harming competitors accordingly:

Search engines have become the Internet’s gatekeepers and are arguably as essential a component of its infrastructure as the network itself. Google’s overwhelming dominance of search and search advertising, coupled with its ability to arbitrarily penalize rivals and systematically favor its own services, makes the need for search neutrality particularly pressing.

This is similar to an argument recently made by a consortium of minority media organizations:

The Minority Media and Telecommunications Council, representing 16 national organizations, [  ] raises the provocative issue of whether net neutrality requirements should be applied to search engines.

A couple of deeper looks at Foundem’s organic search position argue that the company’s ranking on Google may be based on bad SEO practices rather than any manipulation of results by Google. See for example Tribble and EConsultancy for more.

Foundem is requesting some sort of remedy or remedies plural to prevent Google from allegedly “favoring” its own products and services:

There is an urgent need to constrain Google’s domination—either through competition or through regulation.  Despite the recent US and European approval of the search alliance between Yahoo! and Bing, competition in search is unlikely to be sufficient, at least in the short term. Carefully considered regulation is therefore needed to codify the Network and Search Neutrality principles of non-discrimination and transparency and apply them equally across the entire Internet ecosystem.

Let’s ask, hypothetically, what all this might mean in practice. It might mean, for example, not allowing Google to refer traffic to its own properties (e.g., Maps) or it might mean promoting competitors alongside its properties — or both.

In one sense this is not unlike what the EU required Microsoft to do with giving people “browser choice” on new Windows machines. However it would be difficult to implement as a practical matter because it doesn’t involve a one-time settings choice; it involves billions of SERPs and their configuration. Would Google need to list competitors’ sites side-by-side with its own results on SERPs in every instance where Google has a corresponding property?

Google does do this today in a few instances:

But who would get these coveted links? Would only the largest sites in a category be “entitled” to this sort of SERP presentation? Would there need to be a rotation of companies in segment? That would create another round of fighting;  you can imagine the problems.

The Foundem brief cites the example of Google’s removal of links to Yahoo Maps and MapQuest as evidence of what happens when Google “favors” its own properties:

The preferential placement of Google Maps at the top of Google’s search results, which began in May 2007, played a significant role in unseating MapQuest from its position as the US’s leading online mapping service.

For those who don’t recall, Google removed those links to competitive mapping sites in January of 2007. Foundem anticipates the “superior functionality” of Google Maps argument as the explanation for MapQuest’s decline:

Some will suggest that the rapid rise of Google Maps can be attributed to superior functionality.  But an analysis by Heather Hopkins of Hitwise found that many more users were still actively searching for MapQuest than for Google Maps at the time when Google Maps’ traffic was surpassing MapQuest’s; she concluded that Google Maps’ rising traffic was the direct result of its preferential placement in Google’s search results . . .

There’s no question that Google’s ability to refer traffic to its own mapping site has helped Google Maps overtake MapQuest. The truth, however, is much more complicated than Foundem’s contention.

MapQuest failed to add new functionality and invest in the property for more than two years at a time of intensifying competition and feature innovation at Google and Microsoft. Similarly Yahoo didn’t invest in its mapping property either and has similarly suffered. MapQuest brand strength has sustained traffic to some degree and more recently MapQuest has upgraded its functionality. But the truth is that MapQuest was and is generally inferior to Google Maps in most respects.

The MapQuest “case” illustrates the problem at hand. Yes, Google has referred traffic to itself but it also offered a more compelling experience than MapQuest. YouTube is another difficult example; it gets traffic directly from Google but is also the category leader by far. Is Google perpetuating that leadership via Universal Search?

Beyond listing competitors’ sites a potential remedy, as mentioned, would be to prevent Google from referring traffic via Universal Search to any of its own properties. OK. Would that principle then equally be extended to Bing and Yahoo or would it just apply to Google? If it were only applied to Google that would represent a problem of fairness at the very least. Indeed, if the no self-referring rule were to be imposed across the board — no search engine can refer traffic to a property owned and operated by its corporate parent — the search experience would arguably be compromised and innovation might slow down.

Furthermore a 2003 federal court decision (interpreting Oklahoma law) may make all of this “search neutrality” or proposed algorithm regulation moot. According to an opinion in Search King, Inc. v. Google Technology, Inc. Page Rank and search algorithms more broadly are “subjective opinions” entitled to First Amendment protection:

The Court must, therefore, determine whether Google’s manual decrease of Search King’s PageRank was malicious and wrongful, and was not justified, privileged, or excusable. Google asserts that its actions cannot be considered wrongful because PageRanks constitute opinions protected by the First Amendment . . .

Other search engines express different opinions, as each search engine’s method of determining relative significance is unique. The Court simply finds there is no conceivable way to prove that the relative significance assigned to a given web site is false. Accordingly, the Court concludes that Google’s PageRanks are entitled to “full constitutional protection.”  . . .

Having determined that PageRanks are constitutionally protected opinions, the Court must now consider whether, under Oklahoma law, Google is immune from tort liability arising out of the intentional manipulation of PageRanks. In Jefferson County, the Tenth Circuit concluded that under Colorado law, protected speech cannot constitute improper interference in the context of a claim for tortious interference with contractual relations . . . The Court finds that Oklahoma law compels the same conclusion in this case.

Google is dominant in search, but it does have competition and many functions formerly taking place in search (e.g., content discovery) may shift to social media and Facebook in particular. So the idea that Google is the only doorway into content is myopic and incorrect as a practical matter. 

Behind many of the complaints filed is a sense of profound frustration. In addition, there’s also a subtext of entitlement to some sort of position in search results. That argument has even been explicitly made by newspaper publishers who similarly complain of Google’s too-central role in content discovery online. Unable to compete effectively some of these companies are now turning to courts and regulators for relief, alleging that Google is doing improper things or proposing novel theories such as “search neutrality.”

