Traffick: March 28, 2004: Search Engine Enlightenment

BusinessWeek Likes Yahoo's Chances

Further to earlier posts about the bright futures ahead for web portals, BusinessWeek says that Yahoo's stock may not be overpriced after all because of 1) its search advertising revenues, 2) personalization offers unexplored revenue opportunities and 3) you guessed it, data mining user preferences for still more targeted advertising.

The article also points out that Yahoo has a perceived advantage because big brand advertisers tend to prefer those silly Flash ads over tiny sponsored links. Sounds realistic, but I don't know if it's true. Of course, there are many unsubstantiated claims in the article about Yahoo's superiority to Google because of the latter's lack of a close relationship with its search customers, but I find those points less compelling.

Gmail Sighting!

This Google AdWords support page is jumping the gun a bit, perhaps, by showing advertisers where their sponsored links will appear. Scroll down the page a bit, past the partner logos, and shazam! You'll see a sneak preview of the Gmail webmail interface!

The screenshot is small, but you can clearly see how Google plans to integrate ads into Gmail.

One Month Ago in Webmail History

On March 2, 2004, Traffick did the math, decided that Google Mail would mop the floor with Yahoo Mail.

I have to chuckle every time someone lectures me that portals are important because they're big and powerful and have many cool services, or conversely, points to some flaw in a major portal's business model. What these poor saps have no idea is that I've examined every side of those types of arguments for the past five years, and just to be safe, have taken pretty much every possible position on them. I can't possibly be wrong when I've got sound bites to cover all the bases!

But seriously. Over the past three years, this website has reminded people that in spite of its flaws, Yahoo was strong enough that it was just a matter of time before it enjoyed an economic turnaround. We pointed to the huge, untapped area of fee-based revenues and pleaded with Yahoo to stop clinging to its advertising model and to diversify. Not that they were listening, but smart management eventually got that percentage of fee revenue up to nice healthy levels, and Yahoo began to regain its former swagger. People who wanted to give up on it to soon pay too much attention to the newpapers. (Or they work for the newspapers, which gives them a really good reason to pay too much attention to the newspapers.)

Another slight pet peeve we've had is the constant downplaying of search in the reporting on portal ad revenues until it was finally too huge to deny. (In the past twelve months, 40% of Yahoo's revenues came from those "tiny little classified ads" that barely require anyone to schmooze or booze with ad execs... in spite of what a recent Fortune article would have us believe, Terry Semel's main contribution to the financial health of Yahoo has not been teaching its salespeople how to wine and dine.)

But a fundamental respect for the power of portals (the big three especially) doesn't change the obvious fact that Google is king of search and fast becoming a portal itself. Yet journalists and analysts have fallen all over themselves to warn Google that it wasn't up to the job. This has become a recent obsession on these pages: to dispel the media myths that Google's facing unbeatable competition just because Yahoo has a new search product, and has wondrous fee-based email and finance pages and auto classifieds, or because MSN has Hotmail and is working on a new search product and generally rules the world.

The only reason everyone used Hotmail and Yahoo Mail so loyally in the first place was the product quality, period. Traffick argued that in our series of portal product reviews & related comments in January 2000.

By now, though, these "best" email products have fallen behind users's needs, skimping on storage (or charging up to $59.99 for a decent-sized box) and gravitating towards larger and larger animated ad banners even for paying customers, as I've repeatedly complained here. The time was ripe for Google, or anyone, to move in and fix the problem. Granted, they've fixed nothing yet, and we really don't know if their spam solution is as good as Yahoo's or MSN's. But it will be interesting to watch.

So what was Charlene Li of Forrester Research talking about last month when she argued that Google would have trouble competing with the portals because of its "deep-seated cultural focus on search"?

I retorted, in part:

Wha-? Fortunately for Google, the planet has a deep-seated cultural focus on search. And in spite of their grandeur, clever use of punctuation, and recent profitability, the public has retained a deep-seated suspicion of portals while being quite happy to make use of the portal services that prove most useful. Like e-mail. The portal service that is going to lose big market share to Google next quarter.

One hates to break it to Yahoo, but losing huge webmail market share to GMail is a foregone conclusion, and is going to happen as fast as you can say "gigabyte."

As it has been from the get-go of the portal wars, product matters. Brand and monopolistic power matter to an extent, but users have by now learned how to wriggle out of the clutches of big companies who overcharge for junk. (See: AOL.) So Yahoo! Mail and Hotmail are quite simply going to have to improve. They can't just pull the Flash-animated wool over people's eyes.

Hilariously, the latest media spin -- if only in headlines, but they do shape how people form opinions from their daily newspaper -- is "but whoa, hold on, Google will be serving ads in email!"

