Author Archives: ghotchkiss

Yahoo’s Stock Slump is not Indicative of the Industry

Yahoo is currently taking some lumps for a slow down in a couple of sectors, but don’t try to extrapolate this to the industry at large. In a Business Week column Catherine Holahan indicates why the slump is due more to internal issues at Yahoo rather than some industry wide malaise.

What is pointed out in the column that is probably more interesting is the growing fragility of Yahoo’s largely banner dominated network, which has been a steady revenue anchor for the company for some time. Increasingly, online is improving the ability to put the right message in front of the right person in the right place at the right time with increased targeting options, either through behavioral targeting, online property targeting or demographic targeting.  Banner distribution on a tired stable of portals isn’t fitting the bill in any of these regards. The behavioral targeting dollars are currently going away from the top 3 into smaller players like Tacoda and Revenue Science. Google has the edge in targeting ideal online properties through roll out of display in their AdSense network and MSN is definitely building a robust demo targeting platform in adCenter. Yahoo’s push back of Panama, which would at least even the playing field for awhile is definitely an Achille’s heel and the race is picking up here.

In the rush for ad spending accountability, Yahoo is starting to lag behind and that could prove fatal if they don’t catch up fast.

What Happens when the Whole World Becomes Searchable?

First published September 21, 2006 in Mediapost’s Search Insider

There are a few items that crossed the threshold of my inbox recently that led me to speculate about search in the grand scheme of things.

First of all, fellow Search Insider David Berkowitz talked about online data storage, and how it could introduce reams of new content into online depositories, there to be connected to by consumers through search.

Secondly, Apple and Google are in talks about iTV, Apple’s new set-top box that allows you to view downloaded video on your TV, at the same time making it searchable.

Welcome to e-World

The fact is, the whole world is becoming digitized and indexable. It’s not a new trend, it’s been making inroads for the last two and a half decades, but there seems to be a tipping point of convergence that’s rapidly approaching. National and international news is almost fully digitized, and local news is following in the same footsteps. There are now digital editions of most periodicals. And Google is doing its level best to digitize every book ever written. So the print world is well on the way.

The Genetics of Music

For electronic media, music is largely in the digital domain, and the searchability of it is rapidly improving. The biggest bottleneck is in trying to categorize and rationalize what is largely a subjective experience. I either like music or I don’t. How do you make that searchable? Well, interestingly, Pandora’s Music Genome Project is trying to do just that. Since 2000, it has analyzed hundreds of thousands of songs based on over 400 attributes or “genes” (hence the Genome moniker) which include melody, harmony, rhythm, instrumentation, singing styles, lyrics and arrangements, to name just a few. It’s a large-scale attempt to make music searchable by something other than genre, artist or title, which is far too limiting for most of us. The Pandora interface, in its attempt to be intuitive, doesn’t allow for power searching, but it’s still a quantum leap forward in allowing us to help define our likes and dislikes in the musical universe.

What You See is What You Search

If you take this same approach to video entertainment, there is a much more complex, and therefore richer, content depository to mine. Think of the universe of movies, TV shows and documentaries that exists, each loaded with dialogue, topicality, visuals and styles. As complex as music can be, video explodes the content to be categorized and analyzed in a dozen different directions. It provides a huge indexing challenge, but therein lies the promise and profitability. And it appears to be a challenge that Google is ready to take on. Of course, we haven’t even touched on aspects like consumer-generated video content (the YouTubes of the world, which seems to be the latest overladen bandwagon) and social tagging.

We’ve Only Just Begun…

But that’s the globally visible world, the tip of an immensely large iceberg. There is very little in our physical world now that isn’t digitized somewhere. There is a virtual mirror for almost every physical presence. Store inventories exist in the digital domain, and have for some time. Aggregating those inventories and making them searchable turns the entire world into your personal shopping mall. We leave GPS trails as we move from point A to B. Our vehicles churn out detailed performance summaries via the onboard computer as we do so. Mobile computing makes the very stuff of our personal lives; our thoughts, our activities, our appointments, our contacts, all digital and indexable. At work and at school, we all produce content on a daily basis. My daughters are content producers each time they do homework, and increasingly, that work is in bits and bytes.

