Thoughts about the strategic discipline, business context, and production perspective necessary for achieving exceptional business and operating performance as a homebuilding company.
Saturday, August 7, 2010
One Homebuilder's Stress Test: Why We Map Processes
It was clear, at least to some. There was trouble if they did not address these issues. With current reality staring them in the face, they did nothing. That was in 2006. In 2008, they filed for Chapter 11 protection.
We know what we are talking about, because we are the homebuilding industry's leading expert, when it comes to the documentation, analysis, measurement, design and redesign, improvement, and management of operating and business processes. It is our tour de force. SAI Consulting has done more work with processes -- and done it longer -- than anyone in the homebuilding industry.
Every consulting engagement we have ever accepted has dealt -- in some way -- with a client's need to structure itself around its critical business processes, and that is because homebuilding operates under a basic proposition:
The only way a homebuilding company stays in business is by creating value that its customers are willing to pay for, the only way that it creates that value is through the work that it performs, and the only way that it performs this work is through processes (and through a project portfolio, because homebuilding is both process management and project management).
But, that does not do justice to process mapping.
In our view, it is far more than documenting, analyzing, measuring, redesigning, and improving workflow. Process mapping connects work to operating performance, and operating performance to business outcomes. In that sense, this type of engagement is the type of stress test that would benefit any homebuilding company.
Consider a small sampling of the forensics in this case:
In 2006, we were engaged by a well-known homebuilding company to help them map their processes. This was a company that had previously won a National Housing Quality award. Just to apply for a NHQ award, let alone win it, this company's processes had been vetted and judged as part of the examination.
During the engagement:
We pointed out the obvious discrepancies between stated operating performance and stated economic returns. We backed our assertions with indisputable production physics, and questioned whether this performance could have possibly occurred. We pointed out the steady decline of the trends in operating performance and business outcomes, to which they were apparently oblivious.
From a process standpoint -- just one example -- we observed that this company had "a very iterative (product) design process exposed to an impulsive/compulsive design mentality", that this was a process with 132 discrete process activities -- involving a minimum of 33 handoffs, 19 reviews, eight approvals, 14 sections of activities where the work of one person or department was subsequently revised or subjected to rework. Furthermore, that none of these 132 activities, by their own admission, could be classified as value-adding, but almost 30% of them were self-classified as completely non-value-added. It was a process that took 12 months to produce a new plan.
We observed -- and proved -- to this company that they had a cycle time of 279 days in their Start-to-Closing process, and that the resulting asset turn could not possibly be the 5.2 times the company was claiming.
We stressed the need to establish a set of operating and business measures as the performance requirements for the new process designs. Yet, in the end, this company was never able to address the question of performance requirements as a comprehensive, connected set of operating and business outcomes. The only performance requirements that were accepted were the cycle time guidelines imposed by the executive group for certain processes. The need for (or importance of) performance requirements did not strike a chord with either the executive group or the process teams. Given the existing level of operating and business performance, we told them that we found "the level of disinterest -- the lack of resolve -- disturbing".
After all, this was a homebuilding company that had produced a Return on (homebuilding) Assets of only 4.7% in 2005; the targeted economic return should have been eight times that rate. Moreover, this was a homebuilding company that six weeks previous had been forced to take the gut-wrenching action of releasing 40 teammates. We pointed out that the real situation was probably worse, that, in all likelihood, an economic return of 4.7% overstated the company's true performance, since cycle times of 250+ days could not possibly result in an asset turn of 5.2. For that matter, it did not seem that the company could have produced an asset turn of 5.2 with a productivity (revenue/employee) measure of only $700,000.
We told this homebuilding company that we had traveled this country, working with homebuilding companies of all shapes, sizes, rationales, and arguments, companies just like theirs, and that processes like theirs were not just badly-designed processes; they were the outcome of flawed thinking on how to best understand and satisfy the requirements and expectations of their chosen market segment, on how to craft a solution that satisfies the requirements of all of their stakeholders -- company owners and teammates, as well as homebuyers.
We told this company that velocity was a lot of what this type of project was all about. It was about finding ways to design/implement better, more productive processes, in order to increase productivity and reduce cycle time. We told them that processes were simply the logical starting point, the first step in the quest toward a "more-for-less" mentality -- more output, more revenue, for the same investment in WIP and production capacity.
