Showing posts with label Paralysis of Analysis. Show all posts
Showing posts with label Paralysis of Analysis. Show all posts

Monday, January 27, 2014

Strategic Planning Analogy #520: Whiteout!


THE STORY
I’ve been suffering through this winter like most everyone else here in the United States. A couple of days ago, it was snowing and blowing so bad that we had “whiteout” conditions.

A whiteout occurs when there is so much snow blowing in so many directions that you cannot see anything but a wall of “white.” It’s like being locked in a totally dark room where you can see absolutely nothing—except instead of total blackness, you have total whiteness.

I’ve been caught in whiteouts when driving on expressways. It is extremely dangerous because not only can’t you see the road, you cannot see what the other drivers on the road are doing. You may as well be driving blind.


THE ANALOGY
Businesses can also experience a form of whiteout. Except instead of being blinded by snow, they are blinded by an excessive flurry of data. With unending streams of data flying from all directions with no end in sight, it is easy to get lost.

There are those who say “the more data, the better.” But like snow, too much data can be a dangerous thing. It can overcome businesses and make it impossible to find the way forward. One can become so bogged down in looking down at data that looking up to make forward progress comes to a halt.


THE PRINCIPLE
The principle here is that obtaining data should not be the goal. The real goal is to discover and reach your vision. Data is only useful in this endeavor if two things occur:

1.     It is converted into knowledge:
a.      Insight to help discover the vision;
b.     Insight to deliver the promises of the vision; and
c.      Monitoring knowledge to make sure you are on track.
2.     It does not bog down forward progress towards your strategic destination (paralysis of analysis).

In other words, if all you have is a big pile of data, you have nothing. In fact, it is worse than nothing because of all the wasted time and effort to gather it and stare at it. Instead, what you want is a smaller pile of knowledge.

Knowledge is like a small map you can take in your car telling you where to go. The knowledge map is useful because it distills the vast outdoors into just what you need to motor along. By contrast, data is like all that snow that is still blowing around outside. Instead of giving knowledge of where to go, it prevents you from knowing where to go.

We can learn three things about how people deal with snow to help us understand how to deal with data for strategic purposes.

1. Look at Radar, Not Individual Flakes
If you really want to understand how to deal with snow, you need knowledge. And what knowledge is that? As I’ve stated in previous blogs, strategic knowledge of the environment usually boils down to three things: Magnitude, Direction & Speed. If you know this about a trend, then you typically know what strategic action to take.

For example, household car ownership in the US peaked in 2007. The current direction in the percentage of households owning cars is moving down. If we add to this the best knowledge on the anticipated magnitude of this trend (how low will car ownership go) and the speed (how fast will ownership drop), then we can build an intelligent strategy to deal with this trend.

The same is true of snow. If you know Magnitude (size of the storm), Direction (where it is coming from and where it is going) and Speed (how fast the storm is moving), then you will know how to deal with that snow storm. This is what the TV weathercasters talk about—magnitude, direction and speed—because they know this is the knowledge you need to make the right decisions.

They don’t get this knowledge by looking at individual snowflakes. In fact, they don’t have to really look at any snow at all. Instead, the weathercasters get this knowledge from understanding the big picture. And the big picture comes from looking at radar images rather than snowflakes. Radar captures the entire storm at once and lets you measure magnitude, direction and speed.

Strategists need to do the same. They need to stop obsessing with individual data factoids and look at strategic radar which lets one see the big picture—big enough to show direction, magnitude and speed.

You won’t get the big picture on car ownership by staring at everyone’s driveway one at a time. You need to get broader—and look at something other than cars and driveways. For example, what are the key “driving” forces behind choosing not to own cars. Is it:

1.     Population migration to dense urban centers?
2.     A different attitude towards car ownership among younger adults?
3.     A growing concern for the environment?
4.     A poor economy?
5.     Advances in car sharing options (like Zipcar)?
6.     Less need to travel due to being able to get tasks done at home via the internet?
7.     A combination of the above?

