A couple of days ago I posted on Google’s evil scale and although the post was mostly meant to be light-hearted, I also suggested that, at some point, Google was bound to start behaving like a big company, and that this was likely to be triggered by increasing criticism of the company by journalists and activists. It appears that Michael Arrington at Techcrunch agrees with me. He wrote the following today under the heading of “Will 2008 be Google’s end of innocence?“:

2008 may be the year that Google’s innocence ends, as media and governments start to cast a less forgiving eye at the behavior of the company that controls 60% of the search market and perhaps as much as half of all online advertising revenue.

…

There’s no getting past the fact that Google has out-competed everyone in the search game, and is justly collecting the economic rewards of that effort. But society loves to tear down their heroes just as quickly as they supported them as underdogs.

This may be the year that things change for the ten-year-old Google. Their days of innocence may be over - perhaps Yahoo, or Firefox, are the apples that they should not have bitten into.

As I mentioned a couple of days ago, at some point Google will cross over from innovative upstart to established incumbent, and many other changes will follow. I tend to agree with the assertion Arrington makes here: I think when that happens, which is a shift that will happen in people’s minds as much as in the corridors of the Googleplex, Google will find itself facing many more challenges than it currently does. And how it weathers that shift will be the indicator of whether Google is likely to be the next Altavista or Mapquest (companies which once dominated but then declined) or the next Microsoft (a company which, for all its faults has nonetheless remained tremendously dominant and very successful).

I found this pretty entertaining - I guess it’s a little dated at this point but still very relevant in light of the ongoing debate about Google’s position on Chinese censorship. The attention to detail here is impressive - it’s been created to look just like the corporate pages at Google (although all the links point back to the creators at Lot49.com. The introduction, written as if a Google press release, follows:

At the 2006 World Economic Forum in Davos, Switzerland, CEO Eric Schmidt discussed Google’s decision to censor its search results. During his speech, he mentioned that we used an “evil scale” to weigh our actions. Applying that scale, we concluded that to withdraw from China would be “worse evil” than participating in censorship.

Having received a number of queries about our evil scale, we present an explanation here. Our scale divides evil into 15 degrees because we like hexidecimal and because it’s convenient for representing shades of gray online. Also, we find that the shift from numeric to alphabetic characters is useful in separating bad things from those that are really terrible.

We determined that removing certain information from search results on Google.cn rates a 6 on our scale. Withdrawing from China qualifies as an 8. Disorganized information helps no one. In fact, it is a detriment to society. When all messages are equally probable, entropy is maximized: H(M) = log | M |. We are committed to fighting entropy by organizing the world’s information. Working with Chinese censors will help us achieve our goal.

I wonder how Google feels about this stuff. They don’t yet have the reputation of other large corporations like Microsoft and AT&T for being overly sensitive about criticism of their products and policies, but you do wonder whether they will eventually move in that direction. It’s tough for any big company that strongly believes in its own mission and attitude to take criticism. So far Eric, Larry and Sergey seem to be fairly level-headed about it - let’s hope they stay that way.

I just tried to sign up for Google’s new FriendConnect service. I filled in all the details and clicked “Submit” and then up came this page:

It appears Google is using one beta service (Google Spreadsheets - specifically, the feature which allows you to use forms to create spreadsheets for databases) to register for another (FriendConnect). As a result, “something bad happened.” This probably isn’t best practice for a hotly-anticipated new service from Google, much as I understand the urge to eat one’s own dogfood, as it were. Wouldn’t a standard web form with a more robust backend have done the trick?

At any rate, I’m looking forward to trying out the service if and when I can get it working. Looks like an interesting approach to “socializing” non-social networks, but a lot will come down to how it works in practice. I’m also looking forward to trying the other similar initiatives that were somewhat suspiciously all launched within a few days of each other (MySpace Data Availability and Facebook Connect).

