eight

A Timing Analysis

Seven Steps

… not to give a specific answer, but to suggest a way of study.

—Josef Albers¹

A timing analysis is a structured method designed to help you achieve a number of important goals.

First and foremost, it will help you find windows of opportunity. A timing analysis will clarify when the moment is favorable to act and when it is not. Next, it will help you predict the risks associated with action and inaction. Everything we do has timing-related risks. Some may not be obvious. We may be unaware of how quickly a situation can change, for example. A timing analysis will bring these risks to light. It will also help you determine the importance of timing in a given situation. Sometimes timing does not matter. In other situations, timing really is everything. Because managers are engulfed in what seems like a tidal wave of information, timing issues can get lost. A timing analysis will surface them. It will also help an executive decide how to proceed. Should she move quickly or slowly, perhaps pause before moving ahead, or match the pace of her own actions to the rhythm of external events, speeding up or slowing down as needed? I call these decisions about rate, punctuation, entrainment, and other elements of temporal architecture the choice of temporal design.

Done well, a timing analysis should convincingly confirm or challenge conventional wisdom and surface a wider range of options than is first apparent. It will also clarify what is vague or experienced only as an intuition or feeling in the gut.

A timing analysis consists of seven essential steps. I'll summarize each before going into a more detailed explanation of the process.

As you work through these seven steps, keep in mind Einstein's comment about simplicity—namely, that “everything should be made as simple as possible, but not simpler.” In conducting a timing analysis, the devil is often in the details. Several steps are complicated. The road through them has many twists and turns, and the speed limit is probably no greater than 30 mph. If there is a lot at stake, then there is good reason to take your time. There is a lot to see and consider. But once you become familiar with the methodology, the steps will become second nature. You will discover shortcuts as well as substeps you can skip.

Keep in mind that a timing analysis is not a formula. It is a way to probe and examine real situations to find answers to timing-related questions. It works with “conditions on the ground”—that is, with the details, uncertainties, and complexities that are present in the messy world in which we live. Because managers are always fighting fires, I'll to use an example from firefighting to illustrate the seven steps.

Now let's consider each step in more detail.

STEP 1: DESCRIBE THE SITUATION

The first step in a timing analysis is descriptive. Describe the issue before you in your own words. Every profession or discipline has its own jargon. People in IT, marketing, and strategy each have their own ways of thinking and talking about events. It is useful to write down your initial impression so that you can add or adjust details as you go along. This description should be no longer than a few pages. Next, ask yourself: Given what I want to accomplish, what timing issues will I face? Perhaps the issue is when to test-market a product, when to exit a failing business, or how to time a series of steps in a complex process. This initial description should focus on the big picture.

For the firefighter, the goal is to save lives and property. The timing question is how to act as quickly as possible in a manner consistent with safety.

Speed is important in many situations, yet we know from experience that there will be other factors to consider, other questions to ask, and decisions to make—for example, what things should be done at the same time, when is a pause required, can a particular step in a process be skipped entirely, and so on. You will discover these choices and options as you conduct the analysis. At the outset, simply begin by recording your current understanding of the situation.

What kinds of timing issues do you face? For each, list the factors that might affect that decision. Why is acting at one time better than acting at another? In the case of firefighting, the factors involved may include the size and location of the fire, the kind of building involved, whether people are trapped inside, the immediate weather forecast, and the like. This list will come in handy when you begin to sketch a score diagram.

STEP 2: SKETCH A SCORE DIAGRAM

We will use the vertical and horizontal structure of a musical score to create a visual representation or sketch of the factors involved in making a timing decision. (If you are not familiar with what a musical score looks like, examine the score from Beethoven included in the Introduction.) The parts played by different instruments are stacked vertically. The result is that when you look at a score, you can see what follows what (which note follows which), and what is going on at the same time (what chords and harmonies are present). The score diagram that you will sketch—no artistic or musical talent required—will be similarly organized. It will have a vertical and a horizontal dimension.