Yet by its very nature “search neutrality” is something of an oxymoron.

There is a problem when one company is too powerful in any industry. But when that comes about not as the result of market manipulation but rather consumer preference it raises a philosophical question about what to do and how to “restore” or promote competition. However the potential remedies, implied by the Foundem brief, are problematic.

I’m no free-markets zealout but I do believe that if courts and regulators start dictating how search algorithms should operate it starts to destroy search and its utility.

Keywords and search queries are two very different entities. While keywords are the inventory an advertiser buys, search queries are the actual words people type into a search engine. The distinction is important, because search queries typically cover a significantly longer tail than keyword buys, and there is a lot of value in mining search engine query data.

To take a deeper dive and look at the way that head, mid-level, and long tail search queries perform in paid search accounts, we took a random sampling of WordStream client accounts and analyzed the aggregate cost, number of conversions, and cost-per conversion across the sample. The results were very interesting, and to some extent would confound the zealots on both the “short keyword list, no such thing as the long tail” camp and the “the highest value lives in the lowest frequency of clicks” contingency.

Which types of search queries drive more traffic for less spend?

We analyzed about a million dollars worth of spend, spread across roughly 15 million impressions. A few notes on the data:

• The advertisers we looked at were a randomly selected mix of lead-gen and e-commerce companies who are small to mid-sized advertisers.
• We analyzed the total spend, number of conversions and cost-per conversion against a series of segments.
• We divided the data based on the number of clicks per query: 300+, 100-300, 6-100, and 0-5.
• All of this activity was on Google’s AdWords.
• We were also careful to mix in both long-term and new account data, so that the results wouldn’t be overly skewed by the way our software builds campaigns.

It’s also worth noting that paid search is a multi-billion dollar industry, and this sample size is by no means definitive, but there are nonetheless some very interesting insights to be gained.

First let’s look at the distribution of spend across head, mid, and long tail terms:

The interesting thing about this analysis is that a vast majority of the spend in the accounts in question was directed toward long tail search queries. The aggregate number of search queries in the accounts analyzed was greater than the aggregate number of keywords by about 4 to 1.

This leads to a misconception about the distribution of searcher intent. Many advertisers think that most of the traffic is being driven by a short list of keywords, when in reality these advertisers are spending significantly more of their budgets on the long tail.

Next, let’s look at the distribution of conversions across the same segments:

This data maps pretty consistently with the first data series, showing that these advertisers are driving a majority of their conversions from queries with 0-5 clicks. Additionally, 90 percent of the conversions for this data set were driven by queries with 1-100 clicks attached!

Probably the most interesting data set of all is the cost-per conversion across head, mid and long tail queries:

The cost-per conversion in the 6-99 click segment is less than half of the CPC for more competitive terms, while the particularly low-volume terms (0-5) are roughly two-thirds higher per conversion than the 6-99 click segment.

So what does all this query data mean?

I think the story this data is telling is that:

• There is enormous value in the long tail of search queries
• Query data has a longer tail than keyword data
• To achieve high ROI, it’s crucial to aggressively mine negative keywords and to effectively target more specific search queries.

Here are three key takeaways from our analysis:

There is value in the long tail. The fact that 0-5 click queries and 0-100 click queries comprise so much of the cost and total conversions for these companies is obviously an indication that the long tail of search queries contains a lot of traffic and opportunity.

There’s a longer tail for query data. The fact that the total number of keywords is roughly a fourth of the total number of queries is the result of numerous long tail search queries “hiding,” rolled up under a single keyword. A cursory glance at any of the advertisers accounts might lead one to believe that most of the traffic is being driven by a short list of keywords, when in reality a series of broad matched keywords are matching against more specific, less popular queries.

Negative keywords also have value. The final key takeaway is the product of the discrepancy between the cost-per conversion for 0-5 clicks versus 6-100. While 0-5 clicks drove a majority of the cost and conversions across these advertisers, the 6-100 segment had a significantly lower cost-per conversion.

In theory, the 0-5 segment should be more specific variations of words and phrases. This should mean that by and large they’d be better targeted and more likely to convert. So where’s the gap?

The relatively poorer results in the 0-5 segment are largely because those queries aren’t hyper-targeted: many are the result of broad match aggressively pushing impressions at people that are matched to much broader keywords. Additionally, many irrelevant terms are rolled up in those broad keywords, driving up costs by matching terms people didn’t realize they’d be bidding on to ads for what are now very poorly targeted ads.

Where to go next with search query data and the long tail

While this data certainly isn’t representative of every single advertiser’s experience, it does reveal the need for a strategic approach to combining broad match with negative keyword discovery and implementation, and it emphasizes the importance of discovering, grouping and targeting specific search queries.

Since this article was spurred by a series of recent discussions surrounding the long tail, I thought I’d include links to the articles that inspired this inquiry:

• The Pundits Are Wrong: Don’t Cut Off Your Tail! – George Michie wrote a piece on the value of segmentation & granular “long tail” targeting within paid search campaigns, which followed Rimm Kauffman’s own empirical study on the difference between head and long tail keywords.
• PPC And The Importance of The Long Tail – Additionally, George hosted a great webcast on the subject which went into greater depth on the value of broad match and negative keywords, and offered some advice on how best to implement.
• Broad Match + Negative Keywords = A Profitable Long Tail – Brad Geddes offers a nice synopsis of the value of broad match, and the importance of pairing it with a strong negative keyword strategy.
• How To Group Your Keyword, Plus a Q and A With WordStream’s Larry Kim – In addition to effectively using negative keywords, you’ll also need to intelligently segment your keywords and find a means for mining query data to assign positive keyword candidates. In this article Josh Dreller interviews the founder of my company, Larry Kim, about how best to attack the issue of grouping and segmentation.
• Shifting PPC from Low to High Resolution – Craig Danuloff of Click Equations writes frequently on the subject of gaining deeper insight into your paid search campaigns through closer attention to search query data. This is a nice overview of the Click Equations “high resolution” approach
• PPC Man Drowning: Too Many Keywords – Andrew Goodman offers some interesting counter-points to the premise that including large numbers of keywords in a paid search campaign is beneficial
• Ding Dong? The Truth About the Life And “Death” of The Long Tail – In this article from May, I walked through some of the benefits of including more granular keyword targets within a paid search campaign, and attempted to draw out the differences between keywords and search queries and the significance of those differences in a bit more detail

