Archive for August, 2012

The Cost of Tolerating the AS-IS State of Business Processes

Posted August 26, 2012 By Fletcher Groves

SAI has been mapping business processes – documenting, analyzing, and improving workflow – in the residential construction industry for almost 15 years.  During that period, we have managed dozens of engagements, across a broad range of geographical markets and builder types, multiples of processes, involving thousands of discrete steps.

It is revealing to simply look at the numbers from all of these engagements, in their totality.

First, understand what the front-end of process mapping entails.

After the background and parameters for a process mapping engagement have been established, the first order of business is always to document and analyze the current state of a builder’s processes, what we call the AS-IS state of the process;  AS-IS establishes the existing state of a process that has to be improved;  after the AS-IS, comes the SHOULD-BE, the redesign and improvement of those processes to a desired, future state.

The mapping of current-state processes fits within the larger picture framed by an in-depth assessment of operating and business performance, but we look at process workflow itself from the following perspectives:  (1) the cycle time, capacity, and flow of the process;  (2) the problems and issues associated with the design of the process, or the manner in which work is being performed;  (3) the process characteristics associated with complexity, repetition, error, and rework;  (4) the percentage of value-added work being performed;  (5) the workflow patterns visible within groups of process activities;  and (6) the comparison of the client’s processes with the same processes in building companies of similar size and orientation.

Every process mapping engagement we have ever conducted for a homebuilding company has mapped the main value delivery process, what we term Prospect-to-Closing;  PTC is the aorta of any homebuilding company, in terms of workflow, value-generation, and expense.  We exclude the sub-process within Prospect-to-Closing known as Start-to-Completion, because early on we determined that, while STC is a process, it is not really process management;  it is actually project portfolio management with embedded routines and sub-processes.

More on this point later.

Typically, we also map support processes and other strategic processes, but not always the same support and strategic processes in every engagement.  That makes every engagement slightly different.  To give this common footing, let’s focus on the workflow that we elect to treat as process management, and specifically on the perspectives we identify as problems and issues, process characteristics, and percentage of value-added work.


During a process mapping engagement, the team identifies all of the problems and issues associated with either the current design of its processes, or the manner in which the work is performed in those processes.

These include a wide range of problems and issues:  errors and resulting rework;  the failure to communicate in a clear, concise, and timely manner;  the inaccurate source, interpretation, and control of data;  situations in which there was no process, or in which established processes were not followed;  delays, tardiness, lack of coordinated effort, and location;  operating and data systems that are ineffective, difficult to use, inaccessible, lack integration, or lack automation;  vendor performance;  the inaccuracy and flow of documentation;  improper or insufficient training;  plus, a number of higher-level problems and issues, such as corporate behavior, delegation of authority and the sharing of responsibility, unaligned objectives, leadership, planning, and policies.

As a ratio of problems/issues to overall process steps, over every engagement we have ever done for homebuilding clients, this fact emerges:  For every two process steps, there is a problem or issue.  The top six categories?  Problems and issues related to process, data, documentation, coordination, communication, and error and rework.


There are characteristics of work – handoffs, approvals, reviews, inspections, repeated steps, documents, files, and reports, notifications, and error loops, etc. – that are activity types associated with value-enabling or non-value-added work.

Consider the presence of these characteristics:

As a ratio of process characteristics to overall process steps, over every engagement we have ever done for homebuilding clients, these facts emerge:  One out of every four steps results in the work of one individual (or department) being handed off to another individual or department.  One out of every six steps requires the completed work of one person or department to be reviewed, inspected, or approved by a different person or department.  One out of every 13 steps involves the work already performed by one person or department being repeated, redone, or re-routed back through the process in order to correct errors or improve output.  One out of every five steps involves the access or use of a separate document, file, or report.


Each step of a process involves a cost in order to provide it – costs that accrue at each step of the process.  The salient question:  Which steps, from the homebuyers’ perspective, contribute – or add – real value above that accrued cost?