The act of sketching is important. I do not know why children draw, but I suspect one reason is that drawing is our instinctive response to a world we do not yet understand: we master it by giving it visual shape, a shape that we control. Drawing makes the world graspable; it puts it within the reach of our hands, our eyes, and ultimately our mind. It is the same with the patterns of temporal architecture (see the Appendix for a more complete description). We can't image and explore these patterns in our head. They are too complicated. We need to actually sketch or draw them. And although computers are useful for analytics and calculations, there is an awful lot we can learn with nothing more sophisticated than a pencil and the back of an envelope.

List the Relevant Factors Vertically

Begin by giving each independent factor its own horizontal line. You will remember from the polyphony chapter that these lines are called tracks. You can have one set of tracks for whatever is going on inside your organization, another for the actions of competitors, another for changes in government policy, and so on.

Add tracks for everything you think is important. I can't tell you how many tracks to include because every situation is different. This is where your own knowledge and experience come in. Obviously, you can't include a track for every factor or variable, every actor or stakeholder, but it is better to be overly inclusive at this stage than to miss something important. In my experience, most complex situations need at least two dozen tracks to capture what is going on. As a result, most diagrams will be tall.

Describe and Detail the Horizontal Tracks

Next, fill in what you know about each track. Examine them through the six lenses. Ask yourself about

Firefighting: Sketching a Score Diagram

The following is a common sequence of actions that occur after a firefighter arrives at the scene of a burning house or building:²

1. Ventilate the structure. Firefighters make holes in the structure to allow the hot gases and smoke to escape.
2. Search and rescue. While the structure is being ventilated (and often before the first drop of water hits the fire), other firefighters break windows and begin to search inside.
3. Use water to extinguish the fire. Turn on the fire hoses.

In the score diagram in Figure 8.1, I have given each of these actions its own horizontal track. I have also added a track for the fire to indicate the changes in its intensity over time. Take the time to study the diagram. It's less complicated than it seems. It's simply that you haven't seen a diagram like it before.

Let's look at the diagram through each of the six temporal lenses we discussed in previous chapters. First, firefighting is a situation in which speed is of the essence. Hence the rate (indicated in the upper left-hand corner) at which various actions will be taken and events will occur is prestissimo, Italian for “really fast.”

Figure 8.1 Conventional Firefighting Score Diagram

I like to put actions that I have control over in the top section of the diagram, and actions and processes that are part of the context and that I might not have control over lower down. I use the G clef (the S-shaped symbol) for the former, and the alto clef (one half of a figure 8.1) for the latter. Because many things are going on at the same time on different tracks, I have included a label for polyphony. I've placed a time line with tick marks at the bottom of the diagram to indicate important intervals. I also indicate that the timescale of the diagram is one of minutes (M), rather than hours or days, for example.

Each action has its own track. There is a track for ventilation, another for search and rescue, and a third for the use of water, which is the order in which these activities take place. As a general rule, I give each step or stage in a sequence its own track. As you follow each track from left to right, you will find marks of punctuation. Every action or condition has a beginning and ending, which can be abrupt or gradual. I use the letter P as a placeholder to remind me that punctuation may be important. Some fires, for example, may take a long time to fully extinguish.

Notice the intensity of the fire as it changes over time (shape). The fire will subsequently burn more intensely after firemen ventilate the structure and break windows, which provides the fire with a new source of oxygen.

The shaded vertical column defines the window of opportunity for conducting search-and-rescue operations.

What I have drawn is intended as a sketch. When architects design a building, they begin with a sketch rather than with a detailed set of engineering drawings. We are not after mathematical precision. Later, you can superimpose the normal grid of days, weeks, and months on the diagram, and include measurements that are more precise. But that should be the last step. Remember the mantra: before precision, pattern.

STEP 3: PROBE MORE DEEPLY

You now have a basic score diagram that represents a beginning attempt at capturing the essence of the situation you face. The next step is to explore, expand, and refine the diagram using two factors that I discussed in the polyphony chapter: structure and influence. We will begin with structure—the vertical relationships among the tracks.