For those unfamiliar with Google Profiles, see our Hoping To Improve People Search, Google Launches “Profile Results” post from last year. It’s a good introduction to how Google Profiles were repositioned as a way for people to have virtual business cards about themselves, hosted by Google and designed to help ensure that individuals could be found in Google’s special “profile results.” Indeed, Google even did a promotion that transformed Google Profiles into real business cards.

Buzzjacking Google Profiles

By default, Google Profiles were set to show the “About me” tab, as you can see from this screenshot of my profile taken last year:

Today, if you’ve opted into Google Buzz, that’s changed. Rather than “About me” showing, it’s your Buzz activity that appears by default. Here’s how my profile looks now:

That’s right. My profile has been Buzzjacked, as I like to call it.

It makes no sense. If someone is searching for me (or someone else) and finds their Google Profile in search results, why do they want to be shown Buzz activity over a general introductory page?  Or why does Buzz get the nod over Sidewiki comments, which also show on a profile page? Or why doesn’t Google My Maps activity get a tab on the profile page. After all, My Maps is linked to your Google Profile.

How’d Buzz Get Into My Google Profile?

Now let’s first review how Buzz ends up on your Google Profile in the first place. That’s important for understanding how to remove Buzz from your profile, if you decide you don’t want it there.

Anyone with a Gmail account is now offered Buzz. Everyone trying to reach Gmail since Buzz launched will see, at least once, a start-up screen prompting them to take part in Google’s new social service. Say yes on this screen, and you’ll see a list of people that you’re suggested to follow:

Say yes to some of these, and you’ll get another screen prompting you to create a Google Profile, if you don’t already have one, so you can participate in Buzz.

Did you do all that? Then now you’ve got a Google Profile that, by default, shows your Buzz activity.

Just Say No To Buzz? Kiss Your Profile Goodbye

Change your mind and decide you don’t want to be part of Buzz? There are the new Buzz Choices options, which you reach by selecting the Settings link atthe top right-hand side of Gmail:

The first option — to show or not show Google Buzz in Gmail –  doesn’t really opt you out of Buzz. It just hides the Buzz link from showing up in Gmail. If you’ve made any Buzz posts, this doesn’t remove them from your public profile Instead, you have to delete each post individually. Then you have to disconnect any linked sites. Then you have to unfollow anyone you were following. A lot of work, to get the Buzz off your profile.

The easier way unfortunately goes all medieval on your profile. Select that “Disable Google Buzz” option, and you’ll really be taken out of Buzz. Everything you’ve done will disappear, and your Google Profile will get nuked, as well.

It’s crazy. As I’ve covered, Google Profiles existed before Buzz and serve other Google services than just Buzz. Indeed, Google made a big push to get people to set these up last year simply as a way to ensure they could be found for their own names in Google. Killing your profile just because you don’t want to have anything to do with Buzz is overkill.

That’s what’s frustrated Marsha Collier, who was shocked to learn this weekend how her profile had been Buzzjacked. She’d been using her profile exactly as Google had promoted the service last year, as a virtual business card. With the Buzz change, it no longer served that purpose, to her.

Resurrecting Your Google Profile

You can, of course, get your profile back. But you have to rebuild it from scratch, reentering any information you may have had in the profile before Buzz took it over. Once you’ve done that, as long as you don’t reenable Buzz, then your profile will stay Buzz free.

Changes In The Future?

I talked with Google about these issues today. I was told that the feedback so far has been that most people who want to fully opt-out of Buzz also don’t want to have a profile at all — that’s why there’s the “kill Buzz, kill your Google Profile” option. That might change in the future. Also, Google’s aware that some people might prefer to chose which tab shows by default on their profile. So maybe in the future, you’ll be able to make “About me” show rather than “Buzz” by default.

Google’s Continued, Disorganized Social Mess

Overall, the Google Profiles madness is just another sign of how rushed Buzz was and what a mess Google’s social efforts have been in general over the years. There appears to be no coordinated social plan. SideWiki does whatever the hell it wants with Google Profiles, as does Google Maps, as does Buzz. The Buzz comes along and gets the power to wipe out Google Profiles, regardless of whether other services use them. And yet further services ignore Google Profiles entirely.

On the upside, maybe Buzz will be the glue that finally binds all these social efforts at Google together and gets them organized. I also have no doubt the Buzz team is working as hard as it can on feature requests and bug fixes.

Still, things like sacrificial profiles remain frustrating, as does the fact that as we go into our second week of having Buzz, those with Gmail accounts through Google Apps still can’t participate in Buzz. Or that you’re still forced to reveal your Gmail address if you want to be more than a number on Buzz. Or that I still cannot get connected sites to work as advertised. Despite linking our Search Engine Land feed to our Search Engine Land Buzz account, Buzz still refuses to pick up our stories. Ironically, if it weren’t for Twitter, we wouldn’t have a story feed there at all. Buzz recognizes Twitter as a connected site but not our own site!

Sigh.