Get the big picture on this (via strategic radar), and you can start to make educated projections on the direction, speed and magnitude of car ownership. Now you have knowledge instead of data.

2. Plow Away the Unneccessary
Snow on the roads is considered a bad thing. Therefore the plows are brought out to clear away the snow on the road. The drivers of the snowplows do not stop to examine every snowflake on the road to determine which ones are good and which are bad. No, the drivers previously determined that if it is on the road, it is bad and needs to be plowed away. No further examination needed.

The same is true in business. A lot of data is strategically worthless. Its speed, direction and magnitude have no relevancy to advancing my strategy. Therefore, rather than spend time examining it, I should just plow it away so that I can move forward.

For example, Walmart’s strategy centers around owning the low cost, low price position. Therefore, Walmart can focus their attention on only dealing with finding knowledge of issues impacting Walmart’s ability to own the low price position.

When Walmart determined that:
a)     Warehouse clubs and supercenters had the potential to offer lower prices than Walmart discount stores; and
b)     Consumers were starting to prefer these formats (direction, speed and magnitude moving their way),
Walmart changed its strategy and diversified into warehouse clubs and supercenters.

Recently, they made the same determination about internet shopping and are quickly and aggressively moving in that direction.

Everything else was plowed away so that they could focus on what really mattered—owning low price. By knowing where to focus, they could ignore and just get rid of all the debris that does not impact that focus. This allows Walmart to efficiently and effectively own its position over many decades.

That is why focus is so important to strategy. Focus lets you know what to look at and what you can ignore. As I mentioned in a prior blog, it is often more important to know what your strategy isn’t than what it is, because there is a lot more data which is worthless than is worthwhile. Defining what is outside your strategy lets you know what is worthless and can just be plowed away.

3. Fly Above the Clouds
No matter how bad the snowstorm, airplanes can usually avoid the turmoil by flying above the clouds. It’s always sunny and snow-free above the clouds. And that leads to quick and easy travel to the destination.

A similar situation exists in the business world. Current business fads and daily crises can cause all sorts of turmoil. Bouncing from fad to fad or crisis to crises is like bouncing around in a jet going through a storm. It slows you down and keeps you from your intended destination.

Strategists need to get companies to rise above these current temptations which suck up a company’s time, just like jets rise above the clouds to escape turmoil. Once you get above the clouds, you can see clearly. The same is true in business. If you stop getting bogged down in the petty distractions of the moment, you can see more clearly what is truly important.

Rather than follow the current fad, follow the larger game plan. After all, one rarely wins a strategic position when following others to get to the same location as they are. Winning comes from differentiation, not imitation. Rise above the fray to clear the path to your intended destination—the place where you can win.


SUMMARY
Although some data can be useful for strategy, most is just a wasteful distraction. And even the useful data only becomes useful if it is converted into knowledge. Therefore, instead of wasting time trying to absorb as much data as you can, follow the tricks used to deal with snow:

  1. Focus on the Big Picture by looking at “Strategic Radar” (showing direction, magnitude and speed) rather than looking at every snowflake (piece of data).
  2. Just plow away all the data not relevant to your task of moving forward towards your winning point of differentiation.
  3. Fly above the clouds of current fads and distractions so that you can easily see the final destination.

FINAL THOUGHTS
Remember, the winner is not the one who captures the most data, but who gets to the right destination first with the right offering. Don’t let competitors in snow plows pass you by while you stop to look at every snowflake.

Friday, May 22, 2009

Strategic Planning Analogy #259: Quant Jocks


THE STORY
The term “Quant Jock” refers to people who earn their living by being excessively good at developing complex analytics via the computer. Quant Jocks have been an integral part of the Wall Street financial community for years. However, in the past year or two, their reputation for financial wizardry has become a bit tarnished.

First, it was the quant jocks who helped develop all of those sophisticated repackaged mortgage bundles, which nobody really understood and which were a major factor behind the recent housing crisis. Second, it was the quant jocks who developed sophisticated computerized stock trading programs. These sophisticated stock trading programs helped to increase the negative impact of stock melt-down in the fall of 2008.