Explore the Vertical Relationships Among the Tracks

Look at the different tracks in the score diagram and ask yourself:

• What is aligned and not aligned? Does the launch of a product correspond with a spike in demand for it, for example?
• Which tracks lead, lag, or overlap? Are three companies launching virtually identical products at the same time?
• Are there gaps, and do they matter? A company can struggle if there are long gaps between new products.
• Which tracks or parts of tracks are synchronized or out of sync with others? How are the internal systems within the company aligned or out of sync with changes in the market?
• How do the tracks influence one another? Do the contents of one track cause, mask, amplify, or compete with what is going on in another track?

A timing analysis draws your attention to the relationship between the horizontal and vertical dimensions of a situation. In most cases, you will not know at the outset which parts of the score will give you the most information about timing. It could be the realization that one step or stage must follow another, that two processes should not be allowed to overlap, or some other pattern. (Review the polyphony chapter for a more complete discussion of the ways one track can influence another.)

Adjust the Size of the Diagram

Ask yourself:

• Have I included enough tracks and enough detail about sequence, punctuation, and so on within each track?
• Is my diagram “tall” enough?
• Is it also long or wide enough?
• Does it include enough of the known past and probable future to make sense of the situation?

If your answer to any of these questions is no, go back and add tracks and fill in the blanks.

When we don't look far enough into the past or forward into the future, we can be surprised by what happens. For example, everyone believes that good performance should be rewarded over bad performance. But we need to be careful about how we measure “good” and “bad” performance—and what effect that measurement will have on the outcome. For example, a California study of thirty-five thousand physicians who were graded on their performance revealed that some doctors dropped “noncompliant” patients and those with a poor prognosis or a complicated illness. In other words, they dropped patients who would lower their rating.³ A score diagram that included a track for physicians and separate tracks for different types of patients might have identified the problem before the program was implemented. What is obvious after the fact should have been obvious at the beginning.

When something goes wrong, the proposed solution we hear most often is to tighten control—strengthen oversight or regulation. But that skips a step. The first step is to better describe the situation to be controlled. That includes taking into account what can change with the passage of time. To do that, you will need a score diagram that contains enough tracks, and long enough tracks, to capture all the important time-dependent phenomena. Everything we do takes time—happens before, during, or after something else. If we don't consider what doctors are likely to do before agreeing to see a patient or what their decisions might be during the course of treatment, the results of an intervention or change in policy may come as a surprise.

Ask yourself:

• What might happen in the future that will impact my business?
• What from the past continues to impact the present?
• What else is going on at the present time that I may have overlooked?

The fact that we operate in a global environment means that there are always more factors to consider. Ask yourself what would happen if the diagram you are working with were taller—that is, if it contained more tracks—and how these added tracks could change your decisions about timing.

There is a lot to be seen through each timing lens. The firefighting example, simple as it is, illustrates the importance of determining the maximum speed of the fastest process (how quickly the fire can spread), noting the presence of termite tracks (the rate at which plastics replaced cotton), being aware of punctuation (when is the fire really over), and choosing the right sequence (search and rescue followed by water, or vice versa). Were we to look at the firefighting example more closely through each lens, additional issues would be revealed. For example, when I drew lines for ventilation, search and rescue, and the use of water, I didn't know how long each of these processes would last. How long should firefighters wait after drilling a hole in the roof to enter the dwelling? Maybe half a minute one way or another would make all the difference—or perhaps no difference at all.

As you use all the lenses, and add the details that they surface, you may find that the score diagram you are sketching is becoming too big and too complex. There are two strategies for dealing with complexity. The first is an organizing strategy. You can group the tracks. The second is to redraw the diagram using a more macro timescale. I describe both strategies here.

Group the Tracks

I have suggested that you use the G, alto, and bass clefs for tracks that describe the foreground (actions you can take), middle ground (your immediate context), and background (the more remote context). Usually you have the greatest chance of influencing what is going on in your immediate environment—for example, inside your team or group. Few of us can affect the global economy. So one way to group tracks is according to what you can influence. Another way is to group them by importance. Which tracks are likely to have the greatest impact on the outcome you desire? You can also group tracks by rate or speed, putting the fastest tracks together. Another option is to group by stakeholder. There is no best way to group tracks. In fact, using different groupings will help you not only manage the complexity of the diagram but also gain greater insight into the situation you must manage.