But what I find most fascinating and useful about the report is what it says about we view and use the internet now. In particular, the use of search engines such as Google received prominent focus in the report. How can we use this report to better understand our online audiences? See the report for full details of the questions and those surveyed. From a search perspective, what I found most interesting about the responses was that:

• Collectively, we tend to assume the information in search results is accurate
• Google is a habit that we not only aren’t likely to break anytime soon, but instead are likely to grow more dependent on
• Our reliance on search-based navigation has leveled the playing field and remains a great way for anyone to reach an audience

On Google and the use of search engines

The survey asked about Nicholas Carr’s assertion that Google makes us stupid. 76% of those asked disagree and think that the internet is enhancing our intelligence. The more information we have access to, the smarter we are. Of course, this view begs the question: doesn’t the validity of this view depend on the quality of the information we have access to? It seems to me in that sense those surveyed reflect our collective assumption that if it’s ranks well on Google, it must be true.

Some studies have shown that searchers put their trust in what results rank highly, although that may be on the decline. The way the experts answered this question, however, implies that overall, we as searchers equate high Google rankings with accuracy. David Ellis, of York University in Toronto sees this tendency in his classroom: “unless pushed in the right direction, students will opt for the top 10 or 15 hits as their research strategy.” Interestingly, other respondents bring up the issue of accurate results not from the perspective that Google and other search engines don’t necessarily rank based on truthfulness, but that searchers might not be searching correctly.

Peter Griffiths, former Head of Information at the Home Office within the Office of the Chief Information Officer, United Kingdom, notes “the potential for stupidity comes where we rely on Google (or Yahoo, or Bing, or any engine) to provide relevant information in response to poorly constructed queries, frequently one-word queries, and then base decisions or conclusions on those returned items.” This sentiment is echoed by Robert Lunn, consultant at FocalPoint Analytics: “there is a big difference with what a world class artist can do with a paint brush as opposed to a monkey. In other words, the value of Google will depend on what the user brings to the game. The value of data is highly dependent on the quality of the question being asked.”

I find it amazing that so much of the  questioning of the value of the information that search engines returned is based on how well we search and not in the search engine’s ability to return truthful results (as opposed to popular or well-optimized ones). Some of those answering did point this out. For instance, Gene Spafford, Purdue University CERIAS, Association for Computing Machinery U.S. Public Policy Council, pointed out:

“Access to more information isn’t enough — the information needs to be correct, timely, and presented in a manner that enables the reader to learn from it. The current network is full of inaccurate, misleading, and biased information that often crowds out the valid information. People have not learned that ‘popular’ or ‘available’ information is not necessarily valid.”

Combine the sense that generally, searchers seem to assume high ranking results are truthful ones with the data showing that searchers trust organic results more than paid ones and a clear case emerges for an investment in organic search for overall brand credibility.

Another telling aspect of both this question and the way in which the experts answered is that in most cases, respondents used “Google” in place of “search engines” or in many cases “the internet”. Google is a habit that most of us don’t plan on breaking any time soon. In fact, the experts surveyed felt we would continue to become more dependent on Google.

On reading and search-based navigation

The report spends a lot of time on the decline of reading and “literary intelligence”. Are we really reading less? And if we are specifically reading fewer books, is that because of the internet or because of video games and TV? According to Patrick Tucker, Senior Editor at The Futurist magazine, “this type of content generation [that is, blogging, commenting, and the like], this method of ‘writing,’ is not only sub-literate, it may actually undermine the literary impulse…. Hours spent texting and emailing, according to this view, do not translate into improved writing or reading skills.”

Another view, however, is that search-based navigation to content (vs. traditional navigation that begins with a reader picking up a printed newspaper or book) provides a level playing field for skilled writers who may not otherwise have found an audience. Fred Stutzman, School of Information and Library Science, University of North Carolina-Chapel Hill, for instance, noted, “I firmly believe that more people than ever before will be afforded the opportunity to write and create, to find audiences, and engage in content-enhancing feedback loops that will enhance communication.”

On the “next” Google

Susan Crawford, former member of President Obama’s National Economic Council, now on the law faculty at the University of Michigan is “optimistic that Google will get smarter by 2020 or will be replaced by a utility that is far better than Google.” Others surveyed described new ways of searching (perhaps without realizing that’s what they were doing). Rich Osborne, Web Innovation Officer, University of Exeter, for instance, predicted that “it will become commonplace to be able to overlay reviews of a product simply by pointing a screen at it, or check the weather forecast by pointing your phone at the sky.” (And you can do some of that now with products like Google Goggles and Shaazam.)

The report discusses many other issues around the future of the internet, particularly anonymity and privacy. But underlying all of the discussions was the premise that Google is our primary gateway to the content on the internet. Obvious, maybe. But it mostly went unsaid. And that no one needed to say it tells me that the future of the internet, at least for a little while longer, centers on understanding how to be visible in google search results.

New touch screen devices, with iPhone and Android leading the way, are taking the mobile Internet into the mainstream. We now have the Apple app store with numerous touch-friendly apps in addition to the hundreds of thousands of touch-friendly websites that people can access.

Yet, there has been disappointingly little innovation in mobile search. The big mobile search engines are still serving up old-fashioned lists of desktop (i.e. full web) search results on touch devices. They are assuming that the full desktop web page is the best result for every search. Search engines that can crawl and index the touch-friendly web will improve upon today’s mobile search services in four crucial ways.

Speed. The typical size of touch-friendly web pages is 30 Kbytes. Typical page sizes on full websites range are around 500 KB—huge! While these full pages may download fine over the fast broadband networks with PCs, they download much more slowly on touch phones on either 3G or crowded public WiFi networks. A search engine that takes you immediately to mobile-friendly results is going to work a lot faster under real-world conditions than a search engine on a phone that takes you to 500 KB page results. It is important to note that speed is one characteristic that is relevant to both mobile and touch-friendly sites, and is crucial for app and web developers to consider during the design process. While sites initially created for mobile use slimmed down greatly, today’s touch-friendly versions are even slimmer.

