Remember when the media predicted part shortages across the PC industry? Well, that prediction turned out to be wrong. How wrong? DIGITIMES.com is now reporting caution of slow demand in many components:
In addition, raw material costs are on the rise as people question the strength of the US dollar.
In summary, there is a lot of uncertainty in the market, and there will continue to be for a while. Even the best forecasters are still guessing; if they could tell the future they would be retired and living off their stock-market fortunes. Technology moves quickly, and so does the market, so plan to be agile or pay the price.
Where should you place your bets?
This brings me to the opposite of agility, stability. Many integrators claim to offer extremely long life products (10 or more years), because sometimes stability matters more to an equipment manufacturer than cutting-edge technology. Yet the suppliers of PC components don’t offer ten year lifecycles on their products . . . so how can integrators offer these types of lifecycles?
There is only one way: buying supply upfront and putting that supply on a shelf.
How does the integrator know how much supply to buy?
Their customer forecasts (guesses) and signs a check.
To be clear, there is nothing wrong with this as long as you acknowledge the risks, and Dell’s OEM team even does it with our own customers if there is absolutely no other way to address the customer’s lifecycle requirements. But when you take this approach, only one thing is certain – the guess is wrong. By how much, no one can say. And who pays the price for the error in the forecast? The end customer does, because it is (or should be) baked into the manufacturer’s costs or the costs of the integrator.
If that wasn’t bad enough, the customer has now purchased ten years of rapidly depreciating inventory. Ten years ago a Pentium based computer with a 2GB hard disk and 512MB of RAM cost $3,000.
Where will you be in ten years?
In isolation, it might make sense to commit your capital to buying (hopefully) ten years of outdated technology to avoid re-qualifying your software every couple of years, or even avoid a complete software redesign.
However, this forecast fails to take into account what your existing competitors are planning – could they be moving to a commodity hardware platform that will drastically reduce their material and support costs? It also fails to account for new competitors that might enter the market with a full blown SaaS solution. Sure SaaS may not be right for your environment now, but 10 years is a long time in the technology industry.
Amazon only went online 15 years ago, and that was selling books. Now they sell IT infrastructures.
Get on the train
Going forward, my colleague Franklin Flint and I will be discussing the various costs associated with going custom, and how to plan your next product releases so that you can get on the technology train. Unless you have to, it makes sense to swim with the currents of the IT industry rather than against them.