We told them that, in their case, perhaps it also needed to be about something else. Perhaps -- just perhaps -- it also needed to be about finding more ways to create and deliver the value their customers were willing to pay for, were willing to pay a premium for, while still finding ways to become faster and more productive.
At the conclusion, we told this company "there is a long road ahead". We reminded them, that when they "embarked on this journey, (they) knew that this project was more the beginning than the end, that it would be the start of an effort that never really ends; the process of continuous improvement means just that -- a continuous process of improvement".
We asked them the same questions we had begun asking every homebuilding company: "Does the world really need one more average homebuilding company? Will "average" performance -- operating, business, or otherwise -- be sufficient to sustain a homebuilding company in the future?"
We said that it was an urgent question, and we thought the answer was "No". We questioned why there were any average homebuilding companies. We said that we suspected the answer was that there were a lot of builders who did not care, a great many who were still living in the "95-05" mindset and for whom the business of homebuilding had been too easy, but that did not explain the large number who wanted to be extraordinary, exceptional homebuilding companies, yet struggled with mediocre customer satisfaction and average operating and business performance.
We told this homebuilding company that they were not an average homebuilding company in intent or reputation, but they were average -- at best -- in terms of performance. We told them that this project was a start. Whether it was a good start, whether it would be sustained, whether it ultimately would produce the results it was intended to produce, was up to them.
That was 2006. In 2008, they filed for Chapter 11 protection.
"Well, someone might as well ask the question", said the intrepid, results-based consultant. "From both a production standpoint and a financial standpoint, at what point does RB Builders breakeven?"
"One of the advantages and benefits of allocating costs on the basis of how they behave in relation to Revenue is the ability to understand and use breakeven analysis", replied the CFO. "But, I have to admit, before you came, we could not have answered that question. The RB Builders Income Statement was prepared according to the NAHB Chart of Accounts, which is to say that it was comparative, compliant . . . and utterly useless. Now, we also produce a Contribution Income Statement.
"According to the 2008 baseline budget, our breakeven point is 155 closings, based on Revenue of about $39 million. However, because of the way the market has deteriorated, the 2008 target budget has a higher production breakeven point; it requires closer to 170 closings, albeit on only slightly higher Revenue ($40.5 million)."
"How did you calculate that?", asked the VP of Construction.
"Let me show you", said the CFO, creating a new data table and adding the data to the first two columns. "This is what we have said, so far. Some parts we do not know yet."
"Breakeven occurs at the point where overhead is completely absorbed", he said, continuing to write as he spoke. "Overhead is absorbed through the generation of Gross Income, which is comprised of the proceeds that we get to keep from each closing. When you are dealing with averages, one way to figure the breakeven point is to take the average sales price of a home, multiply it by the Gross Margin Ratio, and then divide the resulting Gross Income per home into your overhead."
BREAKEVEN = OVERHEAD / (AVG SP X GM%) BREAKEVEN = $8,500,000 / ($250,000 X 22%) = 155
"That gives you the unit breakeven point, in other words, the breakeven point in terms of closings. The unit breakeven point in the 2008 Baseline is 155 closings."
"Is there another way to look at breakeven?", asked the intrepid, results-based consultant.
"Sure", he said, writing on the board. "You can calculate the breakeven point in terms of Revenue. You calculate that by dividing overhead by the Gross Margin Ratio, which, by the way, is basically the same measure as Contribution Margin.
"Take the 2008 Baseline and Target we were just discussing."
"Like I said, about $39 million and $40.5 million, respectively", he said, pointing back to the data table. "Our overhead under both the 2008 Baseline and the 2008 Target is $8,500,000, but the resulting Gross Margins are different, so the breakeven points are different. In this case, the difference in the unit breakeven point is more substantial than the difference in the Revenue breakeven point.
"Overhead is the same thing as Operating Expense, which is comprised of all our indirect, non-variable costs. Overhead -- or Operating Expense -- is the cost of our production capacity.
"That gives you the breakeven number of closings, and the breakeven Revenue. Equally-important -- since we cannot generate all of our closings and all of our revenue at once -- is the breakeven rate.
"In terms of the scenarios we have been discussing, breakeven in our worst case scenario occurs at around 190 closings, which is at about $46 million in Revenue. That is because we are trying to absorb the same amount of overhead with smaller Gross Margins. It requires more closings, which can become a vicious cycle, with a lot of margin pressure. Anyway, in the worst case scenario, the point at which we fully absorb our overhead occurs at a rate of about 16 closings per month. Of course, we have the production capacity to start and close 25 houses a month."