Interesting news today on Verizon’s approach to the open access requirements associated with the 700MHz spectrum it won in the recent auction:

On Friday, Google urged the FCC to block Verizon Wireless’ $4.7 bil. successful bid for the C Block band of spectrum in the recently completed 700 MHz auction unless Verizon is forced to agree that open access rules apply to handsets it provides its own customers. Specifically, Google claims that Verizon Wireless has no intention of abiding by the open access rules governing the C block spectrum for devices it gives to its own customers and that the FCC should condition Verizon’s grant upon a clear commitment that Verizon will not exclude these handsets from the requirement.

This echoes in reality what I had said might happen in a previous post a few weeks ago:

…isn’t Google mistaken here? Is it assuming that the FCC’s open access rules go further than what Verizon Wireless had already agreed to do? The FCC’s rules are unfortunately vague, and it may be counting on a more favorable interpretation of them than Verizon is.

Google and Verizon have had conflicting opinions on the meaning of the FCC’s open access requirements - each taking the position that most clearly reflected its own views - Google’s being the most expansive interpretation possible, and Verizon’s being the most minimalist. However, in my previous post, I pointed out that even though the rules are vague, they certainly strongly suggest that Verizon will have to go beyond its existing open access project. For example:

Wireless service providers subject to this requirement will not be allowed to disable features or functionality in handsets where such action is not related to reasonable network management and protection, or compliance with applicable regulatory requirements. For example, providers may not “lock” handsets to prevent their transfer from one system to another. We also prohibit standards that block Wi-Fi access, MP3 playback ringtone capability, or other services that compete with wireless service providers’ own offerings. [emphasis mine]

At any rate, looks like we’re in for more fun and games, and more uncertainty for Verizon and its customers. Not what anyone would have wanted, and it could all have been avoided if the FCC had just been clearer about these requirements up front. Verizon can now reasonably argue that it bid based on its understanding of these rules and it’s too late to change that understanding now.

I am constantly astonished by the fact that many websites still use Mapquest on their “directions” pages. It makes me wonder why they do so, when surely no-one really uses Mapquest anymore when there are alternatives like Google Maps and Yahoo Maps around. For several years now, those two alternatives have been considerably better options, with a much more usable interface (both on PCs and on mobile devices) and much higher quality maps (Mapquest’s were until very recently still those funny line drawings where roads are shown as single lines instead of two dimensional objects although this appears to have changed in the last couple of months).

And yet, according to Hitwise, Mapquest still has an over 50% share of the online mapping market. Google Maps, while coming up quickly, is still around 20%. Yahoo Maps, which had been in second place, has recently dropped to third, losing subscribers to Google along with Mapquest itself. Microsoft’s Live Maps product, meanwhile, is hovering around 3% of the market.

I found this astonishing, because I’m not sure I’ve ever used Mapquest, and I certainly never have in the last 5 years. For a time, Google Maps and Yahoo Maps were playing leapfrog in courting my attention by launching new features alternately. But Mapquest was never a serious contender. So why is Mapquest still such a major player?

I think the answer lies in the fact that many people on the Internet are inherently inert when it comes to choosing service providers. Their inertia ties them to the first service they used that worked in a particular space - be it mapping, search or photo sharing. They don’t actively seek out alternatives and so never realize that there are better services available. In the case of Mapquest, it was so dominant in the early days that its name transcended its brand and became generic in the manner of Kleenex, Hoover or Ziploc (”shall I give you directions?” - “no - I’ll just Mapquest it”).

I think this same inertia is what has allowed AOL (coincidentally, the owner of Mapquest) to continue to exist as a walled garden provider even when the same content and services that you can pay $10 to $26 a month for is now available for free at aol.com. The same group of subscribers probably make up much of the customer base for both Mapquest and the old AOL service. And this group of relatively inert Internet users is large and probably growing as more seniors and others who are less adventurous on the net come online.

This all gives a massive advantage to the first company that makes a significant new service work well enough for these relatively conservative users of the Internet. Google is able to overcome this advantage in the case of mapping and other areas such as email because it already has a strong entrenched position in search and is good at leveraging its strength across its properties through those links at the top left of each of its pages.