Use Different Timescales

If a diagram becomes overly complicated, and grouping strategies are not sufficient, break off tracks and develop a separate diagram. Conducting a timing analysis is an art, so be prepared to move back and forth between focusing on a detail and working with the larger composition. For example, in the case of firefighting, you could sketch a detailed diagram just for the search-and-rescue portion, assigning a separate track for each individual and what he or she is doing at each moment. You can also redraw the diagram using a more macro timescale. If you have been looking at a business situation using one timescale, consider using another. If you plot sales every hour, you will see one thing. If you look at the pattern of sales month-by-month over the course of a year, you will likely see something else.

By this point you may have discovered everything that you need to know or have time for. Yet if you want to extract the maximum amount of information from the score diagram, there is more that can be done.

Modify the Diagram

Sometimes the best way to understand something is to change it. Here are three ways to further explore a score diagram.

Change the Direction or Magnitude of the Forces

Actions are always being pushed or pulled in one direction or another by different forces. Ask yourself: What happens, for example, when a deadline is removed? How long will it take to finish a project without the pressure to complete it by a specific date? In the firefighting example, suppose you hear a woman crying out for help just inside the front door of the building. You probably wouldn't wait for water cannons to douse the fire before attempting a rescue. Prescriptive rules about what to do (and when) can be overridden by the force of circumstance.

Change the Alignment of Tracks

Ask yourself: What are the consequences of changing the alignment, overlap, leads, and lags among the different tracks? In the firefighting example, when should water come first rather than last, and what impact would that have on the other tracks?

Vary the “Loudness” of Each Track

Try making some tracks louder or softer—that is, more or less important or prominent than others.

I had a conversation with a manager who had recently joined a company to head mergers and acquisitions. But soon after he arrived, a battle over CEO succession left little time or energy to consider mergers or acquisitions. The manager told me that he wondered if he had made a mistake. When there is a firefight going on inside a company, it will tend to consume the resources needed to find and exploit outside opportunities. When firefighters face a real fire, a prime consideration will be the number of people inside the building and their location. If and when that changes, so will their tactics and strategy.

Use the P4 Strategy

Some people think visually, others in terms of numbers or symbols. The score diagram is a visual image. Let me give you another way to think about it. When you sketch a score diagram, you are executing what I call a P4 strategy, which represents a movement from point to path to polyphony to pattern.

When you face a timing question, ask yourself: What do I know now, at this point in time? Then think about time-extended paths. Ask, What sequences of events are playing out before me, both inside my group or company and in the larger environment? The movement from point to path allows you to begin to envision the horizontal dimension of the score. Think about what follows what (sequences); where things begin, pause, and end (punctuation); how quickly they might develop (rate); how long they could take (duration); how much time separates the different events you have identified (interval); and what time-extended shapes, such as the rise and fall of the business cycle, are present that might influence your decision. Because we exist in a world of simultaneous events, think about what is likely to be going on at the same time (polyphony). That is the vertical dimension of the score. Then look for the patterns that are formed as these parallel processes play out together. Asking these questions is the verbal equivalent of sketching a score diagram.

STEP 4: LOOK FOR WINDOWS OF OPPORTUNITY

A timing analysis is designed to help you decide when to act. The gray columns in the firefighter diagrams point to an interval when search-and-rescue operations can be carried out safely. We often talk about windows of opportunity. When is the time favorable for what we want to do, and when is it not? The metaphor of a window suggests an interval, often brief. The implication is that if you don't act quickly, you will miss it and the window will have closed. But that is an oversimplification. Windows have other characteristics besides their length or duration. Here is what to look for.

Sign

Is a window of opportunity open (+) or closed (−)? Sometimes it is just as important to know when a window is closed as it is to know when it is open. In the firefighting example, it is just as important to know when not to enter a building as it is to know when the time is right to do so.

States and Dates

When will a window open? State rules give one set of answers to the question. State rules are “whenever” rules. A window is open whenever a specified state or condition is present. For example, “Don't fire till you see the whites of their eyes” was a command issued by Colonel William Prescott during the Battle of Bunker Hill. We may not know when that time will occur, but when it does, FIRE!