The steps in a business process can be classified into one of three types, based on value contribution.  They are either:  (1) activities that are required in order to deliver the output (product or service) the homebuyer is expecting and is willing to pay to receive;  (2) activities that, of themselves, create no value for the homebuyer, but must currently be performed by the builder in order to meet the homebuyer’s expectations and requirements;  or (3) activities that neither add value nor enable the creation of value.

We categorize these activities as either (1) value-added;  (2) value-enabling;  or (3) non-value-added.  The categorization of value takes its cue from the anecdotal evidence of the problems and issues, and the process characteristics.

The value-enabling activities enable the builder to currently perform the value-adding activities;  value-enabling activities are the glue that binds together the value-added activities in the process.  They allow a process to function, but they are also the source of errors and delays – and a significant amount of a company’s cost of doing business.  If the value-enabling activities are simply removed without changing the process, the process will collapse.

The non-value-adding activities are those activities that neither add value nor enable the creation of value.  They produce pointless work – waste, rework, redundancy, unnecessary inspections and approvals, and all of the other activities whose presence would not be missed by either the homebuyer or the builder.

The objective is to redesign the selected processes, in order to optimize (or eliminate) value-enabling activities, and to totally eliminate non-value-adding activities.

Based on the data, it’s a target-rich environment.

As a percentage of the overall process steps, analyzed over every process mapping engagement we have ever done involving homebuilding clients, the value-added portion of activity represents only about 12% of the work;  fully twice as much of the work (25%) is decidedly non-value-adding, and, therefore, pure waste.  The remaining almost two-thirds (63%) is all of the value-enabling work – provided at a cost – that a builder would gladly do without, if it could come up with a better, and more consistent, way of doing the work.

Most of our clients are well-managed homebuilding companies, some of them exceptionally well-structured and well-managed, by the standards with which this industry is judged.  The list includes past and present National Housing Quality (NHQ) Award winners and past winners of the Professional Builder Builder of the Year award.  Certainly, all of them were forward-looking enough in the first place to be concerned with Business Process Improvement.

Nevertheless – the data from these process mapping engagements, anecdotal or otherwise, in its totality, suggests that fully 25% of the work these builders perform adds no value whatsoever;  in all likelihood, the percentage is considerably higher.  While it is true that no two process steps have equal weight – it is a case of varying durations performed by resources with different values – the percentage of non-value-added work has validity in assessing whether a process is either badly-designed or being poorly managed.

If you are a homebuilder, the point is this:  What are such badly-designed, poorly-managed processes costing you?

The most basic – the most universal – proposition of business is this:  The reason for an enterprise’s existence, the way it makes money — is through the value that it delivers to customers and other stakeholders.  That value is only delivered by the work that the enterprise performs, and that work has to be performed in some method of workflow.

The effort to deliver that value costs an enterprise something.

If you recall, the Start-to-Completion sub-process is excluded from a typical process mapping engagement, therefore, the data from that sub-process is not reflected in any of the perspectives described above.  The cost of that part of the effort to deliver value – including all of the very real cost of wasted effort associated with scrap, jobsite conditions, jobsite mistakes, scheduling conflicts, design issues, etc. – are not even considered in the percentage of non-value-added work.  That is because the cost associated with wasted effort in the STC is largely a converted or transformed cost reflected in Cost of Sales;  that cost – a lot of it hidden, because it is commonly passed on in the form of higher pricing from the sub-contractors and suppliers to which almost 100% of the work is outsourced – diminishes Gross Income.

Rather, this discussion is about the consumed cost of non-value-adding effort, reflected in a building company’s Operating Expense.  Whether you choose to view Operating Expense as a productivity issue or as an excess capacity issue – as a consumed cost that should be leveraged (our preference) or as a consumed cost that should be reduced;  as a “more-for-the-same” proposition or as a “same-for-less” proposition – there is clearly a pointlessly-incurred cost associated with wasted effort, with effort that produces no value.

That cost diminishes Net Income.

How much?