This is in contrast to Yahoo!, whose website is so cluttered that finding new services is very much more difficult. There are lessons to be learned both from Mapquest and from Google here. From Mapquest we can learn that being the first good provider of an online service is a massive advantage. But from Google we can learn that it is possible to overcome that advantage by leveraging mind share in an existing market into an adjacent one. And we can also learn that the time to market for competing services is very much shorter today than it was when Mapquest first launched, so that the first mover advantage is being eroded over time.

Service providers on the Internet have essentially two choices: they can target this large group of inert users, which takes a long time to jump on a bandwagon but will then ride it for many years, or to target the equally large group of more technically savvy users who are more adventurous and therefore faster to adopt a new service but also faster to jump ship when something better comes along. A third option, the holy grail, is to produce a service which serves both markets and can therefore grow quickly but also retain a large base over time. But that’s a rare service indeed.

Rick Whitt of Google has come clean about its strategy in the 700MHz auction and confirmed what many suspected - that Google deliberately bid up the price of the C Block, even upping its own bid in the absence of a higher competing bid several times, in order to trigger the open access provisions.

Given that it must have been fairly clear early on that Verizon Wireless was the other bidder, and it’s already initiated its open access program, what did Google really gain by doing this, other than using its own cash to force Verizon to part with more of its money? One of three scenarios must hold true:

1. Google wasn’t sure Verizon was the other bidder and wanted to make sure any other bidder (AT&T) would be subject to open access too
2. Google merely wanted to force Verizon to pay more for the spectrum because there weren’t any other serious bidders for it
3. Google believes that the open access provisions attached to the C Block will require more of Verizon than it has already announced it will provide.

The most likely scenario is 3. But isn’t Google mistaken here? Is it assuming that the FCC’s open access rules go further than what Verizon Wireless had already agreed to do? The FCC’s rules (see page 89) are unfortunately vague, and it may be counting on a more favorable interpretation of them than Verizon is. But there are one or two areas where Verizon has not yet agreed to go as far as it will now be required to:

Scope of the requirement for open platforms for devices and applications. Wireless service providers subject to this requirement will not be allowed to disable features or functionality in handsets where such action is not related to reasonable network management and protection, or compliance with applicable regulatory requirements. For example, providers may not “lock” handsets to prevent their transfer from one system to another. We also prohibit standards that block Wi-Fi access, MP3 playback ringtone capability, or other services that compete with wireless service providers’ own offerings. Standards for third-party applications or devices that are more stringent than those used by the provider itself would likewise be prohibited. In addition, C Block licensees cannot exclude applications or devices solely on the basis that such applications or devices would unreasonably increase bandwidth demands. We anticipate that demand can be adequately managed through feasible facility improvements or technology-neutral capacity pricing that does not discriminate against subscribers using third-party devices or applications. In that regard, we emphasize that C Block licensees may not impose any additional discriminatory charges (one-time or recurring) or conditions on customers who seek to use devices or applications outside of those provided by the licensee. Finally, C Block licensees may not deny access to a customer’s device solely because that device makes use of other wireless spectrum bands, such as cellular or PCS spectrum. However, we also note that, in accepting a multi-band device for use on its network, a C Block licensee is not required to extend the requirement for open platforms for devices and applications to other spectrum bands on which the provider operates.

However, the FCC goes on to limit the scope of these rules:

We emphasize that we are not requiring wireless service providers to allow the unrestricted use of any devices or applications on their networks. In particular, we are mindful of the risks network operators face in protecting against harmful devices and malicious software. Wireless service providers may continue to use their own certification standards and processes to approve use of devices and applications on their networks so long as those standards are confined to reasonable network management. For example, providers are free to choose their air interface technology, and to deny service to devices or applications that cannot operate on the same technology, since such a restriction permits significant network efficiencies without significantly reducing consumer access to services and features. We also recognize that wireless providers have legitimate technical reasons to restrict particular non-carrier devices and applications on their networks, specifically to ensure the safety and integrity of their networks. In particular, we believe that it is reasonable for wireless service providers to maintain network control features that permit dynamic management of network operations, including the management of devices operating on the network, and to restrict use of the network to devices compatible with these network control features. Standards to ensure that network performance will not be significantly degraded would also be appropriate.

