Harold Geneen, the legendary chief executive officer of IT&T, also lived by the numbers. He wrote, "The professional's grasp of the numbers is a measure of the control he has over the events that the figures represent." But like the numbers he revered, Geneen's observation contains only some of the truth. Control does depend upon mastery, but also upon how completely and how accurately the figures mastered actually represent a very complicated reality. What people usually forget or ignore when they look at numbers is how imperfect this representation may be. Precision can mask misunderstanding. Numbers frequently don't measure what we think they do, or not all of it, or not accurately, or not in terms that are meaningful.
Accounting systems, which depend for their utility upon the similarities among companies in different businesses, paper over a world of differences that are crucial to the people running a company. The conventions of financial accounting often have no place in the real-time tracking of the day-to-day operation of a company -- or so Larry Stifler believes.
Nonetheless, practically everything Stifler's company does is captured by a number somewhere. So, what's the difference?
For one thing, Stifler's way of using numbers to represent relationships tells him things about how his business works that are not otherwise obvious.
For instance, HMR does very little consumer advertising. But it's offering a consumer-oriented service. Wouldn't it want to advertise? Stifler knows he doesn't have to, and he understands why. He figured it out by looking at a few numbers and thinking about them in terms of their relation to one another. He saw, from records kept, that on average every HMR patient who moves from the fasting stage of the VLCD into the 18-month maintenance stage recommends 2.2 additional patients who enter the program. If that's true, he said to himself, all I have to do is get at least 46% of the patients in the program into maintenance, and the program will self-sustain (since 46% times 2.2 equals 100%). If more than 46% make it into maintenance, the program will grow automatically. "So," says Stifler, "I said, set up the business for quality care and it'll grow by itself."
Looked at another way, the numbers tell him that every dollar spent providing good patient care comes back 2.2 times.
Or looked at still another way, they tell him that reducing the quality of patient care to save $1 actually costs the company more than twice that in revenue. It's a powerful tool to use with his hospital and medical establishment clients whose administrators tend to believe that you improve profits (or reduce losses) by cutting expenses. "A big part of our job," says operations director Povall, "is convincing hospitals not to cut corners."
In the 26 clinics that HMR operates itself, according to Stifler, between 60% and 70% of the patients who enter stick with the plan for the six months that it takes, on average, to get through fasting and into the maintenance phase. He believes that is the highest figure in an industry in which, more typically, half the patients drop out in less than three months.
When Stifler looks at a company he sees a set of relationships among people, things, and events. It's impossible, he says, to understand what happens inside that company simply by looking at what something cost or what revenue it brings in. Compared with what? he would always be asking himself.
Instead, Stifler's tendency is to build mathematical models that attempt to capture important relationships within the business. In the context of these relationships, numbers acquire meaning. Traditional budget line items expressed in dollar amounts don't give you that.
An annual budget will have a line for projected income and lines for anticipated expenses. Let's say there are three: salaries, telephones, and office furniture. Obviously, these three and the income line are related. If revenues rise more than projected, you'll probably have to hire more people, install more phones, and buy more office furniture. But a budget doesn't tell you anything about how these four items are related to one another. Changing one line item doesn't automatically adjust the others.
Here's the same problem seen in another way. You get your department heads' budget proposals for the coming year. Telephone expenses come to $100,000. Last year you spent just $70,000. You go through the roof: too much on phones, you tell them. But is it? So you assign someone to do a study of telephone costs, and three months later that person hands you the results. But then what about office furniture?
Here's how Stifler approaches the situation -- without a budget. The determining variable in his company is the number of patients enrolled in the weight-loss program. (In another company it might be the number of hamburgers sold or clients serviced.) Everything at HMR is related in some way to the number of patients enrolled. Stifler can begin by estimating a reasonable patient-staff ratio. Say, for instance, it's 50:1. And he knows what each employee costs in compensation and benefits. So, there is a direct and easily quantifiable relationship between the number of patients enrolled and the company's payroll costs. Each staffer needs a phone, desk, and chair. So phone costs and furniture costs also relate directly to the number of patients enrolled. Now, he builds a model, which is fairly easy to do, incorporating those relationships. For every 50 patients, he'll need a new staffer, a phone, a desk, and a chair. If they have this model to follow, clinic managers don't have to spend their time building budgets based on estimates and then revising them when actual patient enrollments are higher or lower than estimated. They just keep track of patients enrolled and do what the model tells them. If they get 50 additional patients they hire one new staffer; 100 and they hire two. And they don't have to find the money to pay for the staffers -- or to install the phones and furniture -- because each new patient brings in a known addition to revenues. "I never write a budget or proposals," testifies Dolores Gable, a behavioral health educator at HMR's Newton-Wellesley (Mass.) Hospital clinic. "I just work with patients, and in a year of working here there's nothing we've wanted that we couldn't get. The model tells us."
