“Put the science back in computer science: test your Google bias inside,” read a big banner, over the entrance to the Metreon Mall, which is across from Moscone.

Inside, there’s a “Bing In On” kiosk:

One side is just an invitation to take the challenge; you actually do the challenge on one of the other sides:

The kiosk simply takes you to the Bing It On site, where you’re invited to run five searches and see which results you prefer, when shown them from Bing and Google side-by-side. You’re not told, however, which results are Bing’s and which are Google’s.

I found the machine pretty sluggish to use. The on-screen keyboard wasn’t very responsive, and it seemed slow to load pages. I gave up trying to complete a test, myself.

No Pictures!

Since anything search-related catches my eye, my immediate reaction was to start taking pictures. That caused a mall security guard to tell me I couldn’t. That just made me want to take them even more, especially when plenty of Google Glass-wearing people were wandering through, any of whom could easily take a picture and not be detected.

Indeed, Search Engine Land editor-in-chief Matt McGee, standing with me and wearing his pair of Google Glass, used them for that purpose.

Later, we both headed over to the Metreon for a midnight screening of Star Trek: Into Darkness (great, great movie, by the way). Coming at 2:30am PT today, we were intrigued to see a guy working on the Bing It On machine:

It was opened up physically, and the challenge screen itself was down, revealing what seems to be Windows 7 underneath.

I asked the guy working on it what he was doing, and he basically just said “stuff.” The weird bit was when I went to take a shot from further out:

By this point, the guy in the red shirt had turned up. When he saw me taking a picture, he started saying I couldn’t. I actually have a really nice in focus shot of him looking at me directly saying this, but I thought I’d spare putting his face out there.

No big deal. I had my picture anyway and headed out. I don’t think any type of rigging was going on. I really don’t, no wink-wink, nudge-nudge. It would be pretty hard to do, since there’s no way to anticipate what someone is going to search for.

But, it was sure weird to see such sensitivity over taking pictures of the kiosk in the middle of a very busy mall, filled with smartphone-carrying people.

Was this installed just for Google I/O? I’m checking on that, but I don’t think so. It does appear to be fairly recent, as I can’t find any references to it when I do various searches. Bing has just ramped up its Bing It On campaign, which began originally last September. It may have been installed as part of that or earlier, as kiosks have been in use for months.

Personally, I’ve actually felt this has been one of Bing’s smarter moves, versus the negative Scroogled campaigns. It pits Bing up against Google in a relevancy fight, and Bing can often do very well.

Postscript: Bing has confirmed to me that the installation was put into place specifically for Google I/O. And last night’s repair was to fix touchscreen issues. All sides of the kiosk should now be working today.

Author Images

Many are familiar with authorship images that appear next to some stories, such as this example:

That’s where I did a search for “google bing,” and one of my own articles came up in the top results. My picture appears next to the article because I’ve gone through the steps to identify myself as an author to Google.

Google isn’t rolling out a form of “publishership” that would allow organizations to have the same thing happen for them. In fact, it even pushed back on speculation earlier this month that the rel=publisher mechanism would soon be used to allow this.

The Knowledge Graph Box & Logos

So, what did Google announce today? A new way to get your company logo within the Google Knowledge Graph box, if Google decides to show one for your company. Consider this example from a search for “lear corp,” which brings up a box:

See how there’s a box to the right of the search results? That’s a Knowledge Graph box, one for Lear. The Lear logo is showing, probably because Google figured out a way to guess that it makes sense to show.

Google’s post today says that Schema.org organization markup can be used now as a way for publishers to tell Google what preferred logo they’d like to appear there.

Google’s post didn’t make it clear that this was happening only for the Knowledge Graph box, causing us to originally write that this was going to put logos next to search listings. However, Google has since clarified that logos do not show next to search results as with authorship, but rather, in the Knowledge Graph box that sometimes shows for companies.

Just using the markup doesn’t guarantee that your logo will be used. It only helps suggest this to Google, which makes the ultimate decision.

Going The Google+ Route

Of course, if you really want your logo in that area in a far more controlled and consistent manner, get on Google+. Consider this:

In that example, in a search for “Search Engine Land,” you can see that a Knowledge Graph / Google+ box appears to the right of the listings. Our logo appears in that box, because we’re on Google+.

I’d say for most companies, doing Google+ is going to be a far more effective way to gain logo visibility than using organizational markup. But the option is there, for those who just don’t want to be on Google+.