If I were Verizon, I would be pretty annoyed with Google at this point, whatever its rationale. It artificially bid up the price of the spectrum, triggering both a higher price paid by Verizon and the open access requirements, even though it apparently never had any plans to own the spectrum. I’m not sure it would have any legal grounds for taking action on this, but it certainly appears that Google abused the system for its own advantage and to Verizon’s disadvantage.

At any rate, Verizon is pressing ahead with its plans to use the 700MHz spectrum for its LTE network, as is AT&T with its 700MHz spectrum, and so far it isn’t complaining too hard about those open access rules. Since it has done an about face on the question of “openness” perhaps it is willing to embrace these additional conditions too. But I would guess that it will fight for the loosest possible interpretation of the remaining conditions when the time comes, while Google will probably put some high-paid lawyers on the other side.

Just did a Google News search for “Huckabee” to see if I could find the transcript from his concession speech last night, and Google News came up with this:

So, it seems once you drop out of the presidential race you also drop off Google News’s search results. I wonder if this is a funny quirk resulting from some tweaking in Google’s News database (does he have to somehow be shifted from the bucket marked “current candidates” to “past candidates” or something?). Presumably they’ll fix this soon…

Google Sites launches today - the result of the acquisition and subsequent integration of JotSpot, a hosted wiki provider. It promises to be a useful application, allowing organisations to pull together calendars, wiki-style websites, documents (through Google Docs) and so on into an intranet-style site. It will bring Google ease of use to the hitherto cumbersome process of creating a wiki and have the benefit of integrating other Google properties and services.

However, one major thing is missing, and that’s the ability to create a Google Site for anything other than a school or company. The login screen (shown here) asks for your “school or work email address”. What it means is “give us your domain name so we can figure out which Google Apps customer you’re part of.” All this new functionality is tied into the wider Google Apps family, which is designed for organisations which want Google to more or less provide hosted domain services.

There’s a logic to this, and if you’re an organisation using Google Apps already, you’ll want Sites to tap into that infrastructure. But what if you’re not using Google Apps already, and what if you’re an organisation other than a business or school, such as a charity, a sports team, a PTA or other similar entity? You won’t want to own or register a domain name, and the degree of integration you’ll need with other users will be minimal. You’ll still want to share calendars and news, create a website and use at least some of the other functions of Sites, but this won’t be possible for now:

Google Apps Team Edition is designed to help users easily collaborate with colleagues and classmates within their organizations. Because the service is intended for use within schools and businesses, you won’t be able to sign up using an individual email address, such as:

• user@gmail.com
• user@yahoo.com
• user@hotmail.com
• user@aol.com

If you wish to use Google services with your personal email address, we recommend creating a Google Account. Please note that if you’re using Gmail, you already have a Google Account and can add other services to this account. Google Accounts currently include all the services offered in Google Apps except for Google Sites - at this time Sites can only be used in Google Apps.

If you want to use Google Apps instead of a Google Account and you don’t have an email address issued by a school or business, you may wish to try Google Apps with domain registration.

In other words, the only way you can get around this at the moment is by registering a domain (with Google or otherwise) for the sole purpose of using Sites. This feels like a totally unnecessary imposition for an organisation which wants to use the functionality but doesn’t want a domain.

As an example, I work with other adults to run our Church congregation’s program for teenage boys. We use Google Docs to keep track of calendaring (ideally we’d use Google Calendar, and perhaps someday we will) and various other methods such as group emails and several Facebook groups to communicate with each other and the boys. How much easier would it be if we could centralise everything in Google Sites - calendars, announcements, the latest information on upcoming events and so on? But we can’t do it without registering a domain first. How frustrating. Hopefully Google will expand the availability to other groups soon.