User experience. Widgets are also easy to click on but not seeing much use in smartphones. Touch phone and PC screens are completely different sizes and shapes. Full websites can be accessed on most touch phones, but they require frequent scrolling, pinching and zooming that combine to make it harder for the user to find the content they are seeking. A well designed site for touch phone users e.g. m.espn.com is “finger-friendly” and helps users get what they need in just a few taps. A highly-relevant touch-friendly web page or app store page is often the better search result for the user than the full desktop page.

Mobile widgets can also maximize the experience on touch phones by providing a scaled-down user interface but now they typically perform a single function and are most useful in relation to RSS feeders, which are becoming outdated due to social aggregation sites such as Twitter, Facebook and StumbleUpon. With the increased availability of rich apps, widgets have lost traction in the past year.

Optimized modalities. Search behavior is different on touch phones than on PCs. Users are less willing to iterate their searches by making multiple modifications of their query for two reasons: it’s harder to type on a touch phone and the results usually load slower across the wireless network than they do on a desktop search engine. These two differences between desktop search and mobile search affect the way that people use the devices, therefore changing the type of content they search for on the go.

Commerce & monetization. Today, mobile users spend much less online than desktop users. This is because the user experience for buying bigger-ticket items doesn’t work well. It’s no fun trying to buy stuff from a full web site on a touch phone—too much pinching, scrolling and zooming—easier to just defer it until you get back to the PC. The touch-friendly web unlocks the door to much higher levels of m-commerce by making it much more convenient to transact on the phone.

New payment systems such as Square will draw increased levels of mobile spending, the since the touch web is still quite young, we will see more new mobile payment systems develop over time, bringing increased m-commerce and one-touch payments to touch devices. The ability to store payment information in one easily accessible and organized place (which we previously saw with PayPal, and currently see with the Apple app store) will make touch devices an even more attractive payment method. As more money gets spent on mobile devices, the value of mobile search ads will continue to increase as companies try and capture shoppers.

The New York Times had a good article about the program this weekend, These Battle Lines Are Drawn In Yellow. We also covered the program’s launch earlier this month in our own article, Google New Local Ad Category Invades The “7 Pack”.

So what’s the problem? In The Trouble With Google’s Yellow Pages Experiment from Silicon Valley Watcher, Tom Forenski, notes:

Prior to this, the results were separated. Sponsored links were always on the right, and the main search results were ‘organic,’ they were listed on how relevant they were to the query — not because of payment.

Actually, Google’s been mixing the line between paid and unpaid results for several months. I’ll get into the past history, but let’s see what’s happening with this particular program. Here’s a search for skateboards in houston:

Look at the top result, for Kingpinz.com. You’ll see that there’s a little “Sponsored” disclaimer that appears just below the listing. Here’s a close-up:

Now is this a change from Google’s past practices of keeping the unpaid and paid results distinct? To some degree, not at all.

Tom says that Google has always put unpaid “organic” results on the left of the results page and paid ads on the right. That’s not true. Google, since it ran its very first paid search ads back in 1999, has had paid ads that have appeared “inline” or directly above unpaid results. You can see this in plenty of searches today, such as this one for income taxes:

In the example above, you can see ads for Turbo Tax and H&R Block appearing directly above the unpaid listing for the IRS. There are also paid ads to the right.

Does Enhancing A Unpaid Listing Make It Paid?

Of course, with the new local program, the entire listing is not highlighted in yellow, to better delineate it from the unpaid listings. I haven’t talked to Google about this yet — I or someone from Search Engine Land will follow up more on this — but I’m guessing the entire listing isn’t highlighted because the listing itself is free.

If you review the program’s information, you’ll see that business listings are free. The “enhancement” part only appears in conjunction with a free listing. If you don’t already show up naturally for free, then your further enhancement won’t be shown. And that enhancement doesn’t impact ranking, as we’ve reported before:

Google assured me, again, that there’s no impact on rankings from this. But it’s the first time to my knowledge that “ads” appear in the “7-Pack.” The program is also totally independent of Local Listing Ads, which are supposed to return at some point this year.

Indeed, here’s an example of an enhanced listing that doesn’t get top billing, for golf clubs san jose:

See listing G, for “The Golf Club at Boulder Ridge?” That’s an enhanced listing, but it doesn’t appear at the very top.

So on the one hand, Google’s long-standing goal to keep the clarity between paid and unpaid results seems in the clear. But then again, Tom clearly got confused — and I suspect others might, as well. “See, that listing is showing up for free, but that thing underneath, that’s an ad. Kinda.” That’s hard to explain.

That brings me back to some past history. In November, our Google Experiments With Paid Inclusion & Does “Promoted” Meet FTC Guidelines? article looked at how a new shopping program seemed to violate the FTC’s guidelines about paid inclusion, plus whether paid ads on YouTube used language that clearly identified those as ads, as the FTC requires.

Google’s response was that it felt things were in compliance. With these local ads, we’re checking and will update with any Google statement we receive. As a reminder, this is the FTC’s guidance about such paid ads, issued in 2002:

The staff is encouraging search engine companies to make changes to their paid-ranking search results to clearly delineate them as such, whether they are segregated from, or inserted into, non-paid listings. Factors to be considered in making such a disclosure clear and conspicuous are prominence, placement, presentation (i.e., it uses terms and a format that are easy for consumers to understand, and that do not contradict other statements made), and proximity to a claim that it explains or qualifies.

Personally, I don’t see the enhancements as ads. But I do see them as potentially confusing for consumers, and there’s nothing that explains more to a curious consumer about why that little yellow highlight below a local ad appears. As the very least, the word “Sponsored” could lead to more information.

Postscript: A Google spokesperson sent us a statement:

Enhanced listings are a new feature in the Local Business Center that enable business owners to let potential customers know what they think is most important or unique about their business. The feature enables LBC users to choose one of the following enhancements: photos, videos, website, coupons, directions, menu or reservations. We are currently running a limited trial of the feature in San Jose and Houston.