Hold on yourself. Did I mention that Yahoo not only charges me $59.99, but also show me gigantic screen-dominating ads, making it cumbersome to navigate through basic daily tasks? Oh, right, I've whined about that about eight times here lately. Need I also now mention that keyword-searching my Yahoo inbox sometimes takes upwards of 45-90 seconds to retrieve results? I love Yahoo Mail, and I'll happily admit that some of the advertising in there, like the Vonage ads, is pretty effective. There's room for a variety of approaches, to be sure. But then again, if GMail is 10X bigger, 5X faster, and twice as good at filtering spam, once I get accustomed to the new GMail power, will I ever set foot inside Yahoo Mail again? I have my doubts. Twenty million or so other users might be thinking along the same lines.

Welcome to portalhood, Google. Live long and prosper.

Thursday, April 01, 2004

Rumor: Koogle to Head Up Friendster

Former Yahoo CEO, already an investor in social networking site Friendster, is rumored to be slated for the top job, replacing Jonathan Abrams.

Makes sense: Koogle takes over at Friendster, Yahoo acquires Friendster, Friendster competes with Orkut, Koogle goes back to work at Yahoo, competes with Google. Almost too much sense!

Google to Take "Actuarial" Approach to Pricing AdSense Clicks?

Savvy online advertisers are aware of the pressure that's been on Google of late to follow in Overture's footsteps and allow advertisers to bid separately on content-targeted listings vs. search listings. An AdSense Update email just sent out to publishers refers to "enhancements to our pricing model for advertisers." In particular, the price per click on content-targeted ads will be "automatically" adjusted "based on their expected value to the advertiser." The main technique here seems to be to assess the degree of commercial intent that might be associated with different pages on which ads appear, pricing clicks lower when they appear on (eg.) discussion forums and higher on more commercially-oriented pages, like buyer's guides.

On paper, that makes sense, but I can already hear the complaints that will undoubtedly be coming from advertisers: we still don't have control over the bidding on content targeting, let alone a way of blocking sources of poor-quality traffic, or a way of bidding higher to appear more visibly on the sites which convert the most sales for us.

We'll take what we can get, for the time being: a tacit admission that clicks on contextual ads have often been mispriced, and an albeit arbitrary lowering of the average CPC on content-targeted clicks across the board in order to give some vacillating advertisers more incentive to take the plunge on content targeting.

The unfortunate side-effect, though, might be that for the minority of advertisers who've had anomalously strong performance with AdSense-published ads may see their average CPC's actually rise somewhat, or at least stay stagnant. On the other hand, such advertisers won't ever be charged more than their maximum bids, so they're protected in that regard.

At this stage we expect most advertisers will give this workaround the thumbs-down even if the bottom-line impact is positive on most, mainly because the process is too mysterious. I've been told that "a girl's gotta have a little mystery," but we're not sure that advertisers will feel the same way when the mystery's applied to their ad broker's billing formula.

Here's the really good part of the announcement, though, and this may be where the real intent of Google's new policy lies. Google implies that some publishers may see a decline in revenues, while others will see an increase. Google can mysteriously apply its actuarial tables to assign different values to different "classes" of publishers as a response to the rise of a parasitic class of site owners who are happy to collect AdSense revenues by (somehow) generating clicks on ads while offering little true of true value, in the form of a receptive buying audience, to advertisers.

Google, as they take frequent pains to demonstrate in their public statements, needn't disclose all details of their methodologies here. While the public announcement states examples comparing a "discussion of cameras" as opposed to a "review of digital cameras," there may be other quality notations kept on file. Editorial review of AdSense publishers, coupled with PageRank, might be used as a check of quality and factor into the overall formula to determine the expected value of a click. Sites that are deemed very low-quality, or which seem to resemble other sites which have frequently sent poor quality traffic to advertisers, could potentially see their revenues plummet based on the new formula. Although the formula isn't necessarily based on *actual* conversion data, it might be based on studies and tests that have been done to assess what types of sites typically are leading to high conversion rates to sales in advertisers' campaigns.

Notice I say "might," "seem," and "could." As always, Google leaves plenty of room for speculation.

Complex? Inexact? Sure. But there's never been any perfect way of buying, selling, or pricing online advertising. A colleague's recent experience buying "targeted" ads on a major portal at a "rock bottom" CPM price of 0.60 attested to that. The cost per lead in that venue was 900% higher than the same campaign run on Google AdWords (with content targeting turned on), and 95% of the "good inventory" on that big portal buy came from one source -- big banner ads appearing in web-based email accounts.

Which is a long way of saying that in spite of the many expected grumbles about this latest change to Google's ad pricing, if the price is right, many advertisers will jump all over it. Compared to other forms of advertising, the price is pretty good.