Note: We updated our story to reflect Google’s clarification that markup doesn’t lead to logos showing next to search listings but only in the Knowledge Graph box. Thanks to Tom Roberts below in the comments, for spotting this.

Google has allowed you to speak your search to it on the desktop and mobile devices for some time, but now it’s going to get smarter — talking back to you and continuing the conversation you started.

Google shared details at its Google I/O 2013 developer event today and also in a blog post that’s gone up.

The idea is that your devices —  smartphone or desktop — will apparently be constantly listening for the “OK Google” command. If you say that, then it knows you want to do a search.

That command is already used as part of Google Glass in a similar way (“OK Glass”) — and only works when you actually touch the device. So, it’ll be interesting to see how desktop computers and mobile devices are supposed to be voice activated. But since both are often constantly being touched, it’s likely the microphone will just stay on and listening until there’s a lapse in activity.

Speak your search, and you’ll get back answers. Again, that’s something you can do now. But what’s different is how Google says you’ll be able to continue the “conversation” to keep searching. From its post:

Soon, you’ll be able to just say, hands-free, “OK Google, will it be sunny in Santa Cruz this weekend?” and get a spoken answer.

Then, you’ll be able to continue the conversation and just follow up with “how far is it from here?” if you care about the drive or “how about Monterey?” if you want to check weather somewhere else, and get Google to tell you the answer.

See, it’ll understand where you are, so “here” is already known without having to say a place name. Of if you just searched for “weather,” then it can guess that you want weather for another city even if you don’t say the word “weather.”

We’ll have more details about the coming change later today, either postscripted here or linked from here to a fresh story, including any specific release date other than “soon.”

This Week Wasn’t Penguin

Publishers have already been wondering if a change in rankings that many have noticed this week was some type of Google update. Google won’t say what, if anything happened.

However, Cutts has ruled out that it was the significant Penguin Update he warned in March would be coming. He tweeted:

@mrjamiedodd nope, no new Penguin update this week.

— Matt Cutts (@mattcutts) May 10, 2013

He went on to say that the next Penguin Update is a few weeks out:

@mrjamiedodd we do expect to roll out Penguin 2.0 (next generation of Penguin) sometime in the next few weeks though.

— Matt Cutts (@mattcutts) May 10, 2013

Counting The Pandas & Penguins

Note that Cutts refers “Penguin 2.0″ as the coming rollout. How can that be, when we’ve had three confirmed Penguin updates already, with Penguin 3 happening in October?

This all goes back to a different update, the Panda Update, which first launched in February 2011. That was Panda Update 1. Of course, we didn’t call it Panda 1 then, because as the first Panda Update, it was just called “The Panda Update.”

Two months later, Google made a huge change to Panda, so the next version was called Panda 2. But when the third release happened, and people started calling that Panda 3, Google said that because the changes to the filter weren’t so dramatic, it would better be called Panda 2.1.

That left it to Google to call the shots on whether a Panda Update was big enough to go through a full point change or not. And that became ridiculous when we got to something like Panda 3.92, last September. As we explained then, when the updates started going to two decimal places, we felt maybe just a straight Panda 1, 2, 3 and so on number order made sense, no decimals involved.

Renumbering The Pandas

When what would have been Panda Update 3.93 came around, we decided enough was enough. We renumbered all the Panda Updates that had happened, regardless of how big they were, believing that was a clearer way forward.

The number no longer reflects whether there’s been a major “generational” change or not. The number is just a common reference point for everyone to use, not some type of magnitude.

For the record, here’s where we are with Panda. The impact each update had on queries is shown, when provided by Google, after the number:

1. Panda Update 1, Feb. 24, 2011 (11.8% of queries; announced; English in US only)
2. Panda Update 2, April 11, 2011 (2% of queries; announced; rolled out in English internationally)
3. Panda Update 3, May 10, 2011 (no change given; confirmed, not announced)
4. Panda Update 4, June 16, 2011 (no change given; confirmed, not announced)
5. Panda Update 5, July 23, 2011 (no change given; confirmed, not announced)
6. Panda Update 6, Aug. 12, 2011 (6-9% of queries in many non-English languages; announced)
7. Panda Update 7, Sept. 28, 2011 (no change given; confirmed, not announced)
8. Panda Update 8, Oct. 19, 2011 (about 2% of queries; belatedly confirmed)
9. Panda Update 9, Nov. 18, 2011: (less than 1% of queries; announced)
10. Panda Update 10, Jan. 18, 2012 (no change given; confirmed, not announced)
11. Panda Update 11, Feb. 27, 2012 (no change given; announced)
12. Panda Update 12, March 23, 2012 (about 1.6% of queries impacted; announced)
13. Panda Update 13, April 19, 2012 (no change given; belatedly revealed)
14. Panda Update 14, April 27, 2012: (no change given; confirmed; first update within days of another)
15. Panda Update 15, June 9, 2012: (1% of queries; belatedly announced)
16. Panda Update 16, June 25, 2012: (about 1% of queries; announced)
17. Panda Update 17, July 24, 2012:(about 1% of queries; announced)
18. Panda Update 18, Aug. 20, 2012: (about 1% of queries; belatedly announced)
19. Panda Update 19, Sept. 18, 2012: (less than 0.7% of queries; announced)
20. Panda Update 20 , Sept. 27, 2012 (2.4% English queries, impacted, belatedly announced
21. Panda Update 21, Nov. 5, 2012 (1.1% of English-language queries in US; 0.4% worldwide; confirmed, not announced)
22. Panda Update 22, Nov. 21, 2012 (0.8% of English queries were affected; confirmed, not announced)
23. Panda Update 23, Dec. 21, 2012 (1.3% of English queries were affected; confirmed, announced)
24. Panda Update 24, Jan. 22, 2013 (1.2% of English queries were affected; confirmed, announced)
25. Panda Update 25, March 15, 2013 (confirmed as coming; not confirmed as having happened)

Panda 25 was the first time Google itself didn’t confirm whether a Panda Update had happened, part of its policy that it wasn’t likely to confirm these going forward, since they rollout over the course of days now. Instead, it was left to third-parties to decide if one had hit.

Did Panda 26 Just Happen?

This also means that the update that’s caused chatter this week might be Panda 26. It might be something else. We don’t feel confident enough to declare it Panda 26 ourselves, which is why our list stops at Panda 25. But with Penguin ruled out, it does suggest that maybe Panda 26 had happened this week.

Or maybe not. Isn’t reading Google tea leaves fun?

Penguin 2.0 Or Penguin 4?

That leads to Penguin. This is how those have gone, so far:

• Penguin 1: April 24, 2012 (3.1% queries affected)
• Penguin 2: May 26, 2012 (less than 0.1%)
• Penguin 3: Oct. 5, 2012 (0.3%)

In our numbering system, regardless of how “big” the next Penguin Update is, we’ll still call it Penguin 4.

It will be big. We know that already from what Cutts has said in the past. In fact, it’s so big that internally, Matt said today that Google refers to it as Penguin 2.0.

From what Cutts tweeted to me:

@dannysullivan might want to downgrade previous Penguin 2 and 3 to Penguin 1.1 and 1.2. Upcoming release is the true 2.0.

— Matt Cutts (@mattcutts) May 10, 2013

@dannysullivan your call, of course, but the internal team here has taking to referring to the upcoming release as Penguin 2.0.

— Matt Cutts (@mattcutts) May 10, 2013

@dannysullivan you’re welcome to call it whatever you like. Just letting folks know how we refer to it internally. :)

— Matt Cutts (@mattcutts) May 10, 2013

Oh dear, but if this next one is the “true” Penguin 2, are we going to make a mistake calling it Penguin 4? I’ll argue not as big a mistake as if we called it Penguin 2.

Why We’ll Call It Penguin 4

See, let’s go back to Panda. In October 2011, we wrote that Panda 2.5 was live. Google hadn’t said it was a massive new change, so that seemed the right number. But the following month, Google said that Panda 2.5 would have been better described as Panda 3.0. We corrected that after the fact — but it would have been easier if Google had called it that way from the start.

We can’t depend on Google to consistently tell us how massive a particular update is, or even if an update happens at all. Because of this, linking magnitude to some decimal-based numbering system seems a mistake.

We have to use something that isn’t going to change months later on. The new numbering system has worked well with Panda, and we’ll stick with it for Penguin.

Ideally, I’d love to see Google itself simply list any significant change with the date it happens and some common reference name. I think that’s useful for publishers — not spammers, but any publisher — trying to understand if they’ve been impacted by something that they should correct. You can’t fix what’s wrong if you don’t have a good sense of what it was.