To be clear, enhanced listings have no impact on the ranking of LBC listings.

Enhanced listings are labeled as advertisements via Google’s “Sponsored” designation, and this experiment is intended to help us understand whether this is a useful experience for our users. This feature is currently in a limited trial, and as with all tests, we may make changes to our current experiment in the future.

• Microsites – Microsites aren’t inherently a bad idea, but too many of them can cause brand confusion, external link dilution, and require that all search-related relevance and authority build from scratch with each new microsite.
• Display issues – Many of the advertising brands last year ranked well, but due to technical issues had poor titles and descriptions in the search results.
• Lack of consistency - In some cases, the brand bought AdWords for commercial taglines, but then didn’t display that tagline in the ad. Searchers likely skipped right past those ads as they were looking for a match to the phrase they typed in.

Have things gotten any better in 2010? Generally, yes they have. The use of microsites was greatly reduced this year, with most brands opting to simply display their primary domain name. This made ranking much more straightforward. However, overall there was less integration between the commercials and the web.  GoDaddy was one of the few advertisers that encouraged viewers to visit their web site (which surely viewers did, to see the “too hot for TV” commercial sequels).

Why does this matter? 57 percent of us are sometimes online at the same time we watch TV, and that often translates into searches. A comScore study found that 32 percent of those surveyed planned to be online during the game — 14 percent to watch commercials and 13 percent to visit advertiser web sites. Not surprisingly, Google noted search spikes for several advertisers during and after the game.

So, could the estimated 13 million viewers (13 percent of 100 million) looking for advertiser web sites find them? Let’s take a look.

Group 1 — brands that didn’t advertise a domain

In nearly all cases, these commercials were for existing brands and taglines. The new ones didn’t fare quite as well. Bud Light’s new “here we go” tagline for instance, has no visibility.

Acura ZDX: AJAX workarounds gone wrong

Searching for the Acura ZDX returns the page for the TSX. Why, you might ask? Well, even though the ZDX does in fact have its own URL (http://www.acura.com/ModelLanding.aspx?model=ZDX), the TSX page (which likely has been around longer and has accrued more links) appears to be hiding a whole bunch of text via CSS.

So while the page on the left is what visitors see, the page on the right is what search engines see:

Normally, I recommend against hiding text like this because it’s against the Google webmaster guidelines and can get a site penalized, but in this case, I’d recommend against it because they’re not really even doing it right. They repeat the same text about all cars on every car page. It looks like they may have originally done it to get around AJAX issues (the ranking page 302 redirects to a a URL that includes a #), but this is a prime example of why showing things differently to search engines and visitors can cause problems with diagnosing what’s gone wrong with search visibility.

Group 2 – brands that advertised a domain

As I noted last year, advertisers likely expect that if they display a URL in their ad, viewers will type that URL in their browser address bar. But often, viewers search for the URL (or portions of it or the brand name) instead, so it’s important to rank well for anything you’re advertising to maximize the ad’s effectiveness.

Bridgestone Tire – An overabundance of domains

Bridgestone ranks number one for both [bridgestone] and [bridgestone tire]. Unfortunately, the domain that ranks isn’t the one advertised. The advertised domain does a great job of engaging Super Bowl viewers. But how many of those viewers found it?

You can see that Bridgestone did buy AdWords for the correct URL, but that domain appears nowhere in the organic top ten, as it’s crowded out by other Bridgestone properties. This situation can be difficult to fix, as the umbrella brand (Bridgestone) includes many sub-brands (including Bridgestone Tire), and each is likely managed separately. But even excluding SEO concerns, the current structure is likely causing consumer confusion. And it’s undoubtedly hindered the effectiveness of the Super Bowl ads. The domain that ranks first (bridgestone.com) could at the very least include a large call to action that leads visitors to the correct site. In the graphic below, you can see the site that ranks on the left and the site Bridgestone would like visitors to go to on the right (which includes Twitter and Facebook engagement, along with videos of the ads). As a sidenote, the twirly graphics that require you to chase them around with your mouse to go anywhere on the page on the left isn’t necessarily the more enjoyable experience for a bunch of people who may be a bit tipsy from Super Bowl refreshments.

What about the Honda Crosstour?

Honda did pretty well here. They rank in both paid and organic and their display of both is great (the ad even includes sitelinks). This query is one of the only I saw where a competitor tried to get in on the action (see the Chevy ad below).

You’ll notice that the advertised URL  isn’t the one that ranks. In this case, the advertised URL redirects correctly with a 301 and both URLs contain the same keywords (Honda and Crosstour), so searchers aren’t likely to get confused. You may remember that I took issue with Hyundai’s use of a redirected vanity URL last year, but that was in part because the advertised URL was edityourown.com and the “real” URL was hyundai.com. Anyone looking for “edit your own” couldn’t find it. The only improvement I would recommend to Honda would be to ditch the extra parameters. crosstour.honda.com currently redirects to http://automobiles.honda.com/accord-crosstour/?from=http://crosstour.honda.com/. Is the from parameter really needed in this case? Diving further into the site, it appears that primary navigation uses a folder URL structure and secondary navigation uses a parameter-based URL structure. This makes it difficult for the bots to know when parameters are required vs. optional. Which leads to URLs such as this being indexed:

http://automobiles.honda.com/accord-crosstour/exterior-photos.aspx?SP_RID=MTgzNDExNDQyOTES1&SP_MID=2765376&PROGRAMID=RPABRAVO&CAMPAIGNCODE= AT2036&OFFERCODE=AT2036&CELLCODE=BX&PID=975600811&FROM=EMAIL.

Yikes.

At the very least, Honda should get to know and love the meta canonical attribute.