John Caulfield, writing in, recently highlighted data from The Shinn Group’s 19th annual Financial and Operations Study, based on information provided by 450 builders that are participants in the Shinn Group of Companies’ Builder Partnerships.  More than two-fifths of the builders surveyed reported Operating Expenses of more than 25% of Revenue (Shinn’s recommended target is less than 25%). The report said the highest-profit builders were those that kept their Operating Expenses under 20% of Revenue.

Out of the Shinn data and the SAI data, let’s consider a favorable scenario:  a hypothetical builder with modest overhead (20%) and a low level of wasted effort (25%).  That makes this builder’s cost of wasted effort five-percent of its Revenue;  for every million dollars of Revenue, the cost is $50,000, a cost that comes right off the builder’s bottom-line.

That’s one matter, if you are a homebuilding company the size of Pulte (a mere $197.5 million per year in lost-but-potentially-recoverable Net Income).  It’s another matter, one that likely strikes closer to home, if you are a homebuilder with Revenue in the $40 to $45 million range – a range, incidentally, bracketed by two of the consistently successful builders that we took through the effort to map their business processes, who’s data is included in our calculations:  Jagoe Homes (Owensboro, KY) and Charter Homes & Neighborhoods (Lancaster, PA).

Let’s be clear:  If you are the size of a Jagoe or a Charter, and unlike Jagoe or Charter, you do nothing to improve the AS-IS – if you do nothing to remove the waste and variation associated with the current state of the processes you depend upon to deliver value to your homebuyers – that bogie is a cool $2.1 million per year.


Part V: "It’s muda."

Posted August 12, 2012 By Fletcher Groves

(excerpted from The Pipeline)

“TOC and Lean both use buffers to protect their processes from variation”, explained the intrepid, results-based consultant.  “As we learned from the Law of Variability Buffering, unless we do something to reduce the variation, every – and, I mean every – production process will find some way to buffer or protect itself from the effects of fluctuation, using some combination of reserve/excess capacity, time, or inventory.

“The difference is in how the buffers in DBR and Lean work.  In TOC, the various types of buffers protect the capacity of the constraint.  In Lean, the buffer is dispersed more widely, because the capacity has been leveled.”

The CFO raised his hand.  “Go ahead”, she said.

“Something you just said triggered a thought”, he said.  “When we were talking awhile back about how we are currently buffering for variation and uncertainty – and how much that variation and uncertainty is costing us – I made the observation that we seemed to have large time and capacity buffers, but not a particularly large inventory buffer.”

“That’s right.  RB Builders doesn’t have a large inventory buffer”,  she replied.  “Why?  Why do you think that is?”

“I have no idea”, he said.

“In part, it is because high levels of work-in-process won’t protect a build-to-order process from variation”, she explained.  “As long as starts are pulled into the system at the rate of closings, the FIFO sequencing generally holds down excess inventory, because each resource can only work in a prescribed, first-in-first-out sequence.

“If homebuilding production was a continuous or single-piece flow process, it could more easily protect itself with any combination of the three buffers, but it’s not continuous or single-piece flow.  As we know, homebuilding is a build-to-order process, and a build-to-order process generally buffers itself with excess capacity and extra time, not with excess inventory.”

“I agree we don’t currently have an inventory buffer”, replied the CFO.  “But, at times, we have, in fact, had high levels of work-in-process, and . . . ”

“ . . . you want to know why that happens”, said the intrepid, results-based consultant.  “Simple.  A build-to-order process can certainly contain excessive levels of inventory, but that inventory is not functioning as a buffer that protects the system from variation, because it’s not a case of the system responding to variation.  Excess inventory either happens – or doesn’t happen – as a result of the internal policies that allow it or prevent it.  It’s detrimental, in the same way all excessive inventory is detrimental.

“In the case of a build-to-order process, higher inventory doesn’t protect the system from variation.  To the contrary, excessive inventory exacerbates the problem, and makes managing production and utilizing production capacity even more difficult.

“It’s muda.”

The chant began in the back of the conference room, low at first, but growing in volume and intensity, until it engulfed the entire room.