The irony is that the quant jocks had claimed that all of their sophistication would help to reduce downside risk. Instead, it now appears that their models actually increase risk, particularly if events fall outside the programmer’s narrow assumption parameters (which inevitably will happen).

THE ANALOGY
There are many similarities in the goals of Wall Street and of business strategists. Both are trying to find a way to optimize the blend between profitability and risk. The goal is to create as much profitability as possible within a particular risk tolerance.

For years, Wall Street has used a lot of quant jocks in the attempt to achieve that goal. Now, we are seeing more and more of that complex analytical approach being applied to strategic planning’s goals. Planning techniques like Scenario Planning and Real Options seem to be falling under the influence of quant jocks.

At its worst, overly-quantified scenario planning can become similar to those mortgage-backed securities which helped bring down the housing market (if the quant jocks are allowed to go wild). You bundle up all these scenarios into one massive computer program and come out with some sort of bundled scenario risk formula that doesn’t quite match any particular scenario. This makes it hard to know what to do.

The same can happen to a Real Options approach to minimizing risk—lots of math leading to conclusions that tend to mask what is actually going on in a strategy. If the parameters are in the assumptions are off by too much, they whole thing can collapse.

It’s not that math or analytics are bad per se. Scenario planning and real options can be valuable tools. The knowledge and insight coming from rigorous analysis can be useful—on Wall Street and in strategy. But taken too far, analytics can obscure your view. Instead of knowledge and insight, there are incomprehensibles and too much blind faith in the cold, unthinking calculations of formulas inside a black box. And, as we saw in the story, rather than reducing risk, it can lead to increasing risk—and creating melt-downs. Is this what you want for your strategy?

THE PRINCIPLE
The principle here is that strategies need to be far more than just numbers and formulas. In fact, an excess of “quant jock” thinking can actually increase the risk of failure for your strategy. The logic behind this point of view are as follows:

1. Continuity Vs. Discontinuity
Quant jocks tend to live in a world of continuity. The idea is that the world operates by a set of rules. The role of the quant jock is to model those rules and optimize the nuances for the benefit of the company.

By contrast, good strategists tend to focus on discontinuities. Nearly all great strategic moves are done in a way that totally destroys the rules of the status quo. For example, the great success of the Ipod comes from more than just the creation of a device. It was a reinvention of the rules for the entire industry—of how music was sold (itunes), organized and listened to.

The same was true for the iphone—creating an entirely new apps-based business model. Amazon was not just another outlet for selling books—it was a new way to think about shopping, with lots of new tools and information to create a wholly different shopping experience.

As we saw with the housing crisis and the stock market melt-down, the quant jock systems failed miserably when there was great discontinuity. They weren’t built for such rapid change. I fear the same is true in the strategic world.

Rules-based models don’t help you discover a new set of rules, nor do they tell you how to react when the old rules no longer apply. Rather than overanalyzing the world of today, strategists should be dreaming of how to destroy the world of today for their benefit.

Instead of modeling the world as it is, we should model potential new realities. Yes, there is still some analytics involved, but the analytics are only as good as the dreams being analyzed. Great dreams are more important than great analytics, because the dreams are what create the new business models to be analyzed. Analytics without these dreams is just random noise.

2. Beating the System Vs. Being the System
In the stock market, the goal of the quant jock is to find little holes in the rules, so as to beat the system on very narrow deviations. Unfortunately, what happened was that a large number of firms adopted these models and started trying to beat the system in pretty much the same way.

When that happens, you are no longer beating the system. Instead, you become the system. The herd mentality caused so much trading to be done in this similar analytical manner, it became harder to make the models work (too many people chasing too few holes). Then, when the market fell apart in the fall of 2008, the models all worked in unison to force prices down further and faster.

A similar situation occurs in strategy. “Me Too” strategies, where you try to copy the leader, rarely lead to great riches. The leader usually stays the leader and you fight over the few crumbs left behind.