Boost Mobile – triggering related searches

Google Trends only shows 20 spiking queries, so it doesn’t provide full visibility into surging spike interest. But here’s how you can know an ad is driving search interest: when you do a brand search, and the related searches are about your commercial, you know you’ve caused viewers to search for it.

Sadly, Boost Mobile doesn’t actually rank for these related searches. It appears that their microsite, http://www.unwronged.com/, does rank for [boost mobile shuffle], only the domain is indexed by its IP address.

Oh Boost Mobile, just get rid of the weird microsite domain name altogether and host your crazy shuffle flashback somewhere on your main site.

Group 3 – the interesting ones

Hyundai – either they’ve learned or they’ve given up

I really gave Hyundai a hard time last year. They had such a great idea of integrating offline and online media. They created an interactive, social experience online and drove an amazing amount of search interest via their Super Bowl ad. The trouble was that they drove interest in the query [edit your own], but that site was nowhere to be found in the search results since edityourown.com redirected elsewhere.

This year? No microsite. No offline/online integration. No social, viral, online experience. The ads simply showed their domain name: hyundai.com. I’m sad this might mean they think their idea of engaging an audience with offline/online integration was a failure, when really it just could have used a better execution.

They do rank #1 for their brand, although they’re still advertising a domain that redirects elsewhere. But at least both domains have “Hyundai” in them.

Focus On The Family – the search-related risk of famous people

Using famous people in ads can be great for several reasons, but can be problematic from a search perspective. If the famous person already has a solid set of results, it may be difficult for the brand to rank for that person’s name. And viewers are just as likely to search for the person as for the brand. You can see this scenario with the Focus On The Family commercial, which featured football player Tim Tebow.

Clearly the commercial caused people to search for the Tebow family in much greater numbers than for Focus On The Family. Could Focus On the Family done anything to be found for those [tebow] searches? Well, sure. After all, the Huffington Post article about the commercial ranks number two, so it managed to break through all of the legacy content. But focusonthefamily.com site doesn’t seem to have a full article of content that shows it’s relevant for the query. It does feature a video and image, but very little that search engines can actually do anything with.

This is a cautionary tale of online reputation management as well. Lots of negative articles about the commercial are ranking for these queries and the brand’s positive spin is nowhere to be found.

Doritos – ranking despite themselves

Doritos. Their commercials caused search spikes for their brand, and they do indeed rank. I would call this success despite adversity, since snackstronproductions.com and doritos.com are duplicates, and they seem to love creating microsites for every campaign (crashthesuperbowl.com ranks second for their brand name). According to Google’s cache, the word “Doritos” isn’t even on the page and the text is:

“snack strong productions logo (8K) – no flash page Snack Strong Productions requires Macromedia Flash, version 8 or greater. Please click here to download. “

So go them for managing to show up anyway (likely through links).

Go Daddy – integrating media

You can’t fault Go Daddy. Their commercials get your attention and they drive people right to their web site. Last year, I questioned the quality of that traffic. Do people looking for video that’s “too hot for TV” really then buy domain names? As I found last year, apparently the answer is yes. This year, Go Daddy CEO Bob Parsons said, “We had a tremendous surge in Web traffic, sustained the spike, converted new customers and shot overall sales off the chart.”

Dockers – everyone wants free pants

Looking at Google Trends during and after the game, the dominating search was related to those free Dockers pants. Everyone loves pants. Especially if they’re free.

And here’s the fascinating thing: two of the top searches were to the dockers.com/freepants URL, which was was never mentioned in the ad (and doesn’t exist). Yet enough people typed that in to a search box that it had “volcanic” hotness. Fortunately, Dockers either saw this coming or watched it happening and set up a redirect from that URL to its home page. Although they would have done even better by redirecting to the free pants page: http://us.dockers.com/shop/index.jsp?categoryId=4003744. Or on second thought, maybe they could redirect THAT crazy URL to dockers.com/freepants. Sadly, as with Boost Mobile’s spiking searches, Dockers is nowhere to be found for what the commercial caused people to search for. The Dockers site doesn’t appear for any of the four searches listed in Google Trends, although lots of other sites are taking advantage of the surge.

Why? Probably a combination of reasons that begin with its poor URL structure that includes lots of duplication and makes the content difficult to access. Also, the URL for sharing the video? http://us.dockers.com/shop/index.jsp?categoryId=4003744&camp=h-giveaway. If I might make a suggestion? What about instead using dockers.com/freepants?

They even run ads on other search engines. Search for “Google” on Yahoo and you might see:

(The 3-line description is a result of the way some ads display in Yahoo’s results.)

Various blog posts have offered advice on writing “killer” ad copy, suggesting (among other things) to use the keyword in title of the ad and to also use the keyword in the ad’s description text. Others advise including a strong call-to-action, though the jury still seems to be out on the subjects of best use of capital letters and on whether or not to use an exclamation point. Google itself suggests to test multiple ads in each group.

Checklists are for the grocery store

The reason why search engine marketing can often be frustrating perhaps is already obvious: even for Google, “industry standard” practices can’t seem to be universally applied. The two ads above both use “Google” in the title and description, but one includes an exclamation mark while the other does not. One treats “Homepage” as a single capitalized word while the other treats it as two words (but strangely doesn’t capitalize the word “page”). And neither seems to be part of an ad group that has multiple ads, since these are the only creatives I have ever seen for this search query. If you use a checklist of “best practices” when writing ad creatives, at least be aware that Google’s ad writers don’t seem to do the same.

It might be that Google simply isn’t trying to optimize the performance of their ads on their own website (since they don’t change their own bottom line by charging themselves for clicks). But it seems harder for me to believe that Google isn’t trying to optimize the ads they place on Yahoo and Bing. Even if money isn’t an issue, certainly setting a good example and maintaining the appearance of competence are.

