CNet says today's the day. Google will be launching a subscription-based version of Google Apps for Domains. It will include Google Docs and Spreadsheets integration, 10 GB of Blackberry-accessible Gmail, 99.9% uptime, and 24/7 phone support for $50 per user per year. Google will also provide APIs for migrating data, enabling single sign-on, etc. (Update: the service is live; a free trial is available until April 30, 2007.)

Google will continue to offer free versions for consumers (no premium features) and educational institutions (with premium features, but less storage).

Let's say I work at a company with 10 employees. Google Apps (which include website builder and group start page/calendar in addition to Gmail/GTalk and document/spreadsheet sharing) would cost $500/year. What would the same budget buy at GoDaddy?

* $43.08: annual cost of cheapest hosting plan. While this includes 500 email accounts, each offers only 10 MB of storage, which means I'll probably need...
* $29.99: email hosting for 5+ mailboxes. Unfortunately, this would only give me 2 GB of *total* storage. On the bright side, GoDaddy does throw in...
* $0.00 (19.99 if purchased separately): group calendaring for up to 15 users
* $9.99: cost of 1GB online folder for file sharing.

If I'm not familiar with FTP and HTML, GoDaddy's WebSite Tonight site builder will set me back another $12. Which makes the bundle above ~$400 cheaper than Google Apps. On the other hand, it doesn't include IM, or web-based word processing and spreadsheet apps, or a shared start page that's customizable with 10,000+ widgets.

If I were Bob Parsons, I might have other worries besides competing with Google for customers. A couple of years ago, Bob told the New York Times that:

"Try to call Google and actually talk to somebody. It's not their forte. Now, they could acquire that, but at the moment it's a problem I don't have to deal with, so I'm not thinking about it. If they do, we'll go to work."

It seems the time has come for GoDaddy to get to work, particular since it's looking to hire 560 new employees this year. While Bob handed out $1.3M in bonuses in 2006, as well as "new vehicles, trips, electronics and cash prizes, with taxes paid", such incentives might pale in comparison with stock options from a certain company in Mountain View that might also be in the market for experienced call center staff...

PS - Read on Techcrunch that GE and Procter & Gamble are Google Apps customers. The New York Times and InformationWeek both call Google Apps an attack against Microsoft, but the folks in Redmond aren't the only ones who need to keep an eye on this. See this Techmeme thread for more related discussions.

That's what Business Week says. And Garett Rogers points out on his ZDNet blog that a Google Apps for Enterprises page is already online. Pixar is reportedly dying to ditch its homegrown messaging system for Google-powered seamlessness.

Google is expected to charge "a few dollars" per person per month. (Ars Technica says most likely less than $3, to stay competitive with Zoho.) Considering large scale deployments such as Arizona State University's 65,000 users, Google's 300-person enterprise group might soon be generating significant revenue.

In fact, Nick Carr thinks that Google has far greater ambitions than Google Apps as we know it. He's betting on additional SaaS acquisition following last year's JotSpot deal.

Everyone's asking how Microsoft's Office Live (now at 250,000 users) will hold up against Google's assault. Another good question is how the continued evolution of Google Apps will affect low-cost hosting providers? (GoDaddy and eNom are Google's domain registration providers at least. eNom also has Office Live's domain registration business.) And no, 24/7 support won't be enough for maintaining a competitive advantage; Google's planning to offer that as well.

PS - Just came across Phil Wainewright's very interesting list of mega traps for Google Apps. He wonders how committed Google is to generating subscription versus advertising revenue, and he worries that Google will ditch the paid apps business after setting loss-leading pricing expectations that no other vendor will be able to match.

Red Monk analyst Anne Zelenka wrote an article about Amazon Web Services for Web Worker Daily. She tagged it under "Software" - because software is what it takes to build web apps. Yes, Amazon's pay-as-you-go storage and computing solutions are cool as far as hosting infrastructure goes, but Anne thinks much more is needed:

"Amazon hasn't yet come out with a user-friendly tool for making mashups - web applications that combine data and interactivity from various sources on the web... What we future web tycoons need is a platform that hosts mashups on Amazon's S3 and EC2, that brings together other features of AWS - like search capability and message-passing for distributed program control, and ideally ties in some Mechanical Turk like functionality for leveraging human intelligence online."