I attended a couple of hours of the Money:Tech conference organised by O’Reilly Media in New York today. Tim O’Reilly himself - originator of the phrase Web 2.0 - was the keynote speaker, and was followed by a chat with Jim Cramer, host of Mad Money, founder of TheStreet.com, etc. etc. The conference was about Web 2.0 and financial services, and O’Reilly started out by talking about Web 2.0 and what it means to him. Ovum certainly has a definition of it, which revolves around four parts - social, business, content and technology models which define Web 2.0 services and sites. However, O’Reilly has a simpler definition, which stays away from specific technologies and services, and is simply this:

Web 2.0 is really about harnessing collective intelligence. It’s about creating a network-effects driven data lock-in with accelerating results to the winners. [I'm paraphrasing based on my notes but that was the gist]

In this way, O’Reilly says, it’s similar to Sun CEO Scott McNealy’s “red-shift” concept - that is, as you start to successfully differentiate yourself in something, your lead over the competition begins to grow ever more quickly. It’s all about creating business models which thrive off network effects - examples, according to O’Reilly, include Google (where the network effects come from the number of links people make), eBay (where the critical mass of buyers and sellers is the biggest barrier to competitive entry), Amazon (where he suggests the reviews are the key network effect) and so on. The value lies in accumulating data which leverages network effects in such a way that it is very hard for competitors to emulate what you have done.

Another major theme at the conference was open source software, and a debate during a panel session focused on whether open source adds or destroys value from a market. There were arguments on both sides, but it’s pretty clear to me that it destroys value for existing players, since it replaces proprietary products priced at a premium with free open source products. At the same time, it creates new opportunities for players which didn’t have the in-house resources to develop their own software, and it reduces the cost of doing business for everyone, which increases liquidity and therefore provides broader benefits.

So, how does all this apply to the OpenSocial program, the Social Graph API and efforts to create data portability? Do these effectively do to value in the Web 2.0 world what open source is doing in the software world? Does Facebook’s value proposition go away? Part of the answer may lie in something else O’Reilly talked about, which is Clayton Christensen’s “law of conservation of attractive profits,” which states:

When attractive profits disappear at one stage in the value chain because a product becomes commoditized, the opportunity to earn attractive profits with proprietary products usually emerges at an adjacent stage.

This would suggest that when open source enters a market, the value flees to the adjacent markets. And when data portability enters the Web 2.0 market, value will flee away from the Facebooks and MySpaces and to - where?

I would argue, as I’ve suggested in other entries, it flows to those best able to make use of the new technology - data portability - to create new services which thrive off it. I think this is the logical conclusion, and it’s another reason why Facebook, MySpace and others need to create value in something other than the information they hold about their users, because that will soon become commoditised and easily duplicated. They need to leverage that data in ways others can’t because of special sauce they themselves have concocted. It’s not clear to me that they have figured this out yet, hence (perhaps) their resistance to full data portability. But they’d better figure it out quick or that value really will go to someone else (and who would bet against Google here?).

I’ve just started reading Googler Brad Fitzpatrick’s essay on the Social Graph problem and his accompanying slides. While I agree with a lot of what he says, I find that one of his big assumptions (as stated in the slides) is

some edges/nodes secret (but most public!)

This remains one of my biggest beefs with social networking - that the assumption is we want everything public. I sound unfairly old when I say this, but I’m just not comfortable with the younger generation’s tendency to put everything in the public domain. I have very separate groups of acquaintances (they’re not all “friends” in either sense) in real life and would like to maintain the same distinction online. I will put some of it in the public domain (like this blog), allow Google’s bots to crawl it and so on, but just as I have a personal blog which isn’t linked to here (or crawled by Google), I want to control access to my information, even to the extent that “most” would be “private” in the sense of being shared with certain people but not everyone. And I think this is a key feature of the endgame of social networking I discussed in an earlier post. I also wonder how Google will participate when some of the data is password protected. I think we still need the current model of providing behind the scenes authorisations for one application to access another to download key data, and I don’t want all that going through Google.

Update: looks like I’m not the only one with this concern. Although this article takes a slightly different tack, the problem it points out is essentially the same - not everyone wants all data tracked and searchable by Google. Having said that, there are ways to put up stop signs respected by the Google search bots, but not everyone knows about them and certainly not everyone would have thought it necessary before the Social Graph API came along.