So, I decided to examine Google ads on Yahoo and Bing further to see what lessons we can learn from Google about writing good ad creatives:

Lesson #1: Put all, or most, or some of the keywords in your ad

Yahoo and Bing don’t seem to run Google ads for terms specific to Gmail (Google’s email service), Google Maps, or Chrome (their web browser), but they both run ads for Google’s new Nexus One phone. For the query “nexus one phone,” Bing and Yahoo both run the same 3 ads:

Notice that none of the ads mentions the full keyword in the title, even though there is enough space in the titles of the top two ads to say “Nexus One Phone by Google”, instead of just “Nexus One by Google.” And, none mention the keyword at all in the text description.

Lesson #2: Emphasize benefits, not features (or the other way around)

A “golden rule” of writing good PPC ads is to emphasize benefits, not features. So, for example, you should say that the machine offers “fast performance” rather than a “1 GHz processor.” But looking at the three ads above, it appears that all emphasize features (”4 GB Flash Memory,” “1 GHz Processor,” “Tech Specs”), not benefits. Perhaps this “rule” of killer PPC ads should be revamped to say: “Emphasize benefits, not features… except when you should emphasize features, not benefits.”

Lesson #3: About attention-grabbing headline

Another common piece of advice is to write an attention-grabbing headline. Bing doesn’t seem to run Google ads on terms related to finance, but for the queries “stock quotes” and “financial information,” Yahoo shows:

“About Stock Market?” “Financial information?” I don’t know if I could think of a headline less enticing than those. Maybe, “About information?” Using bland headlines like these might make sense for testing purposes if Google had similar ads in the rotation which had much more enticing headlines. However, in apparent violation of their own advice to test multiple ads in each group, these seem to be the only ads they show for “stock quotes,” “stock quote,” “stock market quotes,” “financial information” and similar queries. Notice also that few words are capitalized in these ads, not even the display URL, even though some suggest to capitalize the display URL. And as far as the advice of using words that evoke emotions goes, these ads are stone-faced. So, perhaps Google has already done their testing and discovered that these lame-looking ads get the best response—in which case, maybe writing enticing titles only attracts poor-quality visitors.

Lesson #4: Keep the ad and the landing page closely related to the keyword (unless you have a new webphone on the market)

Interestingly, the ad which appears on Google itself for “stock quote,” “stock quotes” and “stock market quotes” is:

It seems that rather than take you to their highly relevant Google Finance page, as their Yahoo ads do, they’d rather use your “stock quote” or “financial information” query to show a less-relevant ad which cross-markets their shiny new cell phone. (I wonder what the landing page quality of their phone page is compared to their finance page for this term. For that matter, I wonder what Quality Score a low-relevance ad like this one has…)

Lesson #5: Always-be-testing everything but the display URL

Stock quotes and webphones are one thing, but Google makes most of its revenue from selling pay-per-click ad space, so let’s get to the heart of the matter. On the query “pay per click advertising,” they run at least six different ad creatives on Yahoo and seven on Bing, five of which are common to both search engines. Of the five ad creatives in common two use “pay per click advertising” only in the title, one uses these words only in the description text, one uses these words in both the title and the description, and one uses them in neither the title nor the description text. Here’s how the ads appear on Bing:

These ads on Yahoo are the same, but with periods in place of the exclamation marks. So, it appears that Google might be doing full factorial testing of the titles and text on some terms that are core to its business.

Though some suggest to test multiple versions of your display URL, it seems that Google has disregarded this advice, since the display URL on all the variants of the “pay per click advertising” ads on both Yahoo and Bing are identical: www.Google.com/AdWords.

Lesson #6: Don’t sweat the small details

Surprisingly, in one of the ad variants that appears on Bing but not on Yahoo, Google’s ad has a minor formatting error that has unwittingly tripped up many search marketers:

By not realizing that when ads appear in the “promoted positions” above the natural results, the description text is presented on one line, Google’s ad writers left out a blank space between “Minutes” and “Instant.” Try to not let this happen in your ads!

Lesson #7: Sweat the small details

When testing ad variants, it appears that Google rotates ads evenly on Yahoo and Bing, but uses the “Optimize” feature (which delivers ads with higher clickthrough rates more often than those with lower CTRs) for ads which are shown on Google itself. For example, for the query “search marketing,” Google has at least seven variants on Bing (some of which use the keyword in the title, some which do not, etc. etc.), but on Google itself, they show only two ads, the first of which appears about 90% of the time and the second only about 10% of the time:

Personally, I can’t see anything that would explain the apparently large difference in CTR between these two ads which would cause the first to be shown so much more often, except perhaps that “get” comes from German (”gessen,” by way of Old Norse “geta”), while “gain” is from French, and English words of German origin are often perceived as more “visceral” than their French or Latin equivalents. So, perhaps that mere fact explains their apparently wildly differing CTR’s and very small details can cause enormous differences in behavior.

Lesson #8: Use Google’s new boldfacing algorithm to your competitor’s advantage

As a final note, Google has recently begun boldfacing text for words which are synonyms for, but not identical to, words in the query. This might expand the range of words creative editors would consider using in the title and text of their ads. But, the differences between which words Google boldfaces and which is does not can be interesting. For example, for the query “free web mail” Google seems to run at least three different ads, two of which mention Gmail (one is shown below at left). For the query “phone apps,” they seem to have at least six different ads, one of which mentions the iPhone and is shown below at right:

It’s revealing that Google’s boldfacing algorithm considers Apple’s brand term “iPhone” (a competitor to Google’s phone) to be a synonym for “phone,” but Google does not consider its own term “Gmail” to be a synonym for “mail”.

The general rule for people who write PPC ads seems to be: there are no general rules. Forget about “best practice” checklists because they don’t even seem to work for Google. Instead, just test as many ad variants as possible, even ones that you wouldn’t think might work well.