Tag Archives: system integration

Cross Department Committees

From the Ask Tom mailbag –

Many times when there is an issue that affects more than one department in the company, we assign committees involving members from each department to solve them. While this seems nice from a cultural standpoint, it seems strange that we would ask people in an S-II or S-III role, to solve issues that span multiple departments, typically an S-IV function. I recently experienced this myself where I established a committee, set a clear direction (I thought), and checked in occasionally. The end result was I now had a group who had reached a consensus, but it was the wrong one! We are still able to move forward and correct it slowly, but it feels like we wasted effort. What’s the right answer to this? Be more involved? Assign another committee leader with level 4 capabilities? Provide better direction? Make a larger committee?

Quick review on general accountabilities at levels of work.

  • S-I – Production
  • S-II – Making sure production gets done, coordination and implementation.
  • S-III – System work, designing, creating, monitoring and improving a single serial system (critical path)
  • S-IV – Multi-system integration

So, your intuition is correct that, where multiple departments are involved with either output or impact, department integration is appropriate.

Your question – Be more involved? Assign someone with S-IV capability? Provide better direction?
Answer – Yes.

In any managerial role, with team members one level of work below, the manager cannot simply call the meeting and then not show up. Undirected, the team will make the decision or solve the problem at their level of context. Each level of work understands its decisions and problems from their level of context. That context is measured in timespan.

Problems or decisions involving multiple departments generally require looking at longer timespan outputs, more correctly, longer timespan throughputs. A single department is usually heads-down, internally focused on efficient output. Multiple department throughput typically looks at two things. Does the efficient output of one department provide the correct input for the next department as work moves sideways through the organization?

  • Does the output of marketing (leads) provide the correct input for sales?
  • Does the output of sales include all the data necessary agreements for proper project management?
  • Does the output of project management provide all the accurate data necessary for operations?
  • Does the output of operations provide all the necessary checkpoints for quality control?

Multiple department integration also requires a look at the output capacity of each department as they sit next to their neighbor department. Is is possible for sales to sell so many contracts that it outstrips the capacity of operations to produce? A lower timespan focus might say we just need to communicate better. A longer timespan focus (throughput) will realize that no communication solution will fix a capacity issue.

So, yes, the manager has to be more involved, include another team member at S-IV, provide better direction on the requirements of any solution. A larger committee might actually be counter-productive if it contains team members at the wrong level of the problem. I offer these same guidelines as those of a couple of days ago.

  • What is the problem?
  • What is the cause of the problem?
  • What are the alternative solutions?
  • What is the best solution?
  • How will we test the solution to make sure it solves the problem?

Service System Capacity

From the Ask Tom mailbag – Related to Integration is a Fancy Word. The illustrative example described an imbalance of systems in a manufacturing model, where there is a build-up of finished goods inventory (unsold).

Can you provide an example of an anemic sales function in a service industry. What would you get instead of an inflated inventory?

Thanks for the question. In a service industry, say you have 20 trucks, 20 technicians, optimized to average three service calls per day. The daily average capacity is 60 calls.

If sales only sells 55 average daily calls, you have excess unsold capacity of 5 service calls. You may not even notice. If the average drops to 50, you may begin to notice, and so do your service technicians. How long does it take a service technician to complete two assigned service calls vs three assigned service calls? The answer is 8 hours, no matter which. Parkinson’s law – work expands to the time allotted.

This is functional integration work, to monitor the capacity of each function, to make sure the impact of one function doesn’t outstrip or adversely impact the capacity of its neighboring functions.

In this service example, the math is pretty easy, 20 technicians x 3 calls = 60. Sometimes the service work has more subtle variations where the math is not so clean. That’s why system capacity makes for fascinating study.

Integration is a Fancy Word – Part III

The purpose of system and sub-system integration is not to get rid of our silos, but to integrate them together. The second issue in this integration has to do with individual system capacity and total system throughput.

As organizations grow, there is constant pressure on efficiency (lean, six sigma, MUDA), but as the internal systems multiply, efficient as they are, they begin to get in each other’s way. It is not enough to have a collection of perfectly efficient systems, the organization now has to look at total system throughput. Capacity output and constraints of each system come into play.

Is it possible for Sales to write so many orders, that it outstrips the capacity of Operations to complete those orders? Unfilled sales orders become back-orders. Unfilled back-orders become cancelled orders. Customers go to competitors. What’s the problem? Both Sales and Operations are running full-tilt within the constraints of their function, but one function is outstripping the capacity of the other.

Let’s flip this around. Our Operations function has the latest, greatest state-of-the-art equipment, a cracker-jack operations team and the capacity to crank out finished goods like there is no tomorrow. Yet, if our Sales function is somewhat anemic, not that they are writing no sales orders, but certainly not selling everything that gets produced, what happens to the unsold finished goods? Into inventory they go, stacked in the warehouse. Until the warehouse gets full, then what do we do? We get another warehouse. What is happening to the cost-of-goods-sold?

This second issue of system integration is optimizing the capacity of each function as it sits next to its neighboring functions. There are dependencies, inter-dependencies, constraints, contingencies and bottlenecks that govern total system throughput. It does no good to write sales orders for products and services that cannot be filled.

Integration is a Fancy Word – Part II

Even small organizations assemble systems and sub-systems, and face the same system integration issues as large organizations. Remember, this is not an exercise to eliminate silos, but to integrate them together.

