So finally - after months and months of slogging, writing, re-writing, and whole lot of annoying, we get to release the latest version of Blog Flux.

There were two main things we were targeting - bringing more cohesiveness to our separate ‘parts’, and to start building a proper community. With over 150,000 registered users, we should have been doing more than we were.

One of the most time-consuming elements of this was the design. There are almost a hundred ‘unique’ pages in terms of design, structure, and appearance. All of these pages still need to keep a cohesive look to keep the user experience simple.

So what is actually new?

Blog Flux Talk - custom built message board system from the ground up.

Blog Flux Articles - name says it all.

Quizzes - always a favorite with bloggers, about time we launched our own

Still more to come (as always), including a hosted version of iBegin Share, but right now I think we are on the right track.

So I got back today - I made a little bit of a blunder and we left the airport … when the flight was leaving in an hour. So we rushed through, cut the lines everywhere, and had time to grab a coffee before we got onto the plane.

I get home and Jacob from Bloggy Network is excited. Turns out Yahoo did some feature on ‘top us jobs’ (on the front page) and linked to search results for that phrase. What came up at #2? None other than one of our sites - Life Spy.

So the site ended up making an extra $1500 (give or take a bit) that it would have otherwise.

What really stunned me was the ‘quality’ of the traffic. The front page feature/search link sent us an extra ~45,000 pageviews. But the CTR on that page was around 15% - giving us a yummy CPM of roughly $35.00! To compare, the normal CTR/CPM are less than 1/4th for each.

So there you have it. One link made us over $1500 in just 12 hours. Looking forward to Yahoo doing it again (and I can only imagine what #1 would have done).

During the middle of World Cup 2002 started, when Rivaldo was one of the hottest players, I remember reading about how he had gone from highly-criticized to much loved.

Mind you, this face-clutching episode (read Wikipedia entry) was definitely one of the worst moments in soccer-acting, but that is another point.

Rivaldo credited his resurgence to the coach. While many had called for his head, the coach had told him that he had nothing to worry - his place was guaranteed. With the stress of having to prove himself every game removed, he played at a much higher level, knowing that one bad game wouldn’t get him axed.

Of course, the opposite could also apply (player knows he will start no matter what, plays soft - eg Vince Carter), but again - another point.

So when it came to Bloggy Network and our paid bloggers, we never took on a pay-per-post model. I absolutely hate it. Successful blogs produce compelling content. They are insightful, interesting, and require some research. When a person is being paid per post, what motivation does he/she have in producing excellence? A pay per post model, imo, simply encourages people to post as often as you want. Of course there is a certain level of editorial control, but defining requirements is a non-trivial task. It isn’t fun.

The other model, a flat-fee, is working very well for us. It was stressful at first, but the fantastic growth we have been experiencing validates our model. We hired bloggers based on their ability to write interesting and compelling content, not because of their ability to churn out post after post (great for search engines perhaps, but really a shitty solution). We did of course put in certain conditions (you have), but they also meant the stress factor was a lot lower. Our writers knew they would get paid the same, be it with 5 posts in a week or with 50. They also knew that by working harder on each post, by ensuring a higher quality of work, the long-term potential and upside were far better than a post-churning blog.

It takes time. It can be highly stressful. But I liken it to the shotgun approach vs the sniper approach. The pay-per-post is shotgun-like - keep scattering fire and hope one hits. The sniper approach, far more deliberate and slow, yields much more positive results. A person with a shotgun can get to it right away, but a person with a rifle takes time to gain proficiency. Just need to be patient (and that patience is being rewarded now).

Our approach is in contrast to most other blog networks out there.

Continuing the series - it requires a fair bit of hardware to churn through the traffic we do deliver.

We started off with the traditional one server setup - providing web, sql, DNS, mail, and what not all in one box. That worked for a while, but pretty soon we were running at capacity. Our solution was relatively simple - push our non-login based services (such as PageRank Checker and Button Maker) onto a secondary server. It was a simple split, and we wouldn’t have to worry about SQL replication problems. Our actual blogs were on another server, and Blog Top Sites had its own server.

As we grew, it become evident that this solution wouldn’t pan out. By now Enthropia was growing at a rapid pace, a decision was made to invest into a reliable infrastructure. I posted about things to worry about when having your own farm. We ended up with 12 dual dualcore opteron 265s. One of them was the iBegin database server, housing 16 GB ram and 8×72 GB hard-drives.

As previously mentioned, our previous setup of half a dozen servers consisted of solitary servers. Each had to deal with DNS, email, web requests, and SQL lookups individually. This meant a lot of wasted overhead running processes that weren’t required. In our new system, one of our servers functions for both DNS and email. As we grow, adding a secondary server to handle DNS is elementary.

Each server now individually runs apache+SQL. While it may seem like a bizarre decision (we could have gone for a master-slave SQL setup with the other servers acting as web-servers), it makes things simpler for us. Because everything is internal to our own network, whenever we need to do cross-server requests (such as MapStats making sure a user is logged in), it is no problem.

We then have daily weekly and monthly backups to a backup server provided by our host (it can store upto 9TB). We also do monthly DVD backups, as a just-in-case emergency deal.