I've provide more details as they become available, likely later today.

Crying Wolf Leaves Onion Limited Room for Manoeuvre

The publishers of satirical news site The Onion, having produced 4,259 satirical or just plain made-up news items in the past year, was feeling a little hamstrung as a wave of contrived hilarity swept Planet Earth.

Between 7:42 and 7:44 a.m. today, the publication ran stories on the 30-year bond, global warming, and early childhood education research. Not eliciting the desired (or any) response, for the following hour it ran a long, streaming clip of Tony Blair holding forth on the perils of WMD's (which elicited a few chuckles from Jon Stewart's research staff, but otherwise little response).

Asked by a reporter "if that was it" for April Fool's Day, an Onion staffer sounded downcast. "We got nothin'," she admitted.

Wednesday, March 31, 2004

A Good Gig if You Can Get It: Google to Launch Web-Based Email

Yahoo! dumps Google. MSN guns for Google. So... Google launches a service that directly competes with those two companies' wildly popular Yahoo! Mail and Hotmail services. The gloves are off.

According to tonight's press release and several breaking news items, Google's long-rumored email service is being tested tonight by 1,000 users. You knew that Google wouldn't enter this space without making plans to do something a little different. True to form, they've kicked it up a notch, reportedly offering a gigabyte's worth of storage. That's ten times more storage than the largest premium Yahoo Mail inbox, which costs $59.99 a year. I'll be one of Google's first customers. With my storage needs and lazy "archive everything" policy -- and I suspect many others are in the same position -- Yahoo's 100MB just isn't enough (seems like just a few months ago I was saying the same thing about the 50MB).

Along with offering tens of millions of web-based email users a potentially better, cheaper, faster product, this will provide Google advertisers with a fairly sizeable additional channel for targeted keyword ads as users flock to sign up for "GMail."

Like the established players, Google also promises that Gmail will help users fight spam. Time will tell if they've come up with any new wrinkles here beyond the usual Bayesian filters, challenge-response, and related anti-spam methods.

ClickTracks 5.0 Makes Quick Work of 'Stat Soup'

Last week I had the opportunity to see a demo of the latest version of ClickTracks Professional, released today.

The highly customizable product, as before, does most of what the "big iron era" metrics software can do, at a fraction of the price.

Among the recent improvements includes a "What's Changed" report. If there are major changes in traffic patterns -- let's say a new keyword or a new source starts sending you a lot of traffic -- this will appear in the report, which is modeled on reports like the "Google Zeitgeist" and Alexa "Movers & Shakers" report. While this sounds straightforward, try to get any of that info in easily digestible form with any other logfile analysis software.

What ClickTracks is doing, perhaps more so than any other company in the industry, is taking information that marketing managers and consultants typically must piece together by hand by poring through the "stat soup" provided by their existing metrics package, and offering ways of creating custom reports with less effort. In short, the product feels like it's been designed with marketers' jobs in mind, rather than for the convenience of programmers. The old days of the IT head taking a huge file of raw data and handing it over to marketing and saying "here you go" or "looks like traffic is up this month," we hope, are long gone.

Another nifty feature that gets the thumbs-up from us is the ClickTracks Way, a new approach to contextual help files. If you happen to be using a certain feature of the software, the idea here is that some insight would be provided as to what steps to take to generate a useful report that could be used for a concrete purpose. (The beauty of Internet metrics is, for all the complexity, it's a narrow enough field that it's worth an analytics firm's time to suggest specific ways of using the product instead of throwing their customers to the dogs.)

Perhaps the best way to get a feel for ClickTracks is to see a demo in person. They'll be exhibiting at Search Engine Strategies in Toronto in May. Maybe a good opportunity to meet up with some fellow stats wonks who "get it"!

According to the website, ClickTracks Professional pricing starts at $2,995. The product is worthy of its accolades. Rather than a bland broth of numbers, ClickTracks gives today's marketers the chunky functionality they need to analyze the performance of their campaigns.

Google's Web Alerts a Threat to Other API Tools?

The Register points out something I hadn't thought about: Google's newly announceed beta Web Alerts, while still rudimentary, are squarely taking aim at the small cottage industry of Google API tools like Google Alert that provide ranking-monitoring features.

Gideon Greenspan, the creator of Google Alert, says he isn't worried about Google's own alert system, and says his forthcoming premium service will offer much more than Google's basic service. Since Google really isn't in the business of providing a robust alert service, I tend to doubt their fledgling service will ever truly compete with third-party tools.

Can't wait to see your premium service Gideon! :)

Cliche Watch

Every time a search engine releases a personalization feature, the PR makes the rounds and we're forced to consume dozens of headlines about Google (or whoever) "getting personal." What does that even mean? Stop!