In other words, the perfect environment for an online business might not consist of truckloads of lumber (aka bandwidth and storage), with which they can construct whatever they wish (note: tools sold separately). Instead, what they need are Lego-like building blocks for assembling, dis-assembling and re-assembling into the next big thing. (Hey, isn't that what enterprise IT people call SOA? Well, here's an article on Amazon and SOA.)

You'll notice that from Anne's point of view, Amazon has already become the base case for improving upon. Where does this leave hardware-only hosting solutions?

Of course, there will always be customers who prefer to start at the lumber yard and build what they need from scratch. My neighbor Irakli, for instance, is a true do-it-yourself-er who's in the process of making speakers and amplifiers by hand. He detests vendors who try to sell him value added services. Verizon, for instance, is "straying away from its core competency" by calling itself "your broadband and entertainment provider". When I pointed out that land lines and even cell towers are facing increasing competition from VoIP and voice-over-WiFi, Irakli suggested that Verizon should just tighten its belt and compete by lowering prices. If it's cheap enough, it'll win!

It's a good thing Verizon's taking a different approach. Katie Fehrenbacher from GigaOM says IDC has crowned Verizon Wireless as the market leader in customer spending on the value-added data services that Irakli so abhors. But back to the web hosting market... Do you want to run a lumber yard, or sell Lego blocks?

See also: Web Hosting Core Competency, Part 1.

I've been playing connect-the-dots with the following bits of information:

1. Some time ago, Alert Logic CTO Misha Govshteyn commented on my "functional hosting" post that he's not sure there's such a thing as functional hosting. Photobucket, YouTube, Live Journal, etc are software companies. They live and die by the ingenuity of their software developers - not the content storage which they happen to provide as one of many features.

2. The two blog posts I've most enjoyed writing were about the evolution of eBay's and MySpace's architecture. Misha was right. Neither could have made it to where they are today without having exceptional software development talent on staff. Each has been able to rewrite its entire platform time and again - while giving millions of users uninterrupted access to a complex service with countless moving parts.

3. The Planet is facing a bit more challenge on a similar-ish endeavor. Unanticipated glitches forced the company to put its just-launched customer portal on hold and roll back to separate legacy interfaces for The Planet and EV1 customers.

I'm not mentioning this to give The Planet a hard time. I think you'll agree that they're not a software company. In that respect, they're more the norm than an exception within the web hosting industry. Traditionally, the duty of a hosting provider has been to consistently implement a standard set of provisioning/support/billing procedures - rather than to churn out 200,000 lines of new code every single month. But is that an adequate scope of work going forward?

4. I started thinking about this after reading Egenera CTO Peter Manca's comments about yesterday's Sun/Intel deal:

"The bottom line is that it's no longer about the hardware... customers can play hardware vendors off against each other and get the lowest pricing as a result. Whoever happens to have the hottest chips at the time of purchase will win. If every hardware vendor supports both AMD and Intel, all the better, as that means more vendors to play off each other... Software is where the real value add is. Companies that just sell inexpensive servers as their only value have a difficult future. Taking these commodity components and adding real value on top is where the future lies."

Peter's talking about Sun and HP and Dell, of course - but doesn't his assessment apply equally to web hosting providers who sell bandwidth/storage space on such gear? And if so, should The Planet and its competitors move beyond hardware - beyond professional services, even - into the application space? After all, Webmail.us said it picked Amazon rather than Rackspace as its storage vendor because Amazon has a web services stack that Webmail.us could build on top of...

PS - I hope The Planet is offering legacy EV1 customers a discount on new servers while it readies the combined portal for re-launch. If someone has to use two separate interfaces to manage new versus existing servers, they might be that much more inclined to look beyond their current provider.

I signed up for Mozy the other day. It's a remote backup service - of which you've seen plenty. BUT-

* It asks what types of files (music, for instance, or spreadsheets) you want to back up, then auto-detects new and changed files and offers block-level differential backup.
* It offers continuous backup, even on open and locked files
* It automatically reduces its own resource usage and throttles bandwidth when the computer is in use
* Your data is encrypted during transport (128-bit SSL) and while in storage (448-bit Blowfish)

In short, it's smart. And affordable. $4.95/month lets you back up as many files as you've got on one PC.