State rules can refer to any element of temporal architecture. When is the best time to buy stocks? According to some analysts, wait until there are three or four rebounds, essentially an extended W—a shape-based rule. Sometimes the right time to act is just before a deadline (a mark of punctuation), because that is when people are preoccupied and one has the element of surprise. When you have found—or think you have found—a window of opportunity, note which feature(s) of temporal architecture are the reason you believe it is present.

The simplest timing rule is a date rule: act at a predetermined time or date. “The best time to do that is right at the end of the fiscal year.”

In the firefighting example, all timing rules are likely to be state rules. Firefighters watch the fire, not the clock, to time their actions. In other contexts, such as law or accounting, date rules may predominate.

Length and Height

How long will the window remain open (its length), and to what degree (its height)? The latter indicates how favorable that period of time is for what you want to accomplish. Some windows can be raised only a crack. You may succeed if you act during that time period, but success is far from guaranteed. Other windows can be opened fully, even if they remain so for only a short period of time.

How long, and with what degree of safety, will firefighters be able to continue search-and-rescue operations? The answer depends on the height and length of the window.

Punctuation and Shape

Some windows, no matter their size, pop open and slam shut. Others open and close more gradually. It is always useful to sketch the curve that describes how the window opens and closes. Some windows may open so gradually that you don't even realize that there is a window of opportunity until a competitor takes advantage of it. Others open with a bang: everyone in the neighborhood knows there is an opportunity to be exploited.

Singularity

Is the opening of the window a one-time event (a singularity)? If not, when will it reopen? If you miss an opening, will you have another chance? And if so, when?

If firefighters are forced to exit a building engulfed in flames, when will it be safe to return?

Synchrony

What must be present (synchronous requirements) for the window to open? What must be absent (synchronous risks)? In the case of the latter, what conditions, if they occurred at the same time, would prevent a window from opening or force one already open to close?

In the firefighting example, the fire must be sufficiently contained that search-and-rescue operations can be conducted safety.

Assessing the Cost of an Error

Once you have found a possible window, ask yourself:

• What is the cost of missing it?
• What will happen if I am too early or too late?
• Does how early or how late matter?

To investigate this question, plot a cost-of-error curve (COE), as illustrated in Figure 8.3.

In this example, the cost of being early or late is not the same. The COE curve is not symmetrical. Clearly, you don't want to be too early. But if you are late, it doesn't seem to matter how late you are. The cost is the same: very high. An analysis of a window is not complete until you have plotted a COE curve.

In 2011, a firefighter was very badly burned in a brownstone fire that officials said “fed quickly on home furnishings and inrush of air through open windows.”⁴ Yet delay in beginning search-and-rescue operations can cost lives. COE curves will always be on the mind of those fighting fires.

Figure 8.3 Cost-of-Error Curve

STEP 5: IDENTIFY TIMING-RELATED RISKS

Look at the score diagram you have sketched and ask yourself: What timing-related risks does it surface? There will be both element-specific and pattern-level risks.

Use the summaries at the end of each of the lens chapters to look for element-specific risks. For example, in a sequence, the risk is that something may occur out of order. You expect A to be followed by B, but B occurs first. Each element will have its own set of risks.

By pattern-level risks, I mean risks that result from the way two or more elements relate to each other or interact. In the firefighting example, we saw that a change from cotton to plastics affected when search-and-rescue operations should be carried out.

Clearly, the more you know about the risks related to your plans and projects, the better prepared you will be to avoid or manage them.

STEP 6: EVALUATE YOUR OPTIONS

Appraise and evaluate the timing analysis you have conducted. Ask yourself: Has it surfaced timing issues that I didn't see originally? Has it helped me locate windows of opportunity and understand their characteristics? Use the checklist here to assess the adequacy and completeness of your analysis.

