All Posts in Metrics

|In: Metrics|Last Updated:

The world is awash in bad metrics.

Start with a timely example from your favorite daily newspaper or shouting-heads news commentary: “The top 1 percent paid 35 percent of all income taxes” (usually terminated by an exclamation point.)

No, tax policy isn’t relevant to leading and managing an IT organization, but devising and interpreting metrics is essential to the job, so what the heck. Tax policy is, if nothing else, more fun to tear into than IT SLAs. So let’s dig in, starting here: This statistic has no meaning.

None, because there’s no referent. If the top 1 percent earned 100 percent of all income its tax burden is way too low; if it earned 5 percent it’s probably too high. According to the Tax Foundation, the number is 18.7 percent, which might lead you to conclude the top 1 percent is being unfairly burdened by current tax policy.
Perhaps they are. But we’re far from finished, because:

The 18.7 percent number ignores income sheltering. It also ignores the income retained by corporations in which the 1 percent hold large amounts of stock.

Then there’s the question of why we’re only including income tax. Surely, what matters is the total tax burden – federal income tax, state income tax, payroll taxes, sales taxes, property taxes, and, in some states, personal property taxes.

Interestingly enough, comparisons of income to total tax burden are hard to come by.

For that matter, should we be thinking in terms of total income, or should we be thinking in terms of disposable income after an allowance for a person’s basic needs are subtracted out? The top income brackets have a far greater share of disposable income than 18.7 percent no matter where you draw the blurry line that separates necessities from nice-to-haves.

My point in all this isn’t to suggest the top 1 percent, top 0.1 percent, or top anything percent are paying too much or too little in taxes, let alone whether the country’s total tax rates are too high, too low, or are, Goldilocks-like, just right.

My point is to suggest we all need to start crying FOUL! every time a politician or member of the commentariat parades yet another meaningless number in front of us to prove how awful it all is.

These fine folks probably aren’t deliberately trying to deceive us. I suspect it’s worse: I don’t think they know how to devise and interpret useful metrics at all. They’re just passing their ignorance along to the rest of us.

Oh, by the way: This whole conversation is probably the wrong conversation. A better one would start with the questions: (1) What kind of society do we want to have? (2) What roles should government play in providing it? And (3) what’s the best way to pay for it?

We wouldn’t all agree, but at least we’d know what we’re disagreeing about.

Lest you think this has no relevance to IT, take a look at (to choose an easy target) just about any IT benchmarks ever published. Your company’s CEO probably looks at, if nothing else, the ever-popular percent of revenue spent on information technology, so you’d better look at it too.

It’s a dopey metric for so many reasons they’re hard to count, yet its popularity persists. Many reasons? Sure, like:

  • Is a smaller number better or worse? Most CEOs will assume less is better. Spending less means a smaller number for operating expenses for the same amount of revenue, and that’s good, isn’t it?

Sure, except when it means fewer investments in information technology that would increase revenue or reduce other operating expenses.

  • Are the numbers comparable? Different companies have different levels of IT spending outside the IT department. Some companies have outsourced whole business processes, with the relevant IT owned, managed, maintained and enhanced by the BPO vendor – they’re still paying for the IT, only it doesn’t show up on the books.
  • Has your company invested more in information technology in the past? The more IT you’ve built, the more you spend on operations and applications maintenance to keep the joint running. To avoid exceeding the percent-of-revenue benchmark that means you have to spend less on new development.

Just like the tax question, the percent-of-revenue benchmark also leads to the wrong conversation. The right conversation? It starts with very much the same questions as the tax conversation: (1) What kind of business do we want to run? (2) What roles should IT play in providing it? And (3) what’s the best way to pay for it?

Business managers probably call the IT chargeback a tax, too.

Comments

|In: Cloud|Last Updated:

I have a new set of hearing aids. In the instruction manual, well before the explanation of how to change amplification and programming, is this:

You are not allowed to operate the equipment within 20 km of the centre of Ny Ålesund, Norway.

There’s no explanation for the rule, just the fact, which is why my wife and I were briefly tempted to burn some frequent flier miles, just to break it.

But cooler heads prevailed. Actually, colder heads — we live in Minnesota, which we figured is bad enough (Google Maps reveals Ny Ålesund is on an island roughly 1,000 km due north of Lapland).

Which gets us to another disadvantage of relying on policies, standards, and enforcement to make sure how you want everyone to do things around here becomes how everyone actually does do things around here, beyond those mentioned last week: You have to explain your reasons, which makes your policies and standards burdensomely long. If you don’t, you’ll tempt employees to violate the ones that make no apparent sense, just to see what happens.

Changing the culture simply works better. When enough people internalize how we do things around here, peer pressure becomes your primary means of enforcement.

How to change it? You’ll find a detailed account in Leading IT: <Still> The Toughest Job in the World. Glad you asked.

The short version is to change your own behavior, because culture is the learned behavior people exhibit in response to their environment, and leader behavior is the dominant aspect of their environment.

Before you do, you have to describe the culture you want, and there’s a gotcha. The temptation in describing “how we do things around here” is to be procedural: “When someone contacts the service desk, we first identify the caller, next assign a ticket number, then get a description of their issue,” and so on.

But culture isn’t a matter of procedure. It’s a reflection of shared attitudes. Your behavioral description of culture should reflect this — something like, “When someone contacts the service desk we assume they’re experiencing a real problem, and we take ownership of it.”

<SnideComment>Given my experience with service desks, and in particular with my current mailing service after many subscribers received five copies of last week’s column, I’d say this would represent a radical cultural shift in far too many.</SnideComment>

To change your culture you have to describe both the culture you have and the culture you want. You have to figure out what about how you currently behave results in the culture you currently have, and how you’ll need to behave to get the culture you want.

If there are other managers between you and the employees whose behavior you want to change, you have to pay close attention to how those managers are behaving, how you want them to behave, and what you have to do so they’ll behave that way.

A few subscribers asked if there’s a way to change the culture quickly.

The answer is yes. Actually, there are two.

The first is to lay off a significant number of the employees you have and hire to the new culture. It’s unpleasant to say the least — unpleasant for you, more unpleasant for the surviving employees, and … and I hope this is obvious … even more unpleasant for the dear departed.

Although to be fair, on the pleasantness scale the employees you hire as replacements might very well find the change quite positive, all in all.

Anyway, massive layoffs are quick-culture-change tactic #1. The second one is slightly less draconian — fire all of the managers whose behavior seems to be driving the old culture and replace them with managers who seem to have the attitude you’re looking for.

Yes, it’s ugly. No, I don’t generally recommend it. But if you need to turn around a seriously dysfunctional culture quickly, this is your most efficient alternative.

Start with the ringleader, and perhaps his/her chief acolyte. Reason #1: Fire all the managers at once and the disruption will be too great. Reason #2: Persuading HR to go along will be a challenge. Reason #3: Do you really want to be that kind of person? And most important, Reason #4: Once you’ve fired one or two, the rest will usually figure out you’re serious and change their behavior to match what you’re looking for.

And, in case this isn’t clear, you still have to change your behavior (and attitudes) too. Otherwise, the culture will gradually revert back to the one you say you don’t like.

And you’ll have to go through the unpleasantness all over again.

* * *

Four years ago in Keep the Joint Running, Gartner predicted that in just two short years, 20% of all companies would have no IT assets of their own — it will all have moved to third parties and the cloud. KJR’s rebuttal was suitably pungent.

And eight years ago you read about a popular technique for manipulating people.

Comments