The CFO filled in the missing data for the last two columns.
"Obviously, at only 140 to 150 closings -- 12 closings per month -- and $34.5 million in Revenue, we would be below the breakeven point in both closings and Revenue, which means we would be losing money.
"But, if we can somehow find a way to more fully-utilize our production capacity -- which we will pay to have anyway, unless we cut our overhead -- and somehow find a way to close more homes, albeit at considerably lower margins, we would breakeven at 246 closings and $56.6 million in Revenue. That occurs at 20-21 closings a month, a rate also well below the 25 closings-per-month we have the capacity to produce.
"And, while we are talking about breakeven rates, I hope it reinforces the importance of even-flow production", said the intrepid, results-based consultant, looking at the VP of Construction and VP of Sales. "We cannot be all over the map each month with sales, starts, and closings, and with WIP."
"Is there any kind of housing market in which demand does not respond to lower sales prices?", asked a sales representative. "What happens if we reduce the sales price -- give more concessions, endure lower margins -- and there are insufficient sales? Is there a limit to how far we can drop prices? In other words, what happens if we build it, we drop the ticket prices, and they still do not come?"
"There is always that possibility", said the VP of Sales. "But -- we still have to price to the market. That is not in our control. To the extent that we extract value, with better margins as the outcome, we do have some control. On the other hand, from what I am hearing, the gains from higher productivity are permanent, and the speed/velocity that enables those gains is something we can always control."
The intrepid, results-based consultant turned to the sales representative. "To answer your question, if the situation gets bad enough, higher productivity becomes moot", she said. "RB Builders would have excess -- and probably unusable -- capacity. In that case, higher productivity might not seem as urgent, or as attractive.
"If we cannot make some combination of higher margin and higher velocity work for us, we might have to take RB Builders out of gear, and glide to some sort of safe landing. A controlled crash would be a better description. Like Apollo 13. Forget the moon, just get the Odyssey and her crew home."
"Perhaps", said the CEO. "But, I am determined that we come out of this situation with a level of sustainable competitive separation. I refuse to accept the sacrifice of having endured this much pain without having something to show for it. Higher productivity might not seem as urgent, right now, but it will someday. I want RB Builders to emerge a much faster and more agile homebuilding company.
"Forget "industry best practices". We need to do better than that, because we would be foolish to believe that our competition will forever come from who it comes from now. Someday, the homebuilding industry is going to change, and that change will as likely come from without, as from within."
"Ex Disastrium, Scientia", said the intrepid, results-based consultant, smiling at her adaptation of NASA trivia.
"Learn from adversity, learn from failure, learn from mistakes.
Part IV: "Faced with the prospect of both fewer sales and lower margins"
(Note: "The Pipeline" was written in mid-2007. The fact that it survived the past three years with its relevance intact is, you could argue, mere fortuitous insight, but it is also a testament to the immutable principles of production that it illuminates.)
"Rather than ask where your heads disappeared to, let me just say that what you just described is not the reality with which we are dealing", said the VP of Sales. "This coming year -- 2008 -- is going to be a challenge. So, what happens when we are faced with the prospect of both fewer sales and lower margins? What happens when the market is not going to allow us to use -- to economically leverage, as you like to put it -- all of this new-found production capacity?
"I agree that productivity and Throughput-killing variation is the problem when we are faced with an internal constraint, when we are faced with being our own worst enemy. Under those circumstances, "Max-T" is the right approach.
"But, what happens when we are faced with an external constraint?"
"Give us some idea of what you are talking about", said the intrepid, results-based consultant. "Demand is elastic. How many sales, at what margin?"
"The numbers in the GI Baseline and GI Target are a couple of months old", the VP of Sales replied. "They already reflect the expectation of a deteriorating market. We still might be able to achieve those numbers, but, since then, the market has deteriorated even further. It is precipitous. We are looking at the very real possibility of both fewer homebuyers and lower margins."
"How few and how low?", asked a sales representative.
"During 2007, we closed 200 homes, but we only sold 180 homes", the VP of Sales said. "Not exactly the protective backlog of sales that we want. So, we were already seeing the pressure in the market. Our Revenue was $50 million, our average selling price was $250,000, and our Gross Income Margin was 22%. We kept those numbers in the baseline for 2008. However, the 2008 target is $60 million in Revenue, produced on 250 closings, with an average sales price of $240,000, and a Gross Income Margin of 21%. The higher productivity, higher utilization -- whatever you want to call it -- actually gives us more Gross Income in 2008 than we earned in 2007.
