That's how Stifler and others at HMR use numbers to help them understand how the company works. They also use numbers -- ratios, once again -- to keep track of how well it's working. It doesn't take many. Stifler begins by looking at just two.
The first is productivity, and it's simple. He divides the month's net revenues by the number of people in the company (using FTEs, or full-time equivalents) and gets a ratio -- revenues per employee. If all else fails, he says, "this will tell me that there's something wrong. If productivity is the same or better from month to month, then we're OK." Last November HMR's monthly productivity figure hit a new high at something more than $140,000 per employee.
It's a quick check, a safety. If there's something wrong, the productivity figure doesn't identify it. Also, an improvement somewhere might offset a problem somewhere else, and Stifler's productivity figure won't catch that. But it's easy and current, and unlike the data that accountants work up, it gives him an instant view of the forest instead of a belated head count of the trees. "With this one number, I don't have to look at any other."
But he does look at a second number, a number that takes him down one level of abstraction, one step closer to what's really happening in the company. It tells him at a glance how two parts of his business -- product and service -- are doing relative to each other. HMR makes its money from selling liquid-diet supplement. It charges patients for the services it provides them, too -- the support group sessions and counseling. But the service, including the training provided to clinicians in hospital-run clinics, is basically a cost incurred in order to sell the product -- rather like customer support and field service are costs incurred to sell computers or other manufactured goods. You have to offer the service, but you don't want the cost to get out of line. So Stifler maintains two checking accounts. All the income from products sales goes into one, from service fees of company-managed programs into the other. Expenses, likewise, come out of the appropriate account. Because the company makes money on product sales and loses money on service, every month Stifler has to write a check from the product account to cover the service deficit. It is not the size of the check he worries about. It can grow larger so long as it stands in the same proportion to revenues as last month's check. If it's not, right away he knows that product and service have gotten out of whack. His accountants, he says, think it's a nutty system. But their reports, he complains, take too long. "It takes them months to find out that I've got a problem, but my system tells me immediately, not only that I have a problem but exactly how bad the problem is."
Furthermore, Stifler can figure out where the problem is by dropping down one more level of abstraction to find which cost has gotten out of line. Once again, he won't be checking dollar amounts but a set of ratios.
Not every month, but when he is so inclined or when either his productivity calculation or the check he writes to cover the deficit in the service budget tips him off to a problem, he calculates about 18 ratios. (He could have someone else do the calculation, he says, but doing it himself makes him pay closer attention to the result.) The denominator in every case is net revenue. He looks at 18 or 19 cost categories as a percentage of net revenue. His model tells him what the percentage ought to be; the calculation tells him what it is. "This may sound deceptively simple," says Stifler, "but nobody has to shuffle 150 pieces of paper after four months to see what went wrong. I can do it at the end of the month in three minutes."
The point, again, is that looking at the number of dollars spent in any cost category -- telephones, for instance -- doesn't tell a manager much. Who cares what the dollars are so long as they are the same percentage of revenue as they have been and as the model says they should be? If a ratio is off, Stifler knows exactly where the problem is. He doesn't know what it is. Maybe someone didn't follow the model or the price of phone service went up. The former has one solution; the latter, another. But figuring out where the problem lies doesn't eat up gobs of managerial time. "If you ask Evie how much her department spends," says Stifler, "she hasn't the slightest idea. Why should she? She's the world's greatest trainer."
"If we have to do more training," McFadden says, "it's because we have more staff, which is because we have more patients -- and the dollars are therefore there. In another company I'd probably be in trouble or in an element I wasn't good at. Thinking about dollars would keep me away from what I do best."