Postscript By Barry Schwartz: On Monday, May 13, 2013, Google’s Matt Cutts published a video about the future SEO changes to expect with Google in the next few months. The first point Matt Cutts makes is about this Penguin update and how it will go deeper and impact more SEOs and webmasters than the first generation version. Here is that video:

Related Articles

• It’s Panda Update 2.1, Not Panda 3.0, Google Says
• Panda Update 3.92 Rolling Out (Or Is It Panda 20 Time?)
• Google’s Matt Cutts On Upcoming Penguin, Panda & Link Networks Updates
• Panda Update 3.92 Rolling Out (Or Is It Panda 20 Time?)
• Google: We’re Unlikely To Confirm Current Or Future Panda Updates
• Google’s Matt Cutts On Upcoming Penguin, Panda & Link Networks Updates
• If That Was A Google Update You Felt, Google’s Not Confirming It

The new Timelapse site, created by Time Inc. and Google, is a pretty amazing way to see how any part of the world has changed since 1984 through 2012. The site offers amazing animations such as the growth of Las Vegas and the shrinking of the Columbia Glacier, and you can point it at any part of the world you want.

Google explains more about how it has collected the images with the US Geological Survey since 2009 and sifted through to find those with good quality, and without cloud cover, over every part of the world.

Here are some examples of what you can see. Deforestation in the Amazon:

The growth of Las Vegas:

The growth in Dubai:

The Columbia Glacier shrinking:

The site itself explains more about these locations plus makes it easy to jump to them and some other dramatic examples through the editors’ picks you’ll see over the Timelapse map:

Note the last box, with the arrow I’ve inserted pointing at it. That allows you to search for any place in the world. I found it interesting to zoom to UC Irvine, my college, and see how much the campus has grown, since I left. I could also see how parts of Orange County, where I live, has had farmland and hillsides replaced by homes.

Unfortunately, you can’t bookmark and share locations, nor can you create animated GIFs of particular areas using the tool. Both features would have helped with the clear mission of the site, to spread the word about our impact on the planet.

Overall, it’s a great tool, either for exploring the dramatic and sometimes scary changes happening in the world or just to see how much a place you know may have changed or not.

Yahoo’s “Get Out Of Microsoft Free” Card

The WSJ story says that Yahoo has been “quietly” trying to find a way out of its deal with Microsoft. There are some very specific reasons why Yahoo could break that ten year deal early. One of them is the failure for Microsoft to help Yahoo to reach promised “revenue per search” or RPS goals.

Those RPS goals have never been met. That’s part of the news that came out today, that Microsoft has extended for a second time a guarantee to make up for the shortfall caused by not reaching those goals.

Why didn’t Yahoo just take that failure as the “get out of Microsoft free” card that it seemed to offer? From the WSJ story, we have this:

Yahoo is unlikely to get out of its contract with Microsoft until at least mid-2015, the midway point of the 10-year agreement, when either party can potentially opt out, said a person familiar with the contract.

There is another clause in the deal that allows Yahoo to sever it if the revenue per search falls below a certain level, but the revenue per search has been rising and doesn’t appear likely to fall below that level, this person said.

How Low Does RPS Have To Go?

It’s hard to understand how RPS is so low that Microsoft has to pay a guarantee, but not low enough to allow Yahoo to terminate the contract. To understand more, let’s go to the terms of the agreement, specifically about termination:

Yahoo! may terminate the Search Agreement if the trailing 12-month average of the RPS in the United States (the “U.S. RPS”) of Yahoo! and Microsoft’s combined queries falls below a specified percentage of Google Inc.’s (“Google”) estimated RPS measured on a comparable basis or if the combined Yahoo! and Microsoft query market share in the United States falls below a specified percentage

Note the part I bolded. Yahoo has been doing things to boost RPS for the few areas of search ads that it controls, as opposed to the search ads it carries from Microsoft. It may be that the higher Yahoo RPS, combined with lower Microsoft RPS, might keep the overall RPS high enough to prevent Yahoo from acting on the termination clause but still low enough to cause Microsoft to pay guarantees.

My separate article on Marketing Land today gets into this whole RPS issue in more depth:

• Why Yahoo Will Never Reach The “Revenue Per Search” That Microsoft Promised

Why Yahoo Can’t Leave Quickly: No Tech & Short-Term Cost

Maybe that’s the reason. A more likely reason, in my view, is that Yahoo simply can’t leave Microsoft yet.

Yahoo has no substantial Web search technology left. It gave that up as part of the deal with Microsoft. Drop Microsoft, and Yahoo would have to quickly find a Web search partner to power the bulk of its listings that are so very important to a major search engine. Google is the only real choice, and Yahoo might not be able to go to Google without regulatory approval.

