Kyle's life in technology

Friday, October 08, 2004

Data shufflers, value adders, and leveragers of synergies

The IT industry is about data when it comes right down to it. Collect data from various human & non-human sources, do something with it, send it somewhere, store it somewhere, display it somehow. That's why it's called information technology. Not trying to dumb it down, all of those things can be very complex, but I would challenge you to find something in IT that doesn't fit in those categories. And somewhere in that data shuffling, we in IT hope to add value. Making some business process better, faster, cheaper, or even enabling a new 'thing' that wasn't possible without IT. (Like blogging!) ;-) My problem with it is that in terms of business IT ... it's so time-consuming (man hours) and therefore expensive. Looking at price of your average business IT project -- projects that statistically are highly likely to be late, go over budget, have critical technical or usability flaws, etc.

So the challenge is making the value to cost ratio as high as possible. The two ways to make that ratio higher are to increase value and decrease costs. Let's take a look at these ...

  • Value: What makes IT valuable? Some things that come to mind: ease of use, applicability, relevance, timeliness, and in general, low friction. Think of the software and/or applications you use that really work -- often the ones you don't think about -- and see where they fall the aforementioned categories.
  • Cost: Where does all the time go for an IT project, or just IT in general? Again, the low hanging fruit: integration, user interfaces, defects in the software, defects in the process (worse than software defects), testing, business rules & logic, reporting.
In this post, I want to focus on the two highlighted terms: relevance and integration. Of course there are many areas ... but that's for another day and another post.
  • Relevance: An application/system is relevant if it allows you to easily access or create information that is highly useful to you or others, now. Meaning you don't have to click your mouse three million times or scroll through a massive list to get/do what you want. And it should do it for you quickly. For non-UI type processes, they should run intelligently without daily intervention & tell you when there is a problem.
  • Integration: Alluding back to the first paragraph, a lot of what IT does is collecting, processing, moving, and displaying information. Data comes from different sources, resides in different places, and is in different forms. The majority of the time, applications require data & processes from multiple systems that need to be integrated to produce relevance. Integration costs a lot of money.
IT is still a relatively young industry. (As is industry in general ... the assembly line came in existence only in the last century.) The 'main challenge' of IT has changed dramatically over the past decades. It used to be scale & raw power (think loads of vacuum tubes & rooms full of equipment to do some floating point math...), then gradually shifted to storage & memory (how many floppies to install Windows 3.1?), then to connectivity. Less than 10 years ago, my 28.8kbps modem was amazing. In the business world, before ethernet/networking became ubiqitous, floppies were as common as Kleenex. I just happened to notice yesterday that my latest work laptop didn't come with a swappable floppy drive. And I've been using it for weeks ... Well, today, lots of those challenges have become non-issues:
  • Today's PC processors are amazingly powerful, and clusters of cheap computers are becoming a reality that is bringing lots of power to the masses.
  • The gigabyte is a cheap thing. Terabytes anyone?
  • Most people now have 3-5mbps of Internet bandwidth in their homes. Servers can talk to each other cheaply at 1000mbps. My laptop can talk to my home PC at almost 54mbps without wires for < $150, and required surprisingly little of my networking knowledge to get working.
  • There are some data standards that mean my {Windows|Linux|Mac} computer can talk to just about any other {Windows|Linux|Mac} computer on the planet. Even though TCP/IP is frighteningly missing a few layers, none would argue it has brought the world a truly global network.
  • Storing & searching vast amounts of data is becoming easier. I can search all the public docs on the web in a blazing 0.4 seconds, use a powerful, stable database for free -- (but interestingly can't find an old doc or email on my hard drive in less than 10 minutes.)
To wind down to the point I've taken a very long time making -- it shouldn't be so hard & expensive to make relevant applications that integrate multiple sources of data. I shouldn't have to understand a complicated standards set (here, here), buy an expensive software suite (here, here), or require the effort of someone with a masters degree to make it all work. (Please don't get me wrong -- the technologies/products I referred to in this paragraph are great and I use some of them -- but I think the end game looks quite a bit simpler.) Progress is definitely being made, although I think that simplicity, not complexity, will ultimately be the solution. My belief is that we have all the ingredients (processing power, storage, communications) we need -- what's left is figuring out how to put it all together to move us to the next plane of IT -- one of a greatly increased value to cost ratio. More on this later, but I think that I am going to stop ranting for a bit ...

Thursday, October 07, 2004

Giving this blogging thing a try ...

Hi! This is my first try at blogging. Stay tuned for more ...