A good timing analysis should help you

• Find the information you need for you to decide when to act. It should help you put your finger on exactly what matters.
• Decide whether the timing strategy or rule you started with (for example, act immediately) was the right one.
• Challenge or confirm conventional wisdom.
• Surface a wider range of options than were first apparent. Some may suggest creative or innovative solutions.
• Clarify the relationship between what you intend to do and when you intend to do it.
• Help you decide the importance of timing. Sometimes timing may not matter. Sometimes timing is everything.
• Clarify, confirm, or perhaps even challenge your intuition.

STEP 7: ACT

Act on the basis of what you have discovered: immediately if time is of the essence, later if you have determined that it is better to wait. Sometimes the answer to a timing question is obvious. At other times there is no choice. (Speak now, or forever hold your peace.) Sometimes timing may not matter because you have the power to do what you want anytime. Or perhaps one time is as good as another. There is no substitute for experience and good judgment. But there is one thing of which I am certain: when you conduct a timing analysis, you will discover that you know more about timing than you thought you did. You simply needed to access that knowledge and put it to good use.

Obviously, it is better to conduct a timing analysis before you act, rather than after. But even after the fact, conducting a timing analysis can be helpful. It can be used to defend a decision as timely even if the outcome was less than satisfactory. It can also be used to learn what went wrong and how timing was involved.

As you have no doubt observed, a complete timing analysis takes time and, to do well, training. We have all heard the term analysis paralysis—too much time analyzing, not enough time doing what needs to be done. A timing analysis, however, can actually speed up decision making. What was previously unknown and uncertain will, as a result of a competent timing analysis, be much better understood. When a situation is clarified, uncertainty becomes risk, and risk can be managed. In the face of uncertainty, we tend to be impulsive and act too soon, or we hesitate and act too late. Conducting a timing analysis improves the odds of getting the timing right the first time, and hence saves the time needed to correct a mistake, which can be considerable.

There is another reason why learning to conduct a timing analysis will save time in the end. What makes a timing analysis powerful is the fact that the same temporal pattern can exist in many different contexts. That means you can transfer what you learn from one context to the next. You don't have to reinvent the wheel. Recall the ketchup-and-Rogaine example I described in the Introduction. Both had to do with marketing and product design opportunities that came from seeing the underlying “sequence of use” for each product. Let me give you another example of the value of seeing the same temporal design in different situations. In this case, the issue is one of asynchronous risk—the risk that what should occur at the same time, doesn't.

Computer encryption. When we turn off our computers, the information stored in the DRAM (dynamic random-access memory) chip is supposed to disappear along with the algorithms (the keys) necessary to encrypt the data in the computer. Researchers at Princeton discovered, however, that “when the chips were chilled using an inexpensive can of air, the data was frozen in place, permitting the researchers to easily read the keys.”⁵ Essentially, one part of the computer turns off more slowly than the rest. The fact that one can retrieve data after the computer is turned off makes it possible to steal encrypted information.

The lost violin. On May 7, 2008, the Russian violist Philippe Quint got out of a cab at Newark Liberty International Airport. He walked to the back of the cab, collected his luggage from the trunk, and moved it to the curb. He had already paid his fare. The cabbie closed the side door of the minivan and drove off. Unfortunately, Mr. Quint's $4 million Stradivarius violin was still in the backseat. According to Quint, he did not forget the violin. He left it in the cab for safekeeping while he moved his bags from the trunk to the curb. He intended to retrieve it, but the driver pulled away just as he slammed the trunk. After a number of frantic phone calls, the cab was located, and the violinist and his violin were reunited. A few days later, Mr. Quint returned to Newark Liberty and gave a thank-you concert for the taxi drivers.⁶

When a process is complete, we want everything tied up in a nice knot, no loose ends. If something is turned off, we assume that all components will shut down at the same time. When a cab fare has been paid, we expect the violinist to take his priceless violin with him as he leaves the cab. Asynchronous endings are common, however. Whether you are in charge of managing an exit, closing down a factory, discontinuing a product, or selling an organization, expect different components of the process to have their own pace and to last different amounts of time. When you have the concept of asynchronous endings in mind, you are better prepared to manage the difficulties such loose ends present.

And now, with these two stories, we are truly at the end of this book. There is nothing more that I have time to add. All that's left is the Coda.