All-in-all, we utilize five servers for Bloggy Network. We are running at roughly 30% capacity, giving us a lot of room to grow.

Continuing on from Bloggy Network: the series the first site I want to talk about is Blog Flux.

As I had explained in my previous post, Blog Flux arose from the ashes of EatonWeb. What it did differently compared to all other directories was that it eschewed the traditional hierarchal structure, instead opting for a free-flow category/tag structure. Choose upto 10 categories - it is not my position to tell you what your blog is (or isn’t) about. Initially we had allowed an unlimited number of categories, but due to abuse, it was chopped down to 10.

My original plan with Blog Flux was always to be more than just a blog directory - it was to be a clearing house of anything and everything related to blogs. Having to bounce through multiple sites for simplistic services seemed a bit cumbersome. Bloggers were forced to go to X site for a button maker. And then Y site to perform pings. And then Z to see their statistics. The entire process was extremely cumbersome - remembering multiple logins was a headache.

Not everything went perfectly. We had to retire two subsections - Subscriptions and LinkLog. Subscriptions we had launched just around the time FeedBlitz. was launching. While I think our solution was quite lightweight and simplistic, FeedBlitz was honestly just the better solution. The other one that went down was LinkLog, similar to the original function of MyBlogLog. We extended the system, tracking AdSense clicks, and adding in a lot more functionality. Unfortunately, this was something that collapsed under its weight due to issues with scaling. Dealing with 100,000 clicks a day is easy. Dealing with 1 million clicks a day becomes much more complicated!

While we originally launched services in a rapid manner, what has really slowed us down now are the beforementioned scaling issues. With the size of the operations now, we need to test each system under heavy stress. For example, the Who’s Online system took about a day to write. But what wasn’t so easy was scaling the entire options - it generates over a million pageviews a day. Handling that load with speed requires more finesse than programming 101.

Overall though, Blog Flux has been a huge success. The site is simplistic, and offers pretty much whatever you want under one roof - from statistics to easy tools to blog themes, it requires one login to utilize them all. And users love it and are more than happy to link back - we have 10 active subdomains, and two defunct (Subscription + LinkLog). Every single subdomain that has PageRank has a PR of 7. Google reports 217,000 backlinks, and Yahoo clocks in at 2.9 million backlinks. The site itself generates 1.25 million pageviews a month, but we serve up over 20 million apache requests per day!

There is more coming, The integration of Commentful should be ready next week. We are working on a local blogging system that should launch in 4-6 weeks (powered by iBegin Source. Within the year I expect us to hit 2.5 million pageviews a month.

Bloggy Network (the resultant of a merger) is now roughly 15 months old. It does roughly 4.5 million pageviews a month now, and should crack 5 million pageviews next month (growth is in fact accelerating, not tapering off). All organic without a dime of investment money.

My next few posts are going to be about Bloggy Network and the (many) challenges we have faced. My first post (this one) is going to be about our history, and how we started.

Enthropia Inc has been around for a while (registered corporation for over four years now). As a web-dev company, we pretty much only develop projects for in-house usage. Occasionally we decide we no longer want to be part of a certain market, and jettison our investments (ie sites) in that niche.

I first came across Jacob Gower February 26, 2004, when we were selling one of our sites. He was quite happy with his purchase, flipping it for a very handsome profit a couple of months later.

We kept in touch, with Jacob buying most of the sites we sold (including one of the few sites I regret selling - Webmaster.org). I remember him telling me about having to sell Webmaster.org as he was committing himself to blogs. As we kept on conversing, he kept telling me how fantastic blogs were. Around this time we were still involved with Evo-Dev, and we decided on a joint venture: Blog Top Sites. We had a lot of experience in running topsites, and Jacob was involved in the blogosphere. The site was an instant success, signing up hundreds and hundreds of blogs in the first week. Today the site tracks over 4.5 million pageviews a day across over 40,000 blogs. The site generates over 30,000 pageviews a day now.

As things changed and aligned themselves in the right direction, we ended up purchasing EatonWeb, the oldest blog directory out there (side fact: Brigette Eaton was one of the first people ever fired for blogging, somewhere in 1999 I believe).

The site was popular, but it was horribly coded (PHP3 in fact). I also took a look on the blogosphere - it was extremely fragmented. We did what we do best - launch a new site, accumulating many of these fragmented services into one place. Only this time it wasn’t from scratch - we had Eatonweb to help push it forward Our original launch was nothing more than a blog directory, button maker, and pagerank checker.

Success was at hand. By eschewing the traditional hierarchal structure in web-directories and choosing an open multi-category/tag format, we were doing directories ‘different’. People liked having multiple services under one roof, and we expanded our offerings.

By now Jacob had designed his network ‘Bloggy Network’ (a name suggested by me). He had a few good blogs under his belt, including Forever Geek and Blogging Pro. We had 50% of the largest blog topsite (which spawned dozens of duplicates) and we had the best blog services site.

Enthropia has roughly 6 divisions. Each has its own manager. So when thinking about our future plan into blogs, there was only one logical person: Mr. Jacob.

And the pieces just fit.

Next up: Blog Flux, and why it clicks.