Jotting Number 79, by Joe Podolsky:
From: joe_podolsky@hp.com
Date: June 10, 1998
Subject: The Visible Hand
I'm writing this in June. It's also 1998, but for the sake of this essay, the month is what's important. HP operates on a November to October fiscal year. (The myth is that Bill and Dave set that odd calendar because they got a better deal from their accounting firm back in 1939.) As a result, especially for internal service providers like Information Technology, financial planning for the next fiscal year usually starts in July. June is the warm-up month when we debate accounting policies, do benchmarking on various costs, and begin to identify the projects we want to work on in the next year.
In the June 1, 1998 issue of Forbes ASAP is an article that comes from a speech that technology guru George Gilder gave to Catholic theologians at the Vatican in late 1997. The article, "The Soul of Silicon," is a defense of capitalism on moral grounds.
This article, along with the date and the issues I'm dealing with daily got me thinking about how capitalism works in HP-IT's internal economic system.
Way back in 1992, thanks mainly to an internal consulting organization called the Product Processes Organization, a new set of words entered our planning vocabulary. All of HP's internal economy began to be defined in terms of "level-1, "level-2," and "level-3" funding.
Level-1 funding comes from the enterprise-wide coffers (which, in turn, are filled from a tax on the business units). The Executive Committee controls these funds, and the amounts are negotiated with the corporate-level executives who are sponsoring enterprise-wide activities. For example, systems to support our enterprise-wide financial reporting are funded as level-1 projects. The amount is negotiated with the corporate controller, depending on the development and support efforts required. Level-1 funding is a very small part of HP-IT's total budget.
At the other extreme, level-3 funding is "pay-for-what-you-use" funding. It's mostly based on rates and works very much like a public utility. The point of level-3 funding is to give consumers maximum control over what they spend. Most IT infrastructure services operate on level-3 funding. For example, the cost of e-mail is based on things like the number and length of messages and on the storage space used. The rates are set by the IT provider, usually justified based on benchmarks with other companies or with independent service providers. Volume estimates are forecast based on past usage and future guesstimates, and capacities are committed accordingly.
In the middle is level-2 funding, designed to pay for projects. Level 2 funding is like working with an external consultant. Most development work is funded with level-2 money. An internal customer sets a specification, the project group estimates the costs, and the two haggle until they reach agreement.
This "level -1,-2,-3" funding model is a lot different than what existed when I started at HP. In those ancient days, the annual budget for IT was set once a year, based roughly on past history, new work requests, and the projected financial health of the business. Detailed IT costs were ruthlessly examined by every business manager, causing the IT manager to justify, rationalize, or beg for every dime.
Once the budget was settled, business managers competed for their shares of the fixed IT resource. The IT manager juggled priorities based on the perceived needs and political pressure. If the business health deteriorated, we were asked to cut our IT spending. If the business grew more than forecast, we sometimes got to spend a bit more than we planned, but that was a rare event.
This old financial model was very much top-down and cost-focused. The assumption was that IT was an overhead cost to be minimized. The flip side, however, was that the internal IT function had a monopoly. If a business group wanted IT services, it had to go to its internal IT provider.
The level-2 and -3 of the "level -1,-2,-3" model are, theoretically at least, market driven. The internal providers set prices for their services. The internal buyers can buy those services internally or can go outside the company. The internal providers can refuse internal business, but, in HP at least, internal providers do not offer services outside the company.
As long as level-1 is a very small percentage of the total economy, such a market-driven internal economy should offer benefits to both buyers and providers. The buyers have choices. They aren't stuck with internal providers who are offering inadequate or overpriced services. Likewise, the providers have more flexibility. Rather than being cost constrained from the top-down, they can set prices that are comparable to the outside marketplace, thus giving them money to invest in future capabilities. Also, if their products or services are in high demand, instead of being squeezed by costs set at planning time, they can refuse prospective customers, raise prices, or add capacity until the supply and demand are balanced.
Those are powerful theoretical benefits, but how well do they work in the real world?
One basic assumption of a market-driven economy is that both buyers and sellers have choices. In HP's internal economy, the choices are very real for most level-2 projects. This work can be done by internal IT people, by contractors managed by internal IT people, and/or by outside consulting firms. For most applications these days, there are capable software packages available, leading to bona fide make or buy decisions.
It's less clear, however, that there are real choices for the bulk of level-3 services, which are mostly infrastructure. Any specific service can be outsourced; e.g., computational tasks, communication, help desk, desktop support, etc. But, the beauty of a good IT infrastructure is that all the single services are integrated together seamlessly from the users' standpoint. So, if the internal IT infrastructure provider does its integration job well, the buyers are "locked-in" and discouraged from shopping around for a better vendor for any one service.
Of course, the entire infrastructure could be outsourced, but that's a major, top-down decision, not based only on internal market forces.
This "locked-in" level-3 model certainly doesn't give the buyers the benefits promised in the market-driven model, and the locked-in model is arguably worse than the old cost driven process. In the old process, the buyers had visibility into and could apply pressure on IT costs. In the locked-in model, both buyer and provider go through motions of negotiating prices, but both know that deadlocks will be settled by a higher manager, not by buying the service from another vendor.
So why bother? Why pretend that there's a market when there isn't? That's where I gained insight from the Gilder article. Gilder argues that capitalism "is not material. It is moral, intellectual, and spiritual." He feels that, in today's increasingly information-based economy, "The immaterial character of wealth transforms the modes of production. Wealth now comes not to the rulers of ... labor but to the liberators of human creativity, not to the conquerors of land but to the emancipators of the mind."
Whew! Fancy words. But what they brought to my mind was that, in a market-driven model, the buyers have more choices than merely to use the internal resources or to outsource the entire function. They can creatively change their business, making the old infrastructure obsolete, using, for example, public resources such as the Worldwide Web or by using various external Internet providers. Just using the words "market-driven" can give permission to unleash creativity needed to break the lock of a cost-driven model.
And, of course, internal vendors have similar degrees of freedom. The market model challenges us to take the lead by suggesting creative approaches and rewards us if we succeed. Cost-driven models are zero-sum games; the buyers want to squeeze IT costs so they can use the money for other investments. But, "In voluntary capitalist exchanges," Gilder says, "both participants emerge better off ... or else they would not have willingly made the exchange."
This, then, is the worthy goal of any internal funding model: to create incentives for both the buyers and providers to use creativity, not to reallocate a fixed pie, but to grow the pie in ways that make both better off. We'll never get there with a cost model. The market-driven model is clumsy and imperfect, but it's the best chance we have.
What are the internal economic funding models used in your company? What buyer and provider behaviors to they motivate? How do they inspire the creativity needed to help your organization thrive?
Regards,
Joe