"However, now we are thinking maybe only 140 to 150 sales. We think that the average sales price will still be $240,000, but there will be more concessions. More concessions will result in lower margins, somewhere between 18% and 19%. If that scenario happens, we are looking at Revenue of $34.5 million, and Gross Income of $6.5 million.
"In case you missed it, our indirect, non-variable cost is budgeted at $8.5 million.
"That is an operating loss of $2 million, and, frankly, we do not know where the bottom of this recession is. It could get much worse."
"Nevertheless, we need to talk about what we know now. What would you do?", the intrepid, results-based consultant asked the VP of Sales. Glancing toward the CEO, silently with her eyes, she said, "Just let him answer."
"We cannot afford to lose money. But -- I cannot look at the people in this room, and suggest that we fire people whom we have developed and whom we care about, either", said the VP of Sales. "Certainly not as our first resort.
"The market may not turn out to be this bad, but here is what I would do:
"The current job schedule says that we should be able to build our houses in an average of 120 days, and we have been given 100 units of work-in-process to produce as many closings as we can", he said. "That calculates to 300 closings. Despite the challenges of the market, we need to find a way to get those 300 closings.
"Let me qualify part of that statement. The closings are important, but we need to produce as much Gross Income as we can, because that is all we get to keep from whatever Revenue we generate from those 300 closings.
"From a production standpoint, someone else will need to figure out how to beat 60 days out of the current cycle time. From a sales and marketing standpoint, to sell 300 homes, I believe we will need to drop our prices to an average of $230,000. I know, I know. It is a difficult decision. It is $10,000 below the target, and $20,000 below the baseline.
"Our margins would suffer, dropping to 15%, on average. We would need to become much more intuitive and instinctive in our adjustments, and learn to make decisions as fast and as frequently as necessary. We will have to fight for every sale. But -- even with the lower margins -- if we manage to sell, build, and close 300 homes, our Gross Income would be $10.4 million, produced on Revenue of $69 million. Our Net Income Margin would be less than 3%, but we would be profitable.
"Of course, like I said, it could get worse. Much worse.
"Could we find ways to extract more value, and therefore earn higher Gross Income Margins and generate additional Gross Income on every dollar of Revenue? I think so. Could we get our cycle times down to 90 days, and close 400 homes on the same amount of production capacity? Maybe. That is up to us. I am not sure that is what we would want to do right now. The resulting higher production would have even further implications on prices and margins.
"Still, I am starting to realize some things: First, higher velocity can overcome lower margins. Second, productivity gains are permanent."
"In 2008, higher productivity is a case of survival. And -- it may not be enough, if things get a lot worse", said the intrepid, results-based consultant, writing as she spoke. "In the future, the ability to produce more -- more closings, more Revenue, more Gross Income -- on a finite and controlled amount of work-in-process and overhead will be one of the keys to sustainable competitive separation.
BREAKEVEN ANALYSIS
"This discussion raises a question", she said, pointing to the board. "From both a production standpoint and a financial standpoint, at what point does RB Builders breakeven?
"Knowing the answer to that question gives you more insight than you can imagine."
Part III: The Cost of Variation and Long Cycle Time
(Excerpted from "The Pipeline")
"So -- how much do you think all of this variation is costing your company?", asked the intrepid, results-based consultant. "Let me ask it two different ways: How much is the lack of productivity costing you? How much is chronically-long cycle time costing you?
"Variation, productivity, cycle time. They are all connected."
"Our cycle time is 180 days, and we are doing it with 200 closings produced on 100 units of work-in-process", said the CFO. "Without getting into an explanation of statistics, we have about a 50% variation from the standard of 120 days and a 100% variation from the cycle time we instinctively believe we should be achieving, which is 90 days.
"As you have taught us, a system will protect itself from variation and uncertainty with some combination of longer cycle time, higher work-in-process, and excess/unused capacity. Our cycle time is very long, which translates into a large buffer. Our work-in-process is where we planned it to be, and only 25% higher than the lowest amount it could possibly be, which is 80 units at a cycle time of 120 days. So, I do not think this is a case of RB Builders having a buffer of additional work-in-process.