The first integration issue has to do with work as it travels sideways through the organization, from one function to the next – marketing – sales – project management (account management) – operations – quality assurance – research and development – accounting – human resources.

The first issue is to establish the quality standard of the work output as it travels from one function to the next. What does sales always say about the leads that marketing gives them?
“The contact person you gave me hasn’t worked at the company for three years.”
“The telephone number you gave me is a fax machine.”
“The email address you gave me is misspelled and bounces when I send to it.”

What’s the problem? Marketing is proud of the quantity of leads passed to sales, but, the leads miss the mark. The first step to integration is to establish and enforce the quality standard of work output.

In construction, there is a critical work hand-off between the estimating function and project management function. If there is a defect in that hand-off, the defect will be embedded in the project until it surfaces in either operations or quality inspection. The integration of those two functions (estimating and project management) demands the hand-off meeting has a hard agenda with detailed checklists. Problems identified in this hand-off meeting are easier to correct on paper than later, after steel and concrete.

The second issue related to integration has to do with capacity, subject of our next post.

Integration is a Fancy Word – Part I

Each level in Elliott’s level-of-work schema corresponds to a macro organizational function.
S-I corresponds to production work, timespan 1 day to 3 months.
S-II corresponds to supervisory work, to make sure production gets done, timespan 3 to 12 months.
S-III corresponds to system work (single serial system), timespan 1-2 years.
S-IV corresponds to system integration work (multiple systems and sub-systems), timespan 2-5 years.

Integration is a fancy word, what does it mean?

As the organization matures, grows larger and more complex, distinct functions emerge. These started as single roles, but grew into teams. Here is the quick list –
-Project Management (or Account Management)
-Quality Assurance (QA/QC)
-Research and Development
-Human Resources
There are a ton more, depending on the business model, but you get the drift.

Each of these functions begins internally focused, mostly because we said so. We told each function they had to efficient, internally profitable, no waste, prudent use of internal resources. We told them to be internally focused.

As we stack more functions next to each other, we can see the problem we created. The problem presents itself as a communication breakdown, sometimes a personality conflict. Each function competes for budget, prestige and managerial attention. This is the silo effect.

And we have all been told to get rid of our silos. Not so. We put those silos in place for very specific reasons. It is not a matter of getting rid of our silos, it is a matter of integrating them together.

Integration is a fancy word. What does it mean?

Multi-system Integration

From the Ask Tom mailbag –

When you talk about the COO role, you use the words multi-system integration to describe its purpose. I understand silos, and you said we put the silos there for a reason. You said not to get rid of your silos, but to integrate them together. What does that mean?

Yes, as companies move from Adolescence to Prime, the silos will kill them without system integration. The COO role will necessarily focus on operations confronting two main system integration issues.

  1. What is the performance standard of the work output from each function (department, system, silo) as work moves to the next function? Each function has marching orders (from us) to be efficient, internally profitable. This necessitates an internal focus to the function. Many critical handoffs occur as work moves from sales through project management into operations. When there is a defect in the handoff, we may not discover the problem until we have poured concrete around steel. It is much easier to unwind in the handoff stage than the jackhammer stage.
  2. The other system issue relates to each function’s capacity for output. Without someone paying attention, easy for one system to outstrip the capacity of its neighboring system. This may manifest as a breakdown in communication, when the underlying cause is two internally focused systems, heads down not paying attention to the other functions around. Discussions of single system capacity rarely emerge below S-IV, much less the impact of one system onto another system.

One of the critical functions for the COO is to calculate the company’s operational (ops) capacity as operations is most likely to be the constraint in the midst of the other systems.

Is the COO Irrelevant?

From the Ask Tom mailbag –

I read your book, Outbound Air, again.

And I was thinking there might be a conflict with trendier/newer business models. A lot of companies seem to be pushing flatter structures and mixed function work-groups. So there isn’t really a role for COOs as say, the CEO’s internal quarterback

My understanding is that COOs exist to corral the various functions i.e. highest timespan
while the CEO is dealing with strategy, major threats etc.

So does the timespan model change for these flatter-structured businesses where the COO is supposedly irrelevant?

Elliott’s response to a similar question goes like this –
“I hear these things, and I just have to ask, who is kidding whom?” It is not that the role of the COO disappears, but it is certainly different.

Small Organizations
First, many organizations (small ones) are not level (V) organizations in the first place. Indeed, many companies are level (III) organizations, so they have production, supervision and a CEO, who really plays the role of a level (III) manager. Nothing wrong with this small company, the CEO can make a wealthy living out of it.

Growing Organizations
As the company grows, the level of work will necessarily increase to level (IV). There are multiple functions (systems) inside the company that must be integrated together. Again, the CEO in a level (IV) company will play the role of the integrator. In a larger, more mature company, this would be the role of the COO.

Maturing Organizations
In a level (V) company, the CEO must leave the integration role and truly focus on strategy. Without an effective COO at level (IV), the CEO will necessarily be dragged down into the weeds (back into integration activity). And, as long as the CEO is doing work at level (IV), the company will not grow, likely grow and contract in fits and starts, never effectively integrating their multiple systems. Yes, it is possible to have a dysfunctional level (IV) organization.

Digital