And while we're at it, a plea to journalists to stop using the empty phrase "pay-to-play" or "pay-for-play" when describing paid inclusion, pay-for-placement, paid listings, or whatever you might see fit to call them. Play? Do I look like Alex Rodriguez to you?

Traffick apologizes in advance for our own past and future contributions to the global cliche-flow. We'll try harder, promise.

Monday, March 29, 2004

New Traffick Article:
Make Your Own Pie: Google's Personalization Merely a Taste of Things to Come
By Andrew Goodman - 3/29/2004

Personalization is one of those hip online topics like weblogs, RSS, and streaming music that no commentator has felt the power to resist over the past couple of years.

Following a period that generated vast expectations for search engines (and an Internet in general) that will do a better job of giving you exactly what you want, personalization-punditry finally fell on hard times. Today, Google has released a modest attempt to show how personalization might work for a search engine user. Read more...

Google's New Look

Judging by the scope of Google's PR for the upgraded look and new features today, it doesn't appear that it's a random test. :)

Since my colleagues and I are most attuned to the impact the new look might have on our advertising clients, "but we're also loyal Google users!," -- our initial reactions reflect those (multiple) biases.

Here were a few of the knee-jerk reactions from here:

A couple of us noted with interest Google beginning to ease out of its relationship with the Open Directory Project, dmoz.org. Since the dmoz founders and most of the key volunteers have also largely distanced themselves from it, what's propping this thing up, we wonder?
The fact that the ads on the right-hand side aren't colored, and look more like listings (in spite of being labeled as sponsored links) is a major change. The new look is cleaner, and the old "fruit loops" look had perhaps run its course. But is it just me or does this new layout appear to jut the ads farther out into the user's field of view? Also, at the same time, aren't the search results awfully far over to the left-hand side of the page, almost "behind" where the eye begins to look? We don't know if this will increase clickthrough rates on the ads just yet, but it seems that the ads are subtly given more prominence.
The two ads appearing in premium position at the top of the page are blended together now, because the backgrounds are the same color. This could result in users clicking the wrong ad. We don't like the looks of this.
Froogle is being promoted more obviously now, but it's still in "beta." They're nothing if not cagey, these Google people! Right, and this site is under construction.
Continuity is the order of the day. That home page is still nice and clean. It would be pretty tough to go around telling people that Google's so dominant because of their clean-looking pages if they turned around and cluttered 'em up.

Now I just hope Google doesn't keep changing the look of its SERP's, since I'm taking screen shots of examples for the section about "how your ads will look on Google" for this book I'm now writing, and I'll now have to replace the first batch of images...

Google Unveils Updated Interface, Users Cheer

I'm not sure if this is a random test that I'm seeing, but it appears that Google has finally updated its interface sitewide this morning. Rumors of this have been circulating for months, and it seems to be official now.

Gone are the tabs above the search box on the Google home page that launched a thousand imitators, in favor of plain text links, and a new "invisible" tab (to borrow from Danny Sullivan) called "more >>" that links to a page of all the Google services and tools (and neat, new icons, to boot!).

Do a web search, and this is what you'll see now:

(Click the screenshot above to perform this search for "wine reviews" and see it up close)

At first glance, it's a bit tough to pick out the changes, but here's what I can discern:

There's a new page header for the type of search, in this case it's "Web". A quick glance indicates this new layout is consistent for other search types, such as Images and Groups.
Search term is more clearly linked to a definition of the term.
If the search term matches text in the URL, it is now highlighted in bold, green text.
AdWords has gotten an extreme makeover. The "interest" bar meter is toast, and the font is larger. There's also a dividing vertical line between the natural, organic search results and the sponsored links. Also, the border around the ads is now but a memory.

My take? I like it. I like it uh lawt. With all the fancy interfaces cropping up at competing search engines, it was high time Google updated their venerable look and feel to keep up, and to facilitate easier searching. I think this change also highlights other search options better, which was something Google seemed to struggle with, as it continued to integrate new features into its toolset.

I'll let Andrew talk about the AdWords changes, but my guess is that it will help increase clickthrough rates a bit, since the ads now look more like search results, in my mind.

Google may take its time when it makes changes, but when it finally unveils such changes, it does a first-class job. Perhaps this will hush the critics in the media who constantly speculate about Google's rein as king of the search engines coming to an end soon. But, probably not.

View Recent Posts

Speaking Engagement

Need Solid Advice?

Posts from 2002 to 2010

Speaking Engagement Need Solid Advice?

Posts from 2002 to 2010

Speaking Engagement

Need Solid Advice?