Take for example a simple e-commerce site that has one provider hosting its DNS services, a second provider hosting the site content and product catalog, a third provider handling the checkout and payment process, and a fourth serving up ads. When suddenly the Website slows down or a process stops working, which of these providers is at fault? Perhaps the problem is related to an unexpected surge in traffic, maybe it's the fault of the Web hosting company that isn't meeting its service-level agreements, or there could be a capacity limit that does not allow for spikes in demand.
The Supply-Chain Process
It doesn't matter who is to blame. The CIO must have a system, processes and the right team to stay on top of pending issues and investigate and resolve them swiftly before they affect customers. The ad hoc monitoring systems and tools of the past, when everything was managed internally, won't likely do in today's world. Now, with so many moving parts outside company walls, CIOs need to set strict parameters for service. This entails defining capacity and service levels, regular monitoring, and strong lines of communications with suppliers. If the supplier changes its infrastructure or its security policies, and you don't know about it, that's a problem. CIOs have to understand the inner workings of key suppliers at the same level as if they were managing the technology internally. Without that transparency, it will be harder and more expensive to troubleshoot and resolve performance or quality problems later.
A basic framework for managing services in your supply-chain could look like this:
1. Define services a.k.a. compute capacity, storage, network bandwidth, network infrastructure, content delivery network, DNS services, security, application platform, back-office system for orders, front-office apps, affiliate networks.
2. Track each of those services against a core set of metrics, such as performance, responsiveness, capacity, security, cost per unit and volume cost. An individual service, such as network could require additional sub-metrics, such as spike capacity and overage fees.
3. Score each supplier/partner and deliver them a monthly report that you can then use as a discussion tool and a negotiating platform for discounted fees or add-on services.
But wait. Isn't this supply chain model just a new name to good old outsourcing that IBM, Accenture, and others have been doing all these years? No: this is a new game. The old IT outsourcing school relied on massive financial engineering deals that focused on capping operational costs for customers, while the outsourcers ran operations using, guess what, their own equipment and consulting services and placed the customer's IT staff on their own payroll. The outsourcer and customer both had to handle a lot of infrastructure building, system integration and portfolio management. But this world is over. Thanks to cloud platforms and standard Web services, the need for the middleman is diminishing. Large deployments that used to take years, millions of dollars, and hundreds of people can be done today in weeks, thousands of dollars, and a few data analysts.
Sign up for Computerworld eNewsletters.