"Here is the kicker:
"If we honestly believe we can produce 300 closings with our current production capacity, then we have a huge buffer of excess/unused capacity, which calculates into an overall utilization rate of 67%. In essence, we waste one-third of our capacity. As you have pointed out, that is hardly the picture of high productivity. The inability to utilize our capacity translates into significantly fewer closings. We are paying for the capacity and the work-in-process to produce 300 closings, but we only closed 200 homes. That is a gap of 100 closings.
"In terms of what all of this is costing us, there is clearly a cost associated with excess work-in-process and unused production capacity", the CFO continued. "The additional, "beyond-necessary" work-in-process certainly makes us a bigger company than we need to be, and the excess/unused production capacity alone costs us over $2,800,000 a year.
"But -- I do not think it is about cost. I think it is about opportunity. Unless we opt for cost-cutting and reducing overhead -- a "same-for-less" proposition -- then, I would say that it is "costing" us the opportunity of all the Gross Income on those 100 closings we missed in 2007. Our Gross Income Margin was 22%. We had $50 million in Revenue. If you divide that by the 200 closings that we achieved, the average sales price was $250,000. From there on out, the math is pretty simple."
The CFO walked up to the erasable board, and wrote:
$250,000 X 22% X 100 = $5,500,000
"We gave up $5,500,000 in Gross Income."
The conference room was completely silent. Everyone was aware of the Gross Income Baseline, Target, and Reserve, and the impact an additional $5,500,000 would have on the payout of Gross Income Milestones under the new results-based performance compensation plan RB Builders had just enacted.
The silence was broken by the words of the CEO.
"No. That calculation does not even scratch the surface", he said. "What is the real cost?"
He walked to the front of the conference room.
"If all that this excessive variation, lack of productivity, and longer-than-necessary cycle time cost us was $5,500,000 in Gross Income, that would be bad enough", bristled the CEO. "But -- this is also $5,500,000 that would have dropped straight to our bottom-line, in the form of additional Net Income.
"In terms of the cost of our production capacity, utilizing it would have cost us nothing -- zip, nada, zero. It is non-variable cost. It is overhead. We already paid for it.
"I am not finished with this issue."
The CEO turned to the erasable board, picked up a marker, and added two more rows to the data table:
2005: WIP = 100 CLOSINGS = 225 CYCLE TIME = 160
2007: WIP = 100 CLOSINGS = 200 CYCLE TIME = 180
2008: WIP = 100 CLOSINGS = 240 CYCLE TIME = 150
WIP = 80 CLOSINGS = 240 CYCLE TIME = 120
WIP = 100 CLOSINGS = 300 CYCLE TIME = 120
WIP = ? CLOSINGS = ? CYCLE TIME = 90
WIP = ? CLOSINGS = ? CYCLE TIME = 90
Turning and nodding at the data table on the board, he asked, "Forget the 180 day cycle time we have now. What do you think RB Builders would look like at a cycle time of 90 days, instead of the 120 days specified under our job schedules?
"It is a rhetorical question. For all of about five more seconds."
The CEO made a quick calculation and filled in the Closing and WIP data that was missing from the table.
WIP = 60 CLOSINGS = 240 CYCLE TIME = 90
WIP = 100 CLOSINGS = 400 CYCLE TIME = 90
"We could go with virtually any combination -- any strategy or tactic -- of higher Throughput and lower work-in-process to achieve a 90 day average cycle time, but say that we elect to go with a tactic of "Max-T", a tactic of generating maximum Throughput with a planned, finite, and controlled level of work-in-process. Forget 200 closings on WIP of 100 units, which would be the current cycle time of 180 days. Forget 300 closings on WIP of 100 units, which would be the cycle time specified in our construction schedules. What about 90 days? Although it is a valid option, set aside the idea of making RB Builders a smaller company -- 240 closings, but reducing WIP to only 60 units. Focus on Max-T. What about 400 closings on WIP of 100 units?
"What happens?"
The CFO handled the question. "Well, if our Gross Income Margin remains the same -- which it likely would not, because of price elasticity of supply and demand -- then we would generate an additional $11 million in Gross Income, every penny of which, as you point out, drops straight to our bottom-line in the form of additional Net Income.
"Which means that our unwillingness -- or our inability -- to do anything about the current level of variation that drives our long cycle times is costing RB Builders as much as $11 million in Net Income every year. That is a lot of money."