As the barriers disappear between our hard drive and the Net (the subject of David’s column) all this content theoretically can enter the public domain and be searchable. Increasingly, the question we ask ourselves is “where do I draw the line between my private and my online world?” File sharing becomes a substantially bigger deal.

Brain Melting Questions

Fellow blogger Mitch Joel calls these kind of questions “brain melters.” I like that. It captures the mind-numbing aspects of this stuff. Our electronic footprint is now bigger, and in some ways more real, than our physical one. There is this vast binary universe out there, terabyte after terabyte of data that grows each and every second, capturing the essence of who we are and what we do. And the sole door to that world, the channel we all must pass through to gain entry, is search. In the act of searching, we connect to that universe.

Cast the search question in that light. Realize that we have yet to scratch the vast potential of this fundamental glue that holds the Internet together and bonds us to it. Imagine owning the solitary access point to everything!

Google, Yahoo and Microsoft are jockeying for position to do just that. It should excite the hell out of their respective shareholders, but it should scare the hell out of us. Do we really want this much power in the hands of so few?

These are big questions, and I’d love to get your viewpoint. Leave your thoughts on the Search Insider blog, or drop me an email at gord@enquiro.com.

What Happens When the Whole World becomes Searchable?

My Search Insider column today was big picture stuff, looking at how search can connect us to a digitized world.

Here’s an excerpt:

There is this vast binary universe out there, terabyte after terabyte of data that grows each and every second, capturing the essence of who we are and what we do. And the sole door to that world, the channel we all must pass through to gain entry, is search. In the act of searching, we connect to that universe.

Catch the rest of the column at MediaPost.

The column drew some interesting responses, both on the Search Insider blog and emailed to me.

Martin Edic truly thought globally

In the spirit of creating a ‘brain melter’, imagine the extension of search created by GPS and satellite imaging. Suppose I want to create a search engine devoted to global climate change. If I can access these sources I could literally do a planetary search that included both digital data a geographoc, geological, weather and other environmental data all viewable as imagery, maps, text, etc.

David Gust took exception to my plaudits for Pandora: I initially thought Pandora was great, but eventually it became monotonous. A descriptor genome for the music is great, but it doesn’t decipher the music consumption genome in me.

My point is that indexing means little without context. Context is about behavior and that is where the true focus must be placed to truly unlock value of “Indexing the World”

Derick Harris,w ho obviously has a lot of time on his hands, took me to task for my “pointless” vision of an Orwellian future

I do wish that these marketing rhetoricians of search, such as Mr. Hotchkiss, would “think first” about what they are asking, in terms of “big questions” — instead of wasting our time with patently pointless essays that amount to self-serving indulgences posing as questions that really amount to a whole world Googleized into an information hell.

…Ouch! Sorry Derick, I obviously hit a sore spot.

And in the spirit of wired “Big Brother”, Warren Peace (come on..that can’t be your real name. But if it is, kudos to your parents!) shared his vision of a database schema for a “global object database” or GOD for short…

whereby every kind of digital data could be stored, indexed and cross-referenced, and rated for accuracy (couldn’t find funding for it, though). One issue is that many things are analog, not digital, and digitizing them means losing important information. An image of a person and a list of their interests is NOT the person, just an avatar. Do we really need an avatar of every living thing?

Perhaps that’s what the real “God” is – an analog, searchable object database that details absolute accuracy.

Talk about your brain melters!

Online Video Needs Critical Mass

More on the topic of online video. It seems the majority of stories I’m seeing in this space recently have to do with moving video to the Net. A recent one was the agreement between YouTube and Warner.

YouTube is as hot as a high grade viral infection right now, which is what it essentially is. It’s the latest Net “Buzz” poster child, and it’s reaping huge amounts of traffic. That’s a great step towards sustainability, but as we’ve seen in the past, the Net’s traffic patterns are notoriously fickle. The tide can turn overnight and head to a new spot. What YouTube has to do is grow up without growing old. Kafka gets it right in his article.