If the herd mentality takes over and everyone tries to win in the same way, it tends to commoditize the business. This usually leads to price wars, where the prices drop just like those stocks did (and so will your profits).

Analytics by nature tend to mimic others, because they are trying to model the world that exists. Quant jocks may come up with original ways to push around the math, but they rarely come up with original new strategic options. If you truly want to beat the system, you need to create a new position, working under a new set of rules—a place where you can be the leader and the rules work in your favor. Brainstorming, not whirring computers and complex spreadsheets, are the priority.

3. Creating Vs. Measuring
Quant jocks like to measure things. The problem is that when you are creating a brand new business model in a brand new space operating under a brand new set of rules, there isn’t much to measure beforehand.

If you wait until the market is fully developed, so that you have more to measure, it is usually too late. The market leaders have already established themselves and the rules are already working in their favor. The game is over.

Apple succeeds by blazing new trails, doing new things—be it the Ipod/Itunes model, the Iphone/Apps model or the Apple Store model. It doesn’t wait until the market is fully developed and measurable. It takes calculated risks. Sure, calculated risks are still calculated, but the strategy is not put on hold until perfect information is available.

Sometimes, a little consumer research can help. But even here, if the concept is too radical, the customers may not at first be able to internalize how they would behave in that new world, thereby making the results unreliable. Small beta tests may be better than analytical modeling.

Well, if the new world is not yet available to measure, we can always still measure the current model, right? Unfortunately, overemphasis on measuring the business model you are trying to destroy usually will not tell you the best way to destroy it. That’s a lot of effort for questionable reward.

SUMMARY
Strategy is primarily a creative act—building a new business model that did not exist before. A “quant jock” mentality/focus typically is not the path to get there. Instead, it tends to keep one mired in the past. Sure, a little analytics can help fine-tune the creative idea, but it won’t develop it. Therefore, rather than obsessing on analytics, obsess on creative model building and then use analytics as a secondary support mechanism.

FINAL THOUGHTS
Sir Isaac Newton did not discover gravity through analytics. It was a creative burst prompted by watching apples fall from trees. The analytics came later. Strategists should probably spend more time pondering things like falling apples and less time pouring over computer printouts.

Saturday, April 28, 2007

Too Many Clocks

The Story
My first job out of college was as Assistant to the President of a furniture retailer. This retailer sold a fair amount of grandfather clocks—enough so that there was a separate selling room in the store just for displaying a large number of these clocks.

On one of my first weeks on the job, I noticed that the grandfather clocks were all set at different, and seemingly random, times. I thought I would do everyone a favor and set all of the grandfather clocks to the same time that was on my watch. There were many clocks, so this took a bit of time.

After I finished, the president of the company saw what I had done and told me to go back and reset all of the clocks to different random times. I was a little upset because of all the time it took to set the clocks to the time on my watch, so the president explained why he wanted them switched back. This is basically what he said…

Grandfather clocks are relatively expensive timepieces. As a result, customers want to make sure the clocks keep accurate time. Since the potential customer is not going to stare at the clock all day long to check how well the clock keeps time, the typical customer uses a short-cut. The short-cut is to compare the time on the clock to a reference point to see if the both have the same exact time.

For example, the customer may compare the time on the grandfather clock to the time on his or her watch. The likelihood that I set that clock to exactly the same time as that customer’s watch is highly unlikely. Since the watch and the clock are not exactly identical in time, one must conclude that at least one of them—if not both—are inaccurate. If the clock is not accurate in displaying the correct time, then there is a feeling that the clock may not be accurate in keeping time.

Another reference point would be to compare all of the grandfather clocks in the room to each other. The likelihood that I was able to set all of the clocks to exactly the same time is also highly unlikely. Since all of the clocks would be slightly different from each other in the time they are displaying, the customer could not have confidence that any of the clocks are accurate. The customer would wonder how all these clocks could be accurate if they are all showing slightly different times. In fact, the more clocks in the room, the more likely they would convince the customer that they were inaccurate.