That leads to the financial hurdle. While Yahoo waits on an approval process, it also has little search ad serving technology left. That also was largely handed over to Microsoft. Even if Yahoo could go without Web listings, it would still be stuck having to redevelop search ads — taking a loss until that was done.

Bottom line? I’m sure Yahoo is quietly seeking alternatives to staying with Microsoft but also quietly doing it because it’ll take time to extricate itself from Microsoft in a way that won’t be costly in the short term.

Related Articles

• The Microsoft-Yahoo Search Deal, In Simple Terms
• What’s Yahoo’s “Plan B” For Search?
• Ironically, Search Might Be Less A Priority At Yahoo As Google’s Marissa Mayer Takes The Helm
• As The Yahoo-Microsoft Search Alliance Falls Short, Could A Yahoo-Google Deal Emerge?
• Eric Schmidt Floats Idea Of A Yahoo-Google Search Partnership
• A Yahoo-Facebook Search Partnership? Reality Check Time!
• Yahoo Extends Microsoft Search Deal & Revenue Guarantees
• Why Yahoo Will Never Reach The “Revenue Per Search” That Microsoft Promised

Yahoo Declares An Extension

Reuters spotted the extension as part of Yahoo’s latest quarterly financial filing:

On April 30, 2013, Microsoft extended the RPS Guarantee in the U.S. for an additional 12 months commencing April 1, 2013.

Second Time Guarantee Extended

To translate into plain language, when Yahoo & Microsoft signed a search deal in 2009, Microsoft promised that Yahoo would earn a set amount of money for each search that happens, a “revenue per search” or RPS. If this didn’t happen, Microsoft agreed to make up the difference, what’s called the RPS guarantee.

Microsoft has failed to deliver the expected RPS ever since the deal was signed. It was originally set to expire 18 months after the “paid search transition” from Yahoo to Microsoft had been deemed concluded. But in the fourth quarter of 2011, Microsoft agreed to extend the guarantee through March 31, 2013.

My past story explains more about all this, in depth: As The Yahoo-Microsoft Search Alliance Falls Short, Could A Yahoo-Google Deal Emerge?

That leads to today’s news, that Microsoft has extended the guarantee now a second time, taking it out through April 1, 2013.

Why? Quite simple. If Microsoft can’t deliver, Yahoo potentially can walk away from the deal and go to someone else … like Google.

Will Yahoo Go Google In 2014?

The date to watch is April 1, 2014. If Microsoft fails to deliver for a third time on RPS, Yahoo potentially could abandon the ten year deal. For more on that, see our two related posts:

• Why Yahoo Will Never Reach The “Revenue Per Search” That Microsoft Promised
• Even If Yahoo Wants To Leave Microsoft, Here’s Why It Can’t

Google relaunched its app last October with Siri-like voice recognition and readback, but after peaking at number 16, it started dropping down to around the 150 mark last Sunday.

With Google Now being added, the app has risen up to number 10 in the most recently downloaded free apps list. Coincidentally, it also just got named the 10th most downlaoded free app ever for the iPhone.

Whether Google Now functionality helps it stay in the top 10 remains to be seen, of course.

Where Has All The Google Traffic Gone?

BuzzFeed said that it looked at traffic to 200 publishers in its network and found a marked decline in Google traffic from September 2012 onward. The network includes sites like Rolling Stone and The Huffington Post. My mind is blown that the Huffington Post is somehow in the BuzzFeed Network, by the way. Do they aggregate each other?

Here’s the key chart:

Look at that: Facebook ascendant, Google in decline! That seemingly matches up with what some news publishers have said in the past few years, that social is a bigger driver of traffic to them than search. But they rarely qualify that even in these cases, it’s usually not that search went into decline. It’s that while search keeps growing, social has turned into an entirely new audience stream they didn’t have before, one that’s also rocketing up. The pie got bigger, so that both the search and social slices grew.

But that chart suggests the opposite, that Google is plunging, Why. Why? Why!

BuzzFeed tries to puzzle it out. Is social growing? Is SEO no longer king? Is SEO just harder? Is “Dark Social” masking even more social gains?

Blame Google For Google’s Seeming Decline

Try Dark Google perhaps being to blame. That’s how Google, back in October 2011, started withholding from publishers the search terms people used to find them on Google. Withheld, that is, unless you’re an advertiser.