Kudos to Warner for understanding the ebb and flow of the Internet. You have to watch where the new communities of interest are gathering, and shift your strategy to be in the right place at the right time. Presently, YouTube is the right place. The only question is how long is the “right time” window. YouTube in it’s present form is all hype and little substance. We’re still playing with the novelty of online video. We still get a kick out of watching teenagers lip sync to a popular song (or the theme song for Pokemon) in his/her bedroom. That will get old fast. Not to mention the questionable legality of most of the content on YouTube.

Warner is smart enough to realize that the consumer is at the wheel and will control where distribution occurs. They’re laying their bets on YouTube, and it’s probably a smart bet. At least, it’s a smarter bet that where the competition is placing their chips. Universal is still trying to maintain the illusion of control by going head to head with another red hot property, MySpace.com, with threats of legal action due to copyright infringement. EMI and Universal has also gone with SpiralFrog, a start up. They have obviously given up traffic for greater control.

But for YouTube, the trick will be to provide more meat as it transitions from a viral novelty to an internet mainstay. This trick has been successfully pulled off only a few times in the past. One was Google.

Google and the Future of Video

The talks that Google and Apple are currently in about video will likely start defining the future of video entertainment as we know it. And it’s just one more example of “push” going to “pull”.

The news story is about iTV, the new device that bridges the gap between the TV and the PC, letting you viewed video from your hard drive on your TV. It’s the continuation of convergence that I’ve been talking about for some time now.

But what is interesting about this to me is not so much the hardware as the extension of searchability to online entertainment. It’s just a matter of time before the walls come down between something like YouTube and the world of broadcast TV. They’re already crumbling rapidly. And setting your viewing preferences based on searchability opens up a whole new world. I’ve had just a taste of it through Microsoft’s Media Center and I like it. You can search up to two weeks of programming by keyword, looking for a particular topic, director or actor.

Now, let’s extend this the next logical step. Let’s open up the rapidly exploding world of video. All the movies, all the tv shows, all the documentaries ever made, as well as the crushing wave of consumer generated video content, all as searchable as the web thanks to Google. You in the mood for a show about 9/11 conspiracy theories? A quick search and you’re watching Loose Change. Plus, Google suggests other shows you might be interested on based on your topic. It’s just a matter of time before somebody does for video what Pandora is doing for music, allowing you to explore the world of video entertainment based on similarities to what you already like.

Social tagging opens up more possibilities, allowing you to tap into the most popular choices of the various online communities you belong to. The buzz effect takes over (as we see currently on YouTube) and suddenly watching online video becomes a communal experience.

It’s a revolution in video distribution, and the seeds are being sown currently in the chat that Steve Jobs and Eric Schmidt are likely having as we speak.

Google Dropping Sponsored from the Golden Triangle?

Whoa..this is a bold move!

Just saw a thread on Webmaster World that indicates Google is testing removing top sponsored ads after a number of searches where a user doesn’t click on anything. Tried it myself and sure enough, after 4 or 5 refreshes, top ads were gone.

After refreshing on the same query, the ads disappeared for that query, and any modifications of the query, but still showed for a totally different query. After I went through the same process with the new query, all my top ads disappeared.

If Google sticks with this, it demonstrates a huge dedication to the user experience. Our research has shown how valuable this real estate is from a monetization perspective, but Google’s feeling (and rightly so) is that if you’re skipping past it anyway, the probability of a click on these ads is minimal. Why impair the user experience but taking up prime real estate with something that the user is just filtering out anyway.

I did some more testing with some different patterns to see where the sponsored filter seems to be tripped. If you do a number of different searches without clicking on sponsored listings, it doesn’t seem to kick in. It’s only if you do a lot of return visits to the same set of search results without hitting a sponsored link. But once the ads are gone, they’re gone for every query from then on til you clear your cookies.

Ironically, my only hesitation with this is from the user experience perspective. My feeling is the thresholds might be set too low. Intent plays a huge part in how we interact with search listings, and this can vary greatly from search to search. It’s also very difficult to determine from the nature of the query. So, if I’m in a fact finding mode, even if I’m using what appear to be very commercial terms, and I skip over ads on 4 or 5 subsequent returns to a page, that doesn’t necessarily mean I don’t want ads on any search. One anomalous search could filter out top sponsored results for days, weeks and even months and the user would never know what happened. There is no indication on the page that Google is applying any type of filter. There is no way to turn them back on. For 99.9% of web users, they’d never know what happened.