By contrast, if all of the clocks are set at different and random times that are nowhere near the current time, then there is no reference point. The customer then has no reason to assume that the clocks are inaccurate, because there is no known variance. If fact, given the high price the clocks were being sold for, the customer might assume that the clock is built well enough to keep accurate time.

The irony is that the further away the clocks are from the current time, the more accurate they will appear to the customer. I quickly caught on to what the president was saying and immediately changed the clocks back to random and different times.

The Analogy
Strategic planning and clocks both have something in common. Both are used to better understand the time. Clocks are used to help us understand the current time. Strategic planning is used to help us better understand time in the future. Rather than looking at a clock to tell the time, strategic planners look at research. Strategic planning is all about trying to build a more successful future. Research helps planners understand what kind of future they are trying to build that success in. The better one understands the future environment, the more likely one can design the proper plan for that future.

There is an odd relationship between clocks and the perception of their accuracy. If you have only one clock, when you want to know the time, you just look at it and assume it is accurate, without even thinking about it. However, if you want to know what time it is and you have two clocks to check, your feeling of confidence in their accuracy goes down because they will not be exactly the same. Since you do not know which of the two clocks is accurate, a bit of doubt fills your mind. You are less confident that you know the real time, since you now have more options to choose from—the first clock, the second clock or something in-between. If you add a third clock the confidence goes down even more, because there is even more variability in the times being displayed.

The irony is that the more clocks you check, the more uncertainty you have regarding your confidence in knowing what the exact time is. If you have an entire room of clocks, like the store in the story above, there is a feeling that you will never know the exact time. There are too many different options to choose from. By looking at all of the clocks in the room you may feel reasonably confident that the time is, say, approximately two o’clock in the afternoon. However, you will never know when it is exactly two o’clock.

Adding more clocks will not help the situation. Two rooms of clocks will not give you any more of a sense of the accurate time than one room of clocks. You will know no more than you did before—that the approximate time is around two o’clock.

A similar situation occurs when gathering research for strategic planning. After a certain point, the amount of research you have gathered may start pointing to a rough approximation of where the future may evolve. However, it will never give you an exact knowledge of what the future will look like. There is too much variability in the research—they never say exactly the same thing about the future.

Further research in that area will not increase your accuracy. You will know no more than you did before—the same approximation of what the future may look like. Two rooms full of research will not get you any closer to accuracy than one room of research.

The Principle
There is a phenomenon that often occurs in strategic planning referred to as the “paralysis of analysis.” The phenomenon occurs with people who are afraid to take action until they have a certainty of information. Since there is never certainty about the future, these people procrastinate about taking any action. Instead of implementing a plan, they decide to find additional research. The reasoning is that, if one gets additional information, one can make a make a better decision later. Therefore, the person waits for more research.

However, as we saw above, after a certain point, additional research does not increase the accuracy of your prediction. The additional research just keeps roughly circulating around the same approximate direction. If your plan is to continue searching for research until you reach certainty, you will be researching forever, because certainty will never be found. Hence the term “paralysis of analysis.” The person is paralyzed and unable to take any action towards building a better future because he or she is lost in an endless cycle of gathering and analyzing research. While this person is paralyzed, a competitor may not be, which gives the competitor the opportunity to win a better position in the future than the paralyzed one.

Planning is not an exact science. There is always a little bit of uncertainty. Action will need to be taken with less than perfect information.

Going back to the clock story, most people are able to function and take appropriate action in the current moment even if they know that the clocks in their life are not 100% accurate. As long as one knows that it is approximately two o’clock, then that is good enough to take action. This same attitude needs to be applied to thinking about the future. One needs to be able to take appropriate actions towards the future even if one knows that the research possessed about the future is not 100% accurate.

There are three reasons why any additional research beyond a base level will never achieve greater clarity about knowing the future:

1. Consumers are Unreliable in Predicting What they will do in the Future
2. The Future is Not Destiny
3. By the Time Something is Completely Knowable, it is Too Late to Make a Meaningful Strategic Impact.

