Dear Bob ... We are finally getting permission to stop being seen as a "cash drain" on the company, and start doing "chargebacks" to various departments as they come to us with systems they want built and hosted. We also host and maintain 24x7 support for a subsidiary of ours. I have been working with building SLAs for these new systems, as well as putting together a spreadsheet for the cost for hosting sy Dear Bob …We are finally getting permission to stop being seen as a “cash drain” on the company, and start doing “chargebacks” to various departments as they come to us with systems they want built and hosted. We also host and maintain 24×7 support for a subsidiary of ours.I have been working with building SLAs for these new systems, as well as putting together a spreadsheet for the cost for hosting systems (per U of rack space, half rack, full rack, etc.). One of the items I am having trouble determining is how to manage and report the amount of external bandwidth each system is using. Should I look at charging per user that will be accessing the system, or is there a software package out there that will allow me to monitor bandwidth and report high, low, and average usage over a specified period of time? Any other thoughts, while you’re at it?– BillingDear Billing … Here’s my take, for what it’s worth: This is why, in general, I’m not in love with chargeback systems.Your specific question is about how to charge for bandwidth. Just about anything you do here will be controversial. I confess that I’ve lost track of this technology, so what follows is an expectation rather than a fact: I doubt you’re going to find a package that can parse out which packets, from which systems on your network, have internal vs external destinations. Yet that’s precisely what you’ll need, because otherwise you’re charging for internal bandwidth, which is essentially all fixed cost. So I think this is a cost you’ll have to allocate on a per capita basis, except for externally hosted applications where you can monitor the bandwidth the application uses and allocate it to the organizations that use it.Speaking of fixed vs variable costs, separating these is the most important aspect of any successful chargeback system. Treat fixed costs as overhead, charged per capita. If you don’t, you’ll overcharge for use of IT resources, which might make the organization more frugal, but will discourage use of technology when it actually makes good business sense. Charge for variable costs (such as the incremental cost per transaction for your core systems, cost per megabyte of network storage space, time required for discretionary help desk calls, and so on) on a unit basis. If you don’t, you’ll encourage abuse of the resource.Then, remember to budget a headcount or two to resolve pricing and billing disputes. You’re now invoicing for your services, and department heads love to argue about the accuracy of these invoices. They’ll also go shopping, and they won’t care whether the prices they get outside are for services that are truly comparable. They’ll simply beat you up every time they visit Best Buy and find a laptop computer that costs less than what you charge them.Then there’s the biggest challenge in creating a useful IT chargeback system: Striking the right balance between pricing that’s sufficiently accurate that it drives responsible behavior, and one that’s so excessively accurate that it simply creates full employment for accountants. Are you sure you want to go through with this? There are alternatives for helping the business understand the cost and value drivers of information technology.– Bob ——– Technology Industry