How should infrastructure be funded?

In any function -- IT, Facilities, HR, Engineering, etc. -- "infrastructure" means the assets that internal service providers own and operate in order to provide services to internal clients. In IT, for example, that includes networks, servers, middleware, shared-use tools like email, and even applications when IT owns the application and runs it as an application-service provider (ASP).

Where's the money to buy these assets supposed to come from?

To shed light on this question, consider how you'd feel if the phone company charged you for wiring your house, and then claimed to own the in-premise wiring and charged you a service fee for its use?

"Not fair," you say!

But this is exactly what some IT departments are doing. They're demanding funding from clients for wiring new buildings, for augmenting the wiring in existing buildings, and for installing wireless access. Then IT claims to own the entire network and charges clients for the connectivity service.

Naturally, clients are upset. They feel they're getting double-charged, even if the calculations can show that's not true.

Beyond this, clients demand control over their spending. We live in a market economy where your money is yours to spend. Nobody tells you that you have to buy something you don't want.

Thus, if clients are asked to pay for any portion of your infrastructure, it would be natural to assume that they have a right to decide what they buy. This leads to controversy when you want to change the infrastructure and clients say no, or when clients want to use a different technology than what you propose.

If you force clients to buy things they don't want -- under the guise of "policy" or "technical imperatives" -- then clients feel out of control of their own money and resentment is inevitable.

Furthermore, we respect private property. We're accustomed to a simple rule: If you paid for it, you own it and have every right to control it. When clients fund infrastructure, clients have every right to feel they "own" the assets they paid for. But this undermines your ability to manage an enterprise resource.

And when it comes to the above example of building wiring, there's even more confusion. Business units are generally tenants in buildings owned by a corporate facilities department. If clients pay for the wiring and then move, does the next tenant (if any) have to reimburse them? Or does the facilities group pay for and own the wiring, in which case it gets to decide what kind of network to put in?

By the way, the network isn't the only area in which this kind of confusion occurs. Many IT departments charge the cost of a workstation to clients, and then claim to own them and retain the right to control their use and replace them when the IT department decides it needs to.

Another example is applications servers. If they're charged to clients as part of an application implementation project, clients have every right to feel they own them. This becomes an obstacle to enterprise capacity management and server consolidation or virtualization.

The Crux of the Problem

To be fair, consider the challenge from the viewpoint of the internal service provider. It needs funding for the one-time costs of installing infrastructure. Then, it needs funding for the ongoing operational costs of that infrastructure.

Whether the internal service provider charges clients directly{WHITEimeCHARGEBACKS#3>, allocates its costs, or has its own budget (funding that clients believe is provided to benefit them), clients ultimately pay.

Don't think that you can ignore the issue for long just because you don't charge back your costs to clients. If clients look at where your budget went, they'll see that it went to buy infrastructure instead of the projects they requested. Psychologically, it's all their money.

Sources of Funding for Infrastructure

Given that you can't have clients pay for infrastructure and then claim that you own it, what are your alternatives?

One alternative is to get funding from clients and then grant that they own what they paid for. But this, too, is unacceptable. If you lose control of your assets, you can't ensure that services are delivered securely and reliably; and you can't manage enterprise capacity.

The right approach is to not ask clients directly for funding at all. There are two other sources that can be tapped.

First, for small or ongoing infrastructure investments, a fund can be created by building a small margin into the rates charged for infrastructure-based services.

Even if you don't chargeback, you need to define exactly what projects and services can be expected for a given level of budget. That involves pricing your products and services. In this calculation, each deliverable must bear its fair share of indirect costs, including the ongoing cost of the wiring and wireless.

Second, for large investments, you need to designate in your budget a project that's funded by the enterprise, like a loan from the bank. I call this "venture funding." This asset may then be depreciated, with the depreciation expense built into the rates charged for infrastructure-based services.

For those internal service providers which don't charge back, this is portrayed as a distinct line-item in the budget, reducing the number of client deliverables that the rest of the budget can fund.

For those internal service providers which are expected to recover all costs from clients through chargebacks or allocations, this is still portrayed as a distinct line-item in the budget. In other words, it's excluded from the concept of full-cost recovery (though subsequent depreciation is recovered through rates and returned to the enterprise). It's as if the enterprise is serving as a "bank" that loans you money, and is repaid through mortgage payments.

My Bank Won't Loan

There are some obstinate CFOs who don't understand the role of "bank" and the need for venture funding, and continue to (inappropriately) insist on full-cost recovery. If logic, politics, and common sense do not prevail, what can you do?

It's still critically important for clients to understand that they are not paying for your infrastructure. But clients can be asked to provide funding -- as long as it's portrayed as a loan rather than a purchase.

In one IT department, they had no choice but to get the funding for infrastructure from clients. But they worked out a clever arrangement for doing so. They explained to clients that the money provided by the business units was a loan that would be paid back over then next three years in the form of discounts on the services delivered using that infrastructure.

In other words, the depreciation (loan payments) was deducted from the total cost of service to give the clients a discounted rate. Once the loan is repaid, the rates go back up to the fully burdened cost.

This served two purposes. It clarified ownership of infrastructure. It also allowed the publication of rates that are sustainable and applicable to other clients.

Is This Just Bureaucracy?

All this may sound terribly complex and bureaucratic, especially to technical staff who may not feel clients' wrath the way C-level executives do. But the fact is, every internal service provider needs to know the full cost of its products and services in any case. That's fundamental to any rational resource-governance processes, and to managing clients' expectations.

The key is to calculate rates (whether charged to clients or to the core budget) that include a fair share of indirect costs. Indirect costs include the ongoing investments (like infrastructure maintenance) and the depreciation of one-time, big investments (ventures).

Whether you charge back or not, you've got to calculate rates. This gives you a chance to fund needed investments without turning over to clients control of your infrastructure.

Abstracts Free library

Books Speech abstracts

NDMA coaching/consulting services

UP.... NEXT PAGE....