Now, it’s not all ads, but only the top ones that disappear. But the fact is, the difference between visibility and performance of ads in the two locations is so significant, that moving the top ads over to the side is almost like removing the ads from the page. Almost everyone starts scanning at the top of the page.

Google’s intentions are noble here, but they’re actually removing control from the user. I’m a big champion of organic results, so I can’t believe I’m saying this, but Google might be too hasty in stripping out top sponsored ads. In two different eye tracking tests, we found that it was clicks on these top sponsored links that actually offered the highest success rates for users. I’ll be watching with great interest to see how the test progresses.

The SEM Hierarchy of the E-mail Inbox

First published September 14, 2006 in Mediapost’s Search Insider

Into each social structure, a little stratification must fall. As our society takes a decidedly virtual turn, I’m finding that my Outlook inbox is the latest place where a class structure is taking shape.

Of course, you have the standard spam vs. non-spam sorting, but this doesn’t really count. That happens pretty much transparently in the background, and every day or so I wade through the muck in my deleted spam folder just to make sure a vital piece of communication didn’t get waylaid. For instance, today an e-mail from my lawyer went there. On second thought, perhaps the filter knew better than I what should be deleted.

No, it’s the e-mail that survives the cut that is subject to endless classification and sorting, as I haplessly try to wrap my priorities around an ever-expanding inbox. At first, I thought the six different flags supplied by Outlook would do the trick, but I quickly realized my complicated world needs much more than six classifications.

So in an attempt to ease the daily burden of countless search marketers, I offer the following suggestions for an SEM Custom Rules plug that would automatically take the following actions in Outlook’s inbox.

The “Anything from Google” Rule

It doesn’t really matter what comes in with an “@google.com” on the back end, you’d better open it right away. These go on the top of the list. If it’s from Matt Cutts or Tim Armstrong, perhaps a siren and flashing red light to draw further attention. I don’t get e-mails from Eric, Sergey or Larry, and I suspect the same is true for most SEMs, but if I ever did, I would like a heavenly ray of light to shine gently on me as a choir of angels sing the “Hallelujah Chorus.”

The MSN Beta Invitation Rule

This could dramatically reduce the manual sorting required by automatically signing up for beta test groups for MSN’s new adCenter products, including the Targeting by Presence of Facial Hair Platform, the Visitors You Wish You Got Report feature and the Integrated adCenter/Xbox 360 Console, which drops you into a virtual 3-D world where you can walk up to leads that didn’t convert and slap them for being stupid.

The “Hey, I Got a Speaking Gig” Rule

This would (until recently anyway) include e-mails from Danny Sullivan, Chris Sherman and Brett Tabke, indicating which panel you’d be speaking on at the next big show. These e-mails have to be referred to quickly so you have time to book hotels and flights, and then start e-mailing to see who else will be at the show, who was going to what after- hours function, if you could catch a ride with them, who else was on your panel, and when is the deadline for getting the presentation done (no, not the official deadline–the “real” deadline).

The “Why the Hell Did I Sign Up for This?” Rule

The average search marketer signs up for approximately 6,428,943 newsletters, 194,597 Google news alerts, 963,693 forum post notifications–and that doesn’t include RSS subscriptions. This is all done in the hopes of gaining some vital piece of information that would give them the leg-up on the competition, who are of course all subscribing to the same things. This rule would scan everything for the 1 in 159,975 chance that there’s a useful tidbit in there somewhere. The one exception is the Search Insider–naturally.

The SEMPO Board Communication Rule

Admittedly for a very small market, this would be nonetheless essential for those who serve on SEMPO’s board. It would be able to detect the difference between the 9,543 e-mails a day you get just because you were part of the e-mail alias, a 12-page-long cc list, and the messages requesting you to get off your butt and do something.

The “Rocket to the Top of the Search Engines” Rule

Although these e-mails are technically spam, I like to read them every so often and feel smug about how superior and morally pure I am, and how far my company has come since the days when everyone tried this marketing tactic.