Google has some strong arguments that this is all about privacy, but there are plenty of loopholes that are still left open. Meanwhile, Google got caught off-guard by something unexpected. Apple.

See, when Google stopped showing search terms, it still at least did things in a way that let publishers know that traffic came from Google Search generally. What it didn’t expect was that Apple might seize upon Google’s “private” search for use when iOS 6 came out last year.

How Mobile Safari Turns Google Traffic Into Direct Traffic

When Apple did that — in September of last year — Google didn’t appear to be ready for it. Rather than ensure that traffic from mobile Safari still at least passed along the fact that someone had done a search at Google, even if you couldn’t know the exact search, Google let the click show no referral at all.

In other words, before September, if someone searched on mobile Safari at Google and went to a web site, they’d show up as if they were Google traffic. After that, they showed up as if they were a “direct visit.”

Now look again at that chart. You can see the “decline” starts in September and stays pretty consistent each month other than in December. Other things might have helped mask the impact of the Safari loss that month. After all, you can see in previous months that there are times when Google has seen gains and drops.

But from my perspective, if you’ve seen a sudden drop in Google traffic over the course of several months — and it’s not due to a Google penalty — perhaps it’s the Safari issue just getting worse. It’s a problem that Google has strangely allowed to fester for months now, one that seemingly is in Google’s ability to fix.

If you really want to know, you’d look at direct traffic. Is that rising? Or look at mobile Safari traffic. Is that coming up higher in relation to the traffic that’s dropping from Google?

Of course, mobile Safari traffic might not be big enough to a particular site to explain all this drop. Many other factors might come into play, including that sites in the network are indeed receiving less traffic or perhaps other things that might be causing some of that Google traffic to be “dark.”

NOTE: I’ve updated this story and headline slightly from the original to better qualify that the mobile Safari issue might be to blame, not that it is definitely to blame. I think it is a contributing factor, but I can’t tell from afar, and it might be other things also at work.

Searching For The Page By URL

The easiest test was to see which of them had the page listed by searching for it by its URL, which is:

http://info.cern.ch/hypertext/WWW/TheProject.html

It looks like this, by the way:

Google found it when I searched for its URL:

Bing did not:

End of test! Google lists the page, has it included in its collection of billions of pages across the Web that were all born from it, while Bing does not.

Searching The Way People Really Search

Most people who are looking for the first-ever webpage wouldn’t do so using the URL, however, especially when that URL hadn’t even worked for over 20 years, until a few days ago.

My assumption is that most people seeking it would do a search for something like “first web page” or “first ever web page.” To test this, I started typing “first web…” into both Google and Bing, to see what they suggested. Those suggestions come from the most popular searches that people actually do.

I got this back:

As it turns out, people seem to be seeking the page by searching for terms like:

• first website
• first website ever
• first website ever made

Finding The First-Ever Website

I did a search for “first website,” and here, some might argue that Bing wins. This is what I got from Google:

The arrow points at the info.cern.ch page, which makes sense. CERN had been redirecting the URL for the first website to the info.cern.ch site for years, until it changed that yesterday, according to a CERN blog post that explains more.

But, CERN’s info site isn’t the first-ever website. It leads to it now, yes. But, it’s not the actually first website. Now consider Bing:

Bing lists the info.cern.ch site, just as Google does. But further down, it also lists a copy of the first-ever website, a copy that’s been located off CERN until this time.

That copy is located here:

http://www.w3.org/History/19921103-hypertext/hypertext/WWW/TheProject.html

It’s hosted by the W3C, an international group that took over leadership of the World Wide Web’s development after its origins with CERN. That copy is actually what CERN used for its restoration project.

If you are trying to find the actual first-ever website, using those words, Bing’s actually pointing to it better than Google. It’s not pointing you to it at the restored location, but neither does Google. At least, Bing gets you to an exact copy of the first-ever page, where it has long been archived.

Stand By For Change

This is likely to change in the near future. For one thing, CERN now links to the first-ever website using those exact words, like this:

Having those words in the anchor text — in the link — is a very powerful link signal to both Google and Bing that the page being linked to should rank for those terms, when it happens from an influential site like CERN.

In addition, as news stories about the restoration project begin to disappear, that will give a greater chance for the actual page to start moving into Google’s top results.

So, stayed tuned. And while you’re waiting, you can learn more about the origins of the WWW from CERN’s information site about the web, its page about the restoration of the first website and from the W3C’s A Little History of the World Wide Web.