Web hosting is a (still) highly fragmented marketplace.

Web hosting is fragmented market because it is a commodity product and there are low barriers to entry. I don’t see google overcoming either of these issues. They built their search brand based on having a superior product. I am not sure that google can offer a truly superior product in the hosting space.

with the millions in extra cash that they have lying around, Google can easily acquire even the most premier hosting company and use it as a base to build a business around. Alternatively, Google could simply buy up tons of customers from smaller hosts.

The dot bomb strategy didn’t work in 1999, why would it work now?

The barriers to entry are extremely low

Small companies enter markets where the barriers to entry are low. Large companies try to create barriers to entry.

Arguably, the single most important element in gaining new customers is online advertising

I am sure that time warner makes a bunch of money on advertising for snake oil. that doesn’t mean that they should sell the snake oil.

Web hosting requires tons of servers, data centers, and skilled technical support staff.

I think these are hygienic factors, not necessarily competitive advantages. Apple has lawyers and GM has lawyers, therefore Apple should make cars? Google’s technological innovations have been in using very cheap redundant hardware. I am not sure that this transfers to commodity paid hosting.

Email and Web hosting go hand in hand.

Google is quite new to large scale email. Also, if you paid for the email, why would you accept the advertising? I don’t see the synergy.

Many businesses and consumers will only do business with a brand that they trust.

There are lots of trustworthy brands already in the hosting market.

A google browser makes sense — one more way to show google ads and block competitors ads. I don’t see hosting as a good fit for google. That doesn’t mean they won’t do it. I just don’t think they would do it until they lacked better things to do, things that don’t drift as far from their core business: selling advertising.

The summary being that software that values simplicity over everything else propagates faster.

Mike Cogan doesn’t understand why LAMP and especially “uncomfortable” PHP should be more popular than “nice neat” Java on various freelance sites. He thinks that the users of these sites, typically smaller businesses, are being short sighted in their choices. perhaps not.

One thing that PHP has is a market of excellent packaged end-user oriented software for websites: vBulletin, oscommerce, x-cart, pMachine, ez-Publish, gallery, phpBB and php-nuke, plus other hosting control panel favorites. These are the killer apps that drive PHP. End users ask for them by name and look for places that host them. End users also install the free open source PHP packages available in their hosting control panels. They pick cheap LAMP hosting, then later ask their host or go on the freelance sites looking for custom programming or installation.

The Hotscripts.com Java ad management software category has 4 entries, one sponsored costing $699. Meanwhile, the PHP Ad management software category has 105 entries, three sponsored all less than $100. If you were a small site owner looking to choose a platform for your site, which do you think it would be short sighted to select?

If you were a small software developer creating a web based software package for sale at a price point less than $250, which platform would you pick to maximize your potential market?

Cogan suggests that selecting PHP would mean that it would be “difficult to integrate any other future business solutions.” But this isn’t true if your future business solutions are also PHP based, as is likely to be the case for small sites.

There is a network effect that drives end user demand for PHP. Each new killer app available on PHP makes the whole platform more attractive, which draws in new apps. The whole cycle will draw in more and more PHP programmers.

Cogan also suggests that PHP solutions would be unmaintainable, unscalable, and unprofessional. I disagree. I think these qualities have more to do with the programmer than with the language.

I think Cogan’s attitude may be typical of the MIT Approach that seems to prevail in the java culture. Zend should take note of this. The road into the enterprise is an uphill battle, fighting against the MIT attitude all the way. Instead, Zend should do everything they can to grow the commercial PHP small business/consumer web application market.

This means installer technology and including a compiled byte code distribution format in standard PHP so that companies can distribute and sell PHP applications without source code. Sure, they would commoditize the market for their cache and encoder products, but they would lock up the small business/consumer market, which is where the benefits of PHP lie. An increased commercial market would create new opportunities. Eventually, the economics will pull in larger and larger businesses.

The doorway into the enterprise for PHP is economic, not technical. Pushing PHP into the enterprise through the IT department is a losing game. Pulling PHP into the enterprise through the end user is a wining game.

Anyway, I think the network effect has already “tipped” for PHP in the consumer/small business marketplace with or without Zend.

PHP IS a disruptive technology.

One last thing, Bill de Hora has a technical reason why lamp hosting is more cheap and plentiful than java hosting.