“Mu-da!  Mu-da!  Mu-da!”


(excerpted from The Pipeline)

And — homebuilding production is what type of process?”

It’s a build-to-order process”, said the CEO.

Very good”, said the intrepid, results-based consultant.  “That’s the refrain in this little hymn.  Or, maybe it’s a responsive reading.  I’m not sure.  Anyway, whenever you hear the question, that’s the answer. 

Let’s give it a try”, she said, as she wrote on the board.


Theory of Constraints uses what it calls Drum-Buffer-Rope (or DBR) to simply convey the concept of scheduling production according to the capacity of a specific resource”, she said.  “Whatever TOC calls its scheduling method is not important.  Drum-Buffer-Rope is simply the terminology that TOC uses to describe a way of scheduling a production process.  Depending on the type of flow that it wants to create, Lean Production uses a combination of a pacemaker, takt time, pull, kanbans, FIFO, and inventory to achieve essentially the same idea. 

In either case, that same idea involves a process that (1) schedules its production according to a selected pace, (2) agrees on the resource the system will use to maintain that pace, (3) ties the work of the remainder of the resources to the pace-setting resource, and then (4) protects the entire system from variation and uncertainty. 

Lean schedules its pacemaker to meet customer demand or forecast, while TOC schedules more to maximize Throughput, by fully-utilizing the capacity on the most constrained resource — on the bottleneck or the constraint.  Other than that, TOC and Lean essentially agree on this point.   

There is one significant difference”, she said.  “TOC purposely unbalances the capacity of the system, in favor of managing a planned or imposed constraint, and subordinates all of the other non-constraint activities/resources to it.  Lean Production purposely levels, or balances, the capacity of the system by means of heijunka (production leveling), demand leveling, and continuous flow – unless, of course, it’s a build-to-order process, in which case, TOC and Lean at least partially agree, because Lean recommends selecting the bottleneck (or the constraint) as the pacemaker. 

And — homebuilding is what type of process?”

“It’s a build-to-order process.”

Under DBR production scheduling, the “drum” is the resource that sets the pace of production, and every other resource is subordinated to the requirements and needs of the drum resource.  Under TOC, the drum resource is always the constraint, which means that it is the resource that determines the throughput of the entire system.   Previously, we talked about a system being like a chain;  the constraint resource is the chain’s — the system’s — weakest link.  It is the activity/resource that either has the least capacity, has the greatest number of dependencies, or has the most work. 

The pace-setting resource in Lean Production is called the pacemaker”, she explained.  “For the most part, Lean doesn’t much like constraints, and views them as something to be prevented, fixed,or avoided, not something to be managed.  Lean prefers to level production and match it to takt-time. 

So, in Lean, the pacemaker is usually not the constraint — unless, of course, it’s a build-to-order process.  In which case, TOC and Lean essentially agree, because Lean recommends selecting the bottleneck resource — its constraint — as the pacemaker.   

And — homebuilding is what type of process?”

It’s a build-to-order process.”

That’s right”, said the intrepid, results-based consultant.  “Under either TOC or Lean, the entire production schedule is tied to the pace set by this single activity/resource — to the pace set by this pacemaker/drum — whether that resource is the constraint or not. 

Under Theory of Constraints DBR, the mechanism that ties the schedule to the drum is called the “rope”.  The rope is the synchronizing mechanism, and the outcome is synchronized flow.  In Lean, the schedule depends on the type of flow — on whether we have continuous flow, pull or FIFO.  If we have continuous flow, release is not an issue and we don’t need a release mechanism;  if we have pull, the mechanism is something that acts like a kanban. 

As you recall, Lean would prefer continuous flow — “flow where you can, pull where you must”.  If Lean cannot create flow, it will resort to pull;  it just has to have something to pull with.”

Like a rope?”, quipped a superintendent. 

Lean and TOC are both known as pull systems, in part, because of the actions of the rope and kanbans”, she continued.  “For a number of reasons, RB Builders very much intends its production system to be a pull system.”