The “Arrange a Meeting/Teleconference” Rule

Why don’t we just accept the fact that it takes 3.6 months and 112 e-mails back and forth to arrange any type of call or meeting, so we should just automate the process? That way we can still feel good that we’re trying to facilitate the phantom meeting by generating reams of e-mails and invitations, while saving us some time. In the end, it will automatically revert to the original time and date proposed, as it turned out that it was really the best for everyone, anyway.

The “Loved Your Column” Rule

Okay, seriously, these e-mails, when they come in (and yes, I have got a few), are the highlight of my day, and they’re the first I respond to. Of course, don’t take this as a hint or anything.

Tales of Pogo Sticks, Bouncy SERPs and Sticky Pages

First published September 7, 2006 in Mediapost’s Search Insider

Much of what little strategy exists in search marketing is aimed towards the first click from a results page (also called a SERP). The position, the messaging and the landing page experience all assume that we’ve captured that all-important first click. But what about the subsequent clicks? In the search business, this is called pogo sticking, the bouncing back and forth from the search page, and clicking on a number of sites in sequence in an effort to find what we’re looking for.

Desperately Seeking Pogo Stats

We know pogo sticking exists, but when I tried to quantify how common it was, I quickly ran into a lot of closed doors. I tried all the major engines and was told that they don’t divulge that type of information, even in aggregate form. I also tried the monitoring services (comScore, Nielsen, Hitwise) but again came up empty.

So, failing anything more quantitative, we had to turn to our own limited data set. The stats below come from the combination of eye-tracking sessions, where we’ve been able to look for pogo sticking. I’m not sure how accurate it is, but it’s the best we’ve got, so I present it with a whackload of caveats.

We saw pogo sticking occur in 49 percent of the sessions we looked at. We suspect the occurrence of this type of behavior would be even higher in real-world settings. So at least one out of every two searches results in a return visit to the results page. In our sessions, 21.5 percent of them results in two clicks from the SERP, 10.4 percent in three clicks, 4.9 percent in four clicks, and 5.5 percent in five clicks. The remainder (6.8 percent) clicked six times or more.

Google has the fewest pogo sticking sessions, with only 36.4 percent of them resulting in a round trip to the SERP. MSN had the highest percentage, with 59.4 percent. Even if you question the numbers (and you have every right to do so) I believe it’s a pretty safe bet that pogo sticking is a pretty common occurrence.

The Power of the Pogo

Why is this important? Because a return visit looks significantly different than the first visit. And if it happens at least half the time, it’s a factor we’d do well to consider as we lay down our search strategies.

I strongly recommend that all search strategies take into consideration the mind-set of your target customer, within the context of what else appears on the page. This exercise can help you forecast the receptiveness of your target to your position on the page, the messaging you present, and the landing page experience you provide.

Let’s walk through a typical scenario. Our target customer searches for “hybrid SUV’s.” Because we’ve done our market segmentation homework, we know our target is early in the buying cycle, and is looking for alternatives for fuel-efficient SUVs, building a consideration set.

Eye-tracking studies have shown there’s relatively little variance in the scanning activity with most searchers at the beginning. They tend to start at the top and work their way down, with a strong bias toward the No. 1 organic spot. Therefore, in this scenario, we have to look at how enticing these top listings are. In walking through this on a search engine, GM and Lexus had purchased the top sponsored spots, where the majority of searchers start their scanning. The first organic spot belongs to the site hybridcars.com, a comparison of available hybrid SUVs. Given our target and his intent, it’s very likely that this site will capture the majority of first clicks from the page.

Beyond the First Click

If we’re playing in this real estate, we have to look beyond the first click to what might happen on the second and subsequent clicks. Scan patterns spread around more evenly on the page on return visits, without the very strong upper-left bias that tends to create the so-called “Golden Triangle” (so-called because we called it that). People tend to fixate on where the last listing clicked, and then can head out in multiple directions from there, either continuing down the listings, skipping up to take another look at the top sponsored, or even a quick glance across to the side sponsored ads. Whereever they choose, their interactions will now be colored by what happened in that first click.

