American Capital just plunked 160 million into GeoSign. This wasn’t a purchase - it was an investment.

GeoSign is one of my favorite companies on the internet. I ran across them roughly two years, but really got to know what they were upto about a year ago. They do domains (owning fantastic domains like GolfCourses.com and Hockey.com), they do content (a ton of writers/editors), and they do local search (TrueLocal). Take my company, giantize it by a factor of roughly 15, and you have GeoSign

As a follow up Frank has an interview with Tim Nye, GeoSign founder.

This likely won’t make the news on sites like TechCrunch or Read/Write, but this is a significant investment. GeoSign is a strong company, claiming over 35 million unique visitors a month. And since I highly doubt they gave up more than 50%, we can say that this investment values the company at over $300,000,000. In seven years, Tim has taken the company from nothing to such a valuation. Kudos to that success.

I can only guess what they are going to do with that money.

FTP. MySQL. HTTP. LOCALHOST. The internet is confusing. For all the talk about how easy WordPress (and others) are easy to install, that only applies to people that know nonsensical words such as FTP and MySQL.

Back when we were still affiliated with Evo-Dev (a disaster story is for another day), we realized that installing web software isn’t a simple process (wth is CHMOD?). So we offered free installation. But installation turn-around wasn’t instantaneous - the customer still had to wait on us). So we created (this is two years ago) an AJAX Web Installer.

The steps:

1. User is told that he/she must supply us with FTP server address, login information, and MySQL server address, login information. Ask your host for that (there is a big difference between getting FTP information and actually having to FTP files)
2. User was then asked to select what product (evoTopsites, evoTopsites Pro, evoArticles)
3. If a paid product, user’s login was cross-referenced with paid product. If no info, asked that user login first.
4. We asked for necessary FTP and MySQL information.
5. We did an FTP connection. We then asked user what folder he wanted to install to, and gave examples (eg /articles/, /topsites/)
6. Files were uploaded (complete with a status indicator). Once files were uploaded, other necessary operations were performed (chmodding files, etc).
7. User was told that the software was successfully uploaded to his/her server, and would you kindly click to complete installation.

The user was then forwarded to the setup script, where the installation was completed.

Unfortunately, this never saw the light of day. It wasn’t cheap to develop, but it was awesome (worked on IE, FF, Opera, and Safari). It was quite fast (we had bought bandwidth directly from Internap), was straightforward, and our beta testers loved it. Errors were dealt with beautifully (incorrect FTP information, incorrect MySQL information, cannot chmod files, etc).

And yet I see nothing like this for any popular script out there. I guess I should get it re-created (that code went when Evo-Dev split off from us).

UPDATE: March 9 - People seem to think I’m just jumping on the AJAX bandwagon, when I’ve written about keeping AJAX in perspective (over a year ago) and about how web 2.0 is mostly crap (almost a year ago).

But in this application, AJAX makes sense.

To think it through, here are some steps:
1. User selects product
2. User inputs information (FTP, MySQL)
3. User either lets it auto-install or chooses a directory
4. Files are uploaded and configured.
5. User is forwarded to the installation script.

Think of #2, #3, or #4 in a non-AJAX manner. A horrible user-interface. I put in the information, and you make me reload the page. Then I want to choose a directory - are you going to reload the page every time? (We had a dynamic folder viewer which let you browse like a normal FTP client). When uploading files, is the user just going to wait around while you upload 10 megs?

No no and no.

User puts in FTP information, you want to show them (using AJAX) the connection in progress (resolving DNS, connecting, putting in password, etc)
User wants to select a directory, you let them browse dynamically.
User waits for files to upload, you want to dynamically notify them. Eg: “5 of 232 files uploaded (0.42/24.1 megs). ETA left: 5:30 minutes”

The AJAX-version would be far superior than the non-AJAX version.

Every day, I get more and more scared of Google. Not in the sense that they are going to suddenly remove me from their rankings, but by how much they shape the internet experience.

A lot of people still think the Internet Explorer ‘e’ icon is the internet. A wizened reader (like you) knows that is absolutely ridiculous. But scoffing at behavior exhibited by millions is a foolhardy exercise (just ask all the domainers making millions because a ton of people insist on writing out domains).

Using the Internet Explorer example, to many the internet is Google. While IE ’starts’ the web, Google ‘is’ the web. The entire internet experience consists of going to familiar sites (ala cnn.com, espn.com, etc). For many the place where their actual internet experience starts is Google. Very strong brands like CNN and ESPN can stand on their own - others must be found through the internet.

So it is with that thought that start pages have the power to be so painful. Any time someone becomes an active user of a website, they essentially invest (their time) into that site. The more a person uses that site, the less likely they are to leave. One can argue that there are superior alternatives for social bookmarking, but the undisputed king is Delicious. Sure exporting bookmarks is easy, but who wants to go through that burden?

So going back to start pages - when a person starts using a startpage (such as NetVibes), and customizes it, and becomes more comfortable with what it does (and doesn’t do), the inertia created is massive. I could pull a superior homepage (faster loading, easier on the eyes, new features super easy to switch), but the amount of people switching over would be minimal. This trend of course applies to anything software - I still use WS_FTP95 for my FTP needs. It bases everything off of C:\ (the ‘root’ is not the desktop), but hell - it works for me, and I cannot be bothered with the newer FTP programs.

So the ‘winner’ of the startpage race will have a lot of weight to throw around. Lets say tomorrow the race is over, and NetVibes is the winner, with 50 million users. These are users that have set NetVibes as their homepage. 50 million people whose daily ‘introduction’ to the web is this one website. Instead of seeing ESPN, they see Yahoo! Sports. Instead of seeing weather by AccuWeather, they see weather by Weather.com. Just imagine if Yahoo! was the default search instead of Google - the revenue from the resulting searches would be a damn lot. Enough to pay NetVibes to make it worthwhile. After all, blind tests show Google, Yahoo!, and MSN have relatively equal searches - most people wouldn’t even care much.

But (and you know this was coming) … every single startpage sucks. NetVibes, Page Flakes, Protopage, Google’s, Yahoo’s - they all are complete crap. I’ll explain why … tomorrow