These are all discussed below.

1. Consumers are Unreliable in Predicting What they will do in the Future. History has shown that consumers are terrible at predicting what they will do in the future, especially as it relates to activities surrounding situations they have not yet encountered. As a result, research that asks your consumers about what they will do in the future has significant limitations. Sony is proud to point out that no consumer has ever asked for any of the myriad of successful inventions they have introduced. If Sony had waited for research to tell them what to invent, none of those items would have been invented. Rather than doing rooms full of specific consumer research, one needs to rely more on generalized research about the basic needs and desires of your consumers and then make inferences about how your plan would play to those basic needs and desires. Consumer research about the future will never give you the specific answer about what to do. At best, it will tell you what general areas or qualities to focus on.

2. The Future is not Destiny. Because I have spent decades studying the retail industry, I am often asked my opinion regarding future trends in retailing. When asked about a particular retail trend, I often respond by saying, “I will not answer your question until you answer a question for me. How aggressively do you think Wal-Mart will be in adopting this trend or in trying to block this trend?” The point I am trying to make is that the future is not written in stone. The future is not an unalterable destiny to be discovered. Instead, the future is changeable. Wal-Mart is so big and powerful, that the decisions it makes can significantly alter the way a retail trend evolves. If Wal-Mart decides to go after a particular business with a particular format, then the trends surrounding that format will be larger and occur faster than what would have occurred if Wal-Mart had not participated.

The same principle applies to you. Since the actions of you and others can alter how the future unfolds, no amount of research alone will ever be able to accurately predict the future. Rather than spending all of your time analyzing two rooms of research, it is better to have spent some of that time considering how you can use your influence to alter the future. Rather than only looking for ways to discover the future, look for ways to create the future.

If the future were 100% certain, then there would be no way to alter it. If one could not alter the future, then one could not alter his or her level of success in that future. If that is the case, then there is no point in planning. It is precisely because the future is not destiny that makes planning so important. Your strategic plans can alter the course of the future. The better your plans, the better your future.

3. By the Time Something is Completely Knowable, it is Too Late to Make a Meaningful Strategic Impact. Often times, the most successful plans are ones in which your company is the first to make a meaningful impact in a new space. If you wait until the space is completely defined and understood before entering, it is often too late. Starbucks rapidly expanded into the new space it created well before it was completely understood. There were no books describing in detail the types of coffee shops that Starbucks invented. In fact, the conventional wisdom based on the research of the time was that consumers wanted their coffee to be inexpensive and convenient. Starbucks was expensive and time consuming.

If you are waiting for all of the research to be written down in complete detail in a book before taking any action, you will be too late. The only way to get research in that much detail is to wait until the market is already fully developed. By the time it is fully developed, your ability to enter and win in that space is severely limited. It is better to get into a new space early and learn via experimentation rather than wait for all of the facts before taking any action.

With all of that said, this is not an excuse to avoid doing any research. Research is still an important part of any strategic planning process. However, the goal should not be to gather rooms full of data in search of exact information. The goal is to gather enough information to get a general impression of how the future may evolve and then start crafting your actions.

Summary
Strategic planning deals with the future. Just as clocks help us understand the current time, research helps planners understand future time. Gathering rooms full of clocks will not give you certainty of the current time. Neither will rooms full of research give you certainty of the future. There are diminishing returns to time spent getting and analyzing research. Just as you can have too many clocks, you can have too much analysis. Gather enough to get a feeling for the general trends and then supplement the research with activities and experiments. Otherwise, you may fall into the trap of “paralysis of analysis.”

Final Thoughts
General George S. Patton is famous for saying that “A good plan violently executed now is better than a perfect plan executed next week.” The goal of strategic planning is not to create a perfect plan. The goal is to create a better future. This requires more than just planning. It requires doing. Good research and good planning can help you reduce the risk of doing the wrong thing. But if you never get around to doing anything because of the search for the perfect plan, the plans do you no good.