Our strategy now has to account for the influence of that likely first click. We have to know how it will alter or reinforce the intent of our user. We also need to know how sticky the landing page behind that first click is. Is it the type of page that will hold him, and possibly send him off in another direction, or is it a quick bounce back to the SERP because it isn’t well aligned to our target’s intent? Does it reinforce our brand, or our competitor’s? What appears above the fold, and what appears below the fold? Again, we know from eye- tracking studies that this is the critical divide of the page in terms of scanning activity.

When one realizes the impact of pogo sticking, it suddenly means that our search strategy doesn’t play out in a vacuum. It’s intimately dependent on what else appears on the results page, and the most likely paths our target will take from that page. It increases the complexity of our strategy exponentially. The only way to successfully navigate it is to have a clear view of the intent of our target. Sure, it makes search marketing more difficult, but it also makes it infinitely more interesting!

Dan Brown Rant

I’ve been surpressing this for a couple weeks now, but I’m threatening to blow a gasket if I don’t do some venting.

Enough with the frigging Da Vinci Code already!

While over in Europe, I fulfilled some lifelong dreams by visiting the Louvre in Paris, as well as Florence and Rome. But everywhere I turned, I kept bumping into adverts for the Da Vinci Code. The Louvre even has it’s own Da Vinci Code tour.

Okay, nothing against Dan Brown. I’ve read the book, as well as Angels and Demons and Deception Point. I enjoyed them. They’re entertaining, but they’re blips on the cultural radar. Mr. Brown is a writer on par with a John Grisham or a Michael Crichton. The plots are mildly interesting, the characters are about as thin as the paper it’s printed on and the dialogue is as stiff as the hard cover. The whole ancient religious sect angle raises it to slightly better than average, but just slightly. There is no justification for this tidal wave of attention that the book seems to be garnering. I thought it was restricted to this side of the Atlantic, but I was sadly disillusioned to see that it’s taken route in Europe as well. Give me a break, Paris, Rome and Florence, you’re better than this!

If you really want to see how this whole angle can be handled masterfully, take a big juicy bite out of some of Umberto Eco’s works. Compared to Foucault’s Pendulum or Name of the Rose, the Da Vinci Code is like a postcard of the original Mona Lisa.

Schmidt Takes a Bite of Apple

Eric Schmidt now sits on the board of Apple, which is leading to a flurry of speculation about what potential partnerships between Google and Apple may be in the offing. So far, the speculation seems to be rather mundane meanderings about integrating Google search technology in iTunes, possible assistance for Google in assembling a suite of apps to compete with Microsoft, and possible entertainment content distribution deals, with Jobs ties to Disney.

To me, the potential lies in the possible creation of power positions on Microsoft’s outposts, rather than assembling the forces for a head to head onslaught on the heavily fortified desk top app market. Even with Apple at their side, Google faces a daunting task in taking on Office with their tremendously entrenched position. At best, I would see them capturing a good percentage of the relatively small “alternative” market, but unless something shifts in market positions, I don’t see them swinging many main stream customers away. These apps are squarely in the pragmatist market, and the customers adverse to the risks inherent in a switch, especially when there will be a seamless upgrade path offered by Microsoft to a live version of Office.

But the cozying up takes on a more interesting dimension when you explore the possibilities beyond Microsoft’s power positions. In the mobile computing world, Microsoft has been struggling to gain market share, and the step from an iPod to a full mobile pc is not that far. Tie that together with Google’s work in creating compelling mobile apps and a web based application matrix and some interesting possibilities present themselves. We’re not far from the horizon where mobile computing starts to replace the desktop. And mobile distribution of entertainment content is rapidly moving through the early adopter stage. A chasm crossing of significant magnitude isn’t far away. I’ve got to believe that Jobs is visionary enough to see this. Apple never managed to beat the Wintel cartel on the desktop, but the hip pocket is a whole new ballgame.

Also of note is what this means for Eric Schmidt himself. Is this move solely for the benefit of Google, or is Eric positioning himself for life after Google? As the Times articles states, this moves him into a pre eminent position as a Silicon Valley power broker, and he has had political aspirations in the past. While Schmidt has done an admirable job in minding the kids at Google, the power triad structure has never been that elegant, leading to questions of its stability.

Whatever the outcome, in the swirl of partnerships that have recently been announced, this emerges as one of the more interesting developments, with some intriguing long term possibilities.