Web Migrations & Taking Sites Live Reality Check

Now let’s say you have been brave enough to either hire a web developer or build your own website yourself, or let’s say you have been assigned to build out a new website for a large corporation. If you know some PHP/Mysql or have some programming skills, or you are a designer, a “web producer” or web product manager or just a plain old entrepreneur, and you are in the middle of trying to get your website live, I understand your pain.

The Brand New Site

Websites run the gamut. They start from a 1-10 page biz card like sites, which only show your basic contact us, about us, services, to a full blown combination of existing systems like WordPress & Joomla adding in customized “serious” app development. They can have 10 lines of programming or in the case of one of my projects over 300,000 lines of programming. Either way, a brand new spanking web site with some level of serious programming will have this kind of logarithmic ending to the project It’s even worse than the old 80/20 rule, where 80% of the work is in the last 20%. It’s more like 95% of the work occurs in the last 5%.

Why so much work at the end? That’s because you typically have a situation where a lot of things are not known till the very end. It does not matter what the developer, project manager, third party guy in India tells you. The hard work in this business starts not on day 1, but day 180, when the petal hits the metal. And this kind of work has more to do with QA than development, and precision, not hand grenade throwing let’s kind of get it working. That’s why many outsourced websites, to overseas folks, die on the vine, or cost 10 times what they projected. This detail work is the work that you, the owner, or a close person to you needs to do. It is not for a guy or gal in a developing nation out there to do. Not to say that overseas development is not cost effective, it’s saying that I have my doubts after the 7th inning stretch.

Civil Engineering vs. Web Engineering

If you compare building a website to building a building. They are similar in that there should be some type of project plan. Where they differ is that a building can’t change, much, once you start building it. The plans for a building are set in stone, or the building could fail. A website is more like a big plumbing project in an old house, even for new websites. You don’t know the full extent of the project, sometimes, till you are in the middle of it. That’s why many web developers are not so willing to take one price stop shopping when selling their skills. Smarter web developers now realize that the big work can emerge towards the end, when a few extra things were discovered.

Web Feature Discovery

I mention web feature discovery a lot in my blogs. That’s because much of the development process on the web is about discovery. A good case in point is you start building a site, and you end up finding a new opportunity along the way. An example recently for me is we were building this college admissions counseling website and we realized about halfway through the project that there is an opportunity to create a series of pages, when SEO’d, would drive thousands of visitors. We would not have thought of this feature, unless we undertook the project. This is not a sequential process. This is a mind map type of process, where you start and many different directions appear. You have to visualize these directions, and rate them and decide which come next, which to ignore and which to take on.

The Migration

Ah, the migration. Another way to say this is what a p in the a. When you are moving a website, it is one of those things that can keep you up every night and you are blind while you are doing it. Even when I had a team of 20 people working for me and with me on a big migration of sites for NTT corp., we could not think of all the details. Our brains can not contain everything. When I recently migrated Pre-dating.com, one of my projects, I did it myself. This was like saying I am going to move myself, and you have a big house of stuff. It will happen, but it is painful. I have moved something like 12 times in my adult life and everytime I have to leave something behind and extricate from my life items and things I don’t want to get rid of. But moving means leaving it.

When migrating, you have this list of all the things that need to happen. Mind you, some techies are great at migrations. But no matter what you know and do, there always seems to be an issue you did not think of. We don’t know everything or sometimes we don’t know much at all. I moved a site recently and I realized after we moved the reverse DNS was set up incorrectly, so when you looked up the site by IP address it was incorrect. AOL blocked the email. Argh! I got it working, but it was in an area that is not my expertise. I fixed it, like fixing a migraine headache. Another issue on a migration I see all the time is slight differences in the servers. The old server and new server may be the same, but for some reason PHP, mysql, sql server, there is always a difference. Hopefully the settings don’t cause a major problem, but it often does. I have even seen a migration and 4 months later the problem is discovered.

What Am I Saying

I am just venting on the issues in taking sites live, and I can’t give you much than a pat on the back when it goes live, whether its a new site or a migration. This is an accomplishment, regardless of what you techie friends would say. There are many sites which refuse to change, move, migrate or improve because of fears of disaster. The disasters, I have seen them, so there are a good possibility no matter what you do. They happen and you deal with it!