"Yeah", said the CEO. "Even for a company that would then have $100 million in Revenue. I want everyone to be clear. Our 2008 baseline is $50 million in Revenue, from which we expect to produce $11 million in Gross income and $2,500,000 in Net Income. Our 2008 target is $60 million, producing $12.5 million in Gross Income and $3.4 million in Net Income. As all of you realize, all of our very-considerable bonuses are tied to the Gross Income Reserve that represents the difference between the baseline and the target.
"By appearance, this kind of performance would be over-the-top. But -- if we pull it off -- it means that instead of progressively splitting a GI Reserve of $1,500,000 three-ways between our owners, Retained Earnings, and all of us, dropping our cycle time from 180 days to 90 days means we get to split a GI Reserve of $12.5 million.
"And -- you know what?
"Despite the fact that its Revenue had doubled, RB Builders would be the same size company. It would have the same amount of WIP, the same overhead, the same working capital requirement, the same level of debt."
The Intrepid, Results-Based Consultant says "I Do"
The intrepid, results-based consultant gazed lazily at the white sand and the water beyond, and thought, it would definitely take a lot of this to kill you. She sighed contentedly, recalling the last 72 hours.
The planning had been thoroughly intense. Or, perhaps it had just been intensely thorough, she wasn't sure. Whatever. Her wedding had come together just the way she had dreamed. The June weather in Northeast Florida had cooperated, for the most part. The boys had enjoyed the ritual marriage event known as Wedding Party Golf. The rehearsal had been fun. The flowers had miraculously come together. The church was beautiful. The service was reverent, and rich with meaning and promise. The music was awesome. There had been tears of laughter, tears of sentiment, and tears of complete joy. The reception had been perfect. Bob Cummings and the Reflections were everything they were expected to be.
Her sister had been a flawless Maid of Honor. Her Mom had been the model of composure, beauty, and grace. Her new in-laws had been the picture of gracious acceptance and support.
Her Dad had defied all odds, by managing to hold himself together.
The intrepid, results-based consultant was undone by all that her marriage represented. Two families joined together. Life-long friendships, some of forty-plus years, honored. Four generations of fraternity brothers, groomsmen, sorority sisters, bridesmaids, friends, and extended members of both families who had traveled thousands of miles, just to be there. A cloud of witnesses that stood -- and would continue to stand, collectively and individually -- with them, by them, and for them, for the rest of their lives. There was the pang of absence for those who had gone before; the ones they always loved, still missed, and would never forgot, but whom they would one day joyfully see again.
It was the promise, challenge, and adventure of a life to be built, and a lifetime to be spent, together. Permanently. Inseparably.
For sure, there would be good times and bad times in the years to come. There would be easy times and hard times. There would be gains and losses. There would be joy and sorrow. There would laughter and tears. There would be planning; there would be spontaneous-ness. There would be children. There would be careers and career changes. There would be changes in priorities and focus. There would be memories made. There would be legacies formed.
There would be times when they knew each other's thoughts; there would be times when they did not think they were from the same planet.
There would be mistakes. There would be grace. There would be forgiveness offered, and forgiveness accepted.
And, through it all, there would be love, honor, and respect.
The intrepid, results-based consultant turned her head to the side, pushed her sunglasses to the end of her nose, and looked at the handsome young man reclined on the beach beside her.
She smiled, and thought to herself, "My Dad would think this is "utterly cool"".
Part II: "Quite the Poster Child for Your Lack of Productivity"
(Inspired and excerpted from "The Pipeline")
One of the sales representatives looked at the superintendent, and just laughed. "Which one of the Productionally-Transmitted Diseases would you like to have? PTD, she owns you."
The intrepid, results-based consultant smiled politely. "In a previous session", she said, "someone said RB Builders closed 200 homes in 2007, on an average work-in-process of 100 houses, which is also the baseline for 2008. But, everyone also agreed that the system should be capable of producing 240 closings on 100 units of work-in-process. Later, someone else mentioned that the building schedules averaged 120 days."
Moving to the erasable board, she wrote the following data in a table:
2005: WIP = 100 CLOSINGS = 225 CYCLE TIME = ?
2007: WIP = 100 CLOSINGS = 200 CYCLE TIME = ?
2008: WIP = 100 CLOSINGS = 240 CYCLE TIME = ?
"Someone calculate RB Builders' cycle time", she said. "How many days?"
After a minute, the sales representatives looked up from her calculator, and said, "If I am doing this right, I calculate that, in 2005, our cycle time was 160 days. In 2007, it was 180 days. And, for 2008, we are targeting 150 days."
The intrepid, results-based consultant completed the cycle time column with the calculated cycle times, but added two more rows with identical values.
2005: WIP = 100 CLOSINGS = 225 CYCLE TIME = 160
2007: WIP = 100 CLOSINGS = 200 CYCLE TIME = 180
2008: WIP = 100 CLOSINGS = 240 CYCLE TIME = 150
WIP = ? CLOSINGS = ? CYCLE TIME = 120
WIP = ? CLOSINGS = ? CYCLE TIME = 120
"Okay", she said. "Tell me what your production system looks like with a cycle time of 120 days."
"There are two ways to look at it", said the superintendent. "We could be closing 240 homes with 80 units of work-in-process. Or -- we could be closing 300 homes with 100 units of work-in-process."
The intrepid, results-based consultant added the new calculations.
2005: WIP = 100 CLOSINGS = 225 CYCLE TIME = 160
2007: WIP = 100 CLOSINGS = 200 CYCLE TIME = 180
2008: WIP = 100 CLOSINGS = 240 CYCLE TIME = 150
WIP = 80 CLOSINGS = 240 CYCLE TIME = 120
WIP = 100 CLOSINGS = 300 CYCLE TIME = 120
"Remember the earlier discussions on margin and velocity?", she asked. "This is where that thinking works. There are two ways RB Builders can increase the amount of Throughput -- the amount of Gross Income -- it generates. Margin is how much money we make on every home we close, and velocity is about how many homes we can build and close in a period."
On the erasable board, she wrote:
MARGIN VELOCITY
"Most of the time, there are improvement opportunities that permit us to attack margin and velocity simultaneously", she said. "Much like the DuPont formula shows when calculating economic return (Return on Assets), Gross Income is a composite of both margin and velocity. We want the best blend of margin and velocity. They do not often conflict, even though there are sometimes tradeoffs, and times when we might be better served focusing more on the one than the other.
"On occasion, we are forced to make a choice on where to focus, more when we are facing an external (market) constraint than when we are facing an internal (production) constraint. What's the difference? Where was the constraint in 2004-2005? The constraint was in your production system. It was an internal constraint.
"The velocity part of the choice decision lies in how well RB Builders is utilizing its true production capacity", said the intrepid, results-based consultant. "The margin part is determined by the condition of the housing market, and whether that market is going to allow us to use -- to economically leverage -- that capacity. We can control truly-variable direct costs, and we can extract more value, but -- at the end of the day -- the market dictates the price you get for a house.
"And, a lot of people are thinking 2008 might be that kind of year -- imagine -- barely a year removed from the recent, final, halcyon days of the "Age of Homebuilder Entitlement"", she said.
"The point is, we have to learn to manage this relationship. We have to find the best blend of margin and velocity, the composite that generates the greatest amount of Throughput. We have to generate the greatest amount of Gross Income that we can, given the reality of our playing field, given the parameters imposed by the market."
2005: WIP = 100 CLOSINGS = 225 CYCLE TIME = 160
2007: WIP = 100 CLOSINGS = 200 CYCLE TIME = 180
2008: WIP = 100 CLOSINGS = 240 CYCLE TIME = 150
WIP = 80 CLOSINGS = 240 CYCLE TIME = 120
WIP = 100 CLOSINGS = 300 CYCLE TIME = 120
"Back to the issue of variation and uncertainty", she said, gesturing towards the data table, and then looking toward the VP of Construction. "Earlier, you noted a widely-accepted sense that RB Builders should be capable of building every house in less than 90 days.
"Not 180 days, not 150 days, not 120 days", she said, pointing to each number.
"In 90 days."
She wrote the following questions on the board:
WHERE IS VARIATION BEING BUFFERED? WHAT DOES VARIATION COST?
"Variation is the deviation from the standard. Variation can be applied to duration. It can be applied to cycle time. The standard cycle time is 120 days, and some of you think it should be 90 days. However, it takes you anywhere between 150 to 180 days to build your houses. Quite a bit of deviation from the standard, I should say.
"Your cycle time is quite the poster child for your lack of productivity."
The intrepid, results-based consultant gazed intently around the room, and then asked, "Considering both your gut-instinct and your interpretation of the data in the table, exactly how and where do you think RB Builders' production system is buffering itself -- protecting itself -- from all of this variation and uncertainty?
"And -- how much do you think this variation is costing your company?"
Part I: Waste, Variation, and Other Productionally-Transmitted Diseases
(Excerpted from "The Pipeline")
"In this session, we are just going to talk about how we better manage whatever it is that is preventing the system from producing more with what it has. First, I want to talk about variation and uncertainty", said the intrepid, results-based consultant.
"It was mentioned earlier that variation is a form of waste, because, to the extent that it does not add value, variation is wasteful. That thinking largely comes from the Toyota Production System and the Lean Production methodology that TPS later spawned. But -- Taiichi Ohno did not include variation as a form of waste (muda) in the TPS, opting instead to use the term for unevenness (mura), and associate it with the separate principle of stability; production processes need stability, and variation causes instability.
"Interestingly, in both Lean Production and the TPS, the principle of stability sits between the principle of waste and the principle of standardization, which includes the concept of standards, visual management, and problem-consciousness, which are linked to PDCA problem-solving.
"So, there is that distinction that would seem to differentiate variation from other forms of waste", she said.
"I think there is a reason Mr. Ohno chose to not place waste and variation in the same category. There is no possibility of having either zero waste or zero variation; the goal is to reduce waste and reduce variation, which is achievable. However, given their inherent characteristics, a desire to eliminate waste is more reasonable than a desire to eliminate variation."
Turning to the board, the intrepid, results-based consultant wrote:
DISTINCTIONS: WASTE V. VARIATION PRODUCTION PHYSICS: LAW OF VARIABILITY BUFFERING
"Then -- beyond that distinction --there is the series of basic laws of production physics that directly or indirectly form our understanding of variation. Five laws, to be exact, but two in particular", she said. "The first law, called the Law of Variability, states that higher levels of variation degrade the performance of the production system.
"The second law, the Law of Variability Buffering, says that variation will always be buffered by some combination of inventory, capacity utilization, or time -- always through a combination of higher work-in-process, excess/unused capacity, or longer durations."
"So", she continued, "If all RB Builders does is attack variation by attacking waste (in the form of errors, rework, etc), and it fails to directly attack the variation that causes instability, then it will have to live with a production system that protects itself with some combination -- buffers itself with some level -- of additional work-in-process, longer-than-necessary cycle times, or wasted capacity. Our production system will default to longer durations. Time is the self-determining buffer, the "buffer of last resort", so to speak. If we do nothing about variation, yet limit work-in-process and capacity, the result will be long cycle times. Guaranteed.
"Buffers -- high levels of work-in-process, long cycle times, and unused capacity -- allow the system to compensate for variation, but, regardless of the combination in which they occur, they all result in lost Throughput.
"And -- the true cost of variation is the financial throughput -- the Gross Income -- that RB Builders surrenders to that variation.
"Taken together, you begin to get a sense that variability is a very big deal", the intrepid, results-based consultant continued. "Jack Welch used to say, "Variation is evil." Some variation and uncertainty is natural, and some of it is necessary and planned. But, the instability that variation and uncertainty cause is a decidedly evil form of waste.
"In process religion, the gods of production will not be mocked", she said. "If all we ever do is attack variation by attacking waste, in the form of errors, rework, etc., and we fail to also directly attack the variation that causes instability, then we will have to live with a production system that protects itself with some combination of too much work-in-process, long cycle times, or reduced throughput."
"But -- isn't protection a good thing?", deadpanned a superintendent. "Shouldn't we be practicing safe production?"
"Yes, protection would be a good idea", responded the intrepid, results-based consultant, equally deadpan. "Especially for boys like you, who should be worried about contracting a PTD.
"Let's put it this way", she said. "Which one of the Productionally-Transmitted Diseases would you like to have? Exactly which combination of longer-than-necessary cycle times, higher-than-necessary levels of work-in-process, and lower-than-possible rates of throughput (because of excess and unused capacity) do you really want to contract? I hear they are all really painful.
"Some level of variation and uncertainty is natural, inevitable, and unavoidable", she said. "We have to buffer that. In addition, some variation is necessary, just to protect ourselves in the marketplace. For example, we don't offer only one floorplan and elevation. Lastly -- whenever we are protecting the output of the system from variation and uncertainty -- some level of protective capacity or buffering is needed.
"However -- protecting a system from variation comes at a cost, and to the extent that the variation that necessitates the buffering is unnecessary, avoidable, excessive, or uncontrollable, it is a very bad thing."