I thought of Mozy when I walked by SwapDrive's office yesterday; they're right in my neighborhood. I was a customer of theirs ages ago, but I canceled my $7.50 account when I maxed out my 100 MB limit. They wanted $50/month for upgrading me to 2GB. Can you believe it??

More importantly, while its website brags that it's the first company to offer web-based remote backup, its technology hasn't changed at all since then. I'd be surprised if they're getting very many new sign-ups.

In terms of pricing, web hosting providers have done a much better job than SwapDrive in keeping up with the times. In fact, many shared hosting plans have gone waaaay overboard, offering even more storage space and bandwidth than the average dedicated server.

But as far as technology and usability, the shared hosting product has hardly changed since I first got into the industry in 1997. And guess what? The average guy on the street still doesn't know what FTP is, much less what's in his "CGI-BIN" or what the "Server Sides" include. By the way, "Perl" looks like it's spelled funny, and how do you pronounce "MySQL"? Maybe they're waiting for the Mozy of web hosting - a more advanced service that does more for you? As a point of reference, I'd pay for Mozy even if Swapdrive were free.

So I think web hosting provider should listen to Mike over at Techdirt:

One of the most interesting things a company can do is cannibalize its own offerings. Intel famously does this on a regular basis [which is why YOUR data center hardware becomes obsolete so quickly]. If you aren't willing to do that, someone else will do it for you. One possible approach is to build a separate group, whose job is effectively to act as the competition. Let them develop the next great competitive advantage - and if it destroys your existing business, better an internal team than someone else.

Other ideas I've seen include Google's 20% time (during which engineers are free to pursue new ideas) and Yahoo's Hack Day, during which teams of employees compete to build cool things.

What I regret most about my time at EV1 was how oblivious I was to the outside world. In my mind, the business was all about today's sales inquiries and customer requests. My thinking was, we've got to focus to get through the workload! The problem is, tomorrow's customers might not need what we have to offer today. Which makes your revenue base awfully transient. So set some time aside for thinking ahead - before a new competitor (or several) comes along and does that for you.

I don't know if you've been following the Google self-sponsorship controversy?

1. Earlier this month, CentralDesktop CEO Isaac Garcia pointed out that Google holds the top ad slot for "spreadsheet", "calendar", "email", "blog" and bunches of other search terms. In addition to occupying premium ad positions, Google products often dominate top search results as well. (Google responded that their product teams compete fairly and squarely with the rest of us for placement.)

2. Last week, Philipp Lenssen posted a couple of screen shots showing that Google is pimping Blogger and Google Calendar via graphical tips on top of search results.

I'll bet sooner rather than later, searches for "web hosting" will bring up ads/tips/listings for Google Domains and Google Apps. In addition, Yahoo! and Microsoft might jump on the bandwagon and start using their search engines to promote Yahoo! Web Hosting and Office Live. Even Wikipedia founder Jimmy Wales' new social search project has a free hosting twin. (I'm not saying there will be a connection between the two initiatives, but there's certainly the potential?)

That's unfair, you say? Too bad. Their search traffic is theirs to monetize.

But you spent $100,000 last month on keyword ads? Of course you're a highly valued advertiser! On the other hand, Mary Meeker from Morgan Stanley says (PDF) that within 5 years, Google and Yahoo! could generate as much as $30 worth of revenue from each user. So in the grand scheme of things, holding on to your ad dollars might be less important than building their own user base.

You're not scared because you've got better customer service? Great! Ask your tech support department for a list of customers who've never opened a ticket. If these folks - who've never come in contact with your wonderfully personable reps - took off for Google's free offers, you'd be left with a truly appreciative group of customers who'd help maintain a steady ticket volume - at a third of the recurring revenue.

The bottom line is, if your business model is dependent on SEO or PPC traffic, you're sunk. And if your product consists primarily of bandwidth and disk space, that's not enough. Because web hosting is not about selling these cheaper-by-day commodities - while hoping to grow your revenue. It's about convincing consumers and businesses to extend your platform and populate your ecosystem. If you succeed in building the kind of community that MySpace has, Google will be spending $900 million to advertise with you - not vice versa.

I've just read on Ars Technica and TechDirt that starting in January, Microsoft will flag sites with Extended Validation SSL certs with a green "safe bar".

According to draft guidelines (PDF) from the CA/Browser Forum, that green safe bar won't be easy to come by. For one, the cert issuer must verify applicants' addresses against government records. If there isn't a match, an in-person visit is required, along with photographic documentation of the business' building exterior and actual workspace.

VeriSign has started selling EV certs for $2495 (!). GeoTrust's got them too, for $899. As of yet, there's no reference to EV on Comodo's website. GoDaddy seems to be the only web host who plans to offer it, starting in "early 2007".

"Goodbye, SSL padlock", says Nate from Ars Technica. In which case, goodbye also to web hosting providers' revenue stream from SSL cert sales - and good luck to those who've invested in substantial SSL inventories. On one hand, the EV certs aren't priced for the mass market. At the same time, site owners will be reluctant to pay for a second class cert that doesn't come with the coveted green bar.

Everybody is talking about Wikiasari today. I first read about the Times of London story on Paul Kedrosky's blog. Then I noticed posts on Pulse 2.0, HipMojo and Mashable and ZDNet as well.

Wikiasari will be a human-powered search engine. Wikipedia founder Jimmy Wales is getting ready to take on Google, Yahoo and Microsoft's Windows Live Search because "the basic task of a search engine is to make a decision: this page is good, this page sucks. Computers are notoriously bad at making such judgments... but we have a really great method for doing that ourselves. It only takes a second to figure out if a page is good, so the key is to build a community of trust."

Initial reports indicated that Amazon would be a partner in the Wikiasari, but Wales has clarified that Amazon has no role beyond its recent investment in Wikia, Wikipedia's commercial arm. Amazon's involvement would make sense though. Wikiasari could leverage the uber-cheap labor pool on the Mechanical Turk (which already powers Askville, Amazon's new Q&A service) to sort search results.

BTW, remember Jimmy Wales' other recent announcement? He plans to set content free by offering free bandwidth, free storage space and free software.

Earlier today, I was talking to Jason Bates from Greenlush about SEO. Unfortunately for Greenlush (not to mention every other hosting company) Wales' new projects might end up pushing both its business model and a major source of new customers into irrelevance.

Via Google Blogoscoped: "Welcome to your new Dell PC - in partnership with Google". Dell machines now come with Google Desktop and Google Toolbar preinstalled. My first reaction was, watch this space for Google Domains, Google Apps and Google Docs. Jason Calacanis has much grander ideas:

"I predict that in 2007 Google will release an Operating System and a super discount PC. Think, a $300-400 PC that's sold at cost in order to make money off of their services. Think about being Dell or Gateway and being offered a free OS and 1-5% of the revenue from Adsense on the computer? How can you turn that down when it's going to be double or triple your margin on a PC?!"

Microsoft, by the way, is bundling Office Live with Sony and Toshiba computers. When the news came out last month, Paul Engels from Hostopia questioned the effectiveness of preinstalled bloatware. But if the bloatware were subsidizing your PC, might that change the picture?

PS - In addition to GoogleOS and GooglePC, GigaOm say Google Mobile Phone might be on its way as well.

I'm intrigued with CPUShare. It's gotten 1592 Diggs over the past 4 days. It's sort of like UC Berkeley's BOINC, which lets you donate idle time on your computer for scientific research. The only difference is, CPUShare creates an open market where you can sell your CPU time.

A couple of months ago, when I read on BOINC's website that companies can use its free software to create private grids, I wondered what would happen if web hosting customers installed it on their servers. Could they build their own utility computing platform for trading idle resources (thereby making their data centers' electricity bill go through the roof)?

With CPUShare, server owners have a lot less to worry about. All they have to do is download the software and name a price for their computing resources. No, there aren't any buyers yet, but the payment processing system won't be up and running for another few weeks still. But if I were a data center owner, I wouldn't be too happy about these calculations from a Digg user: $300/year per machine in additional power costs, plus additional heat that a fully utilized CPU generates, plus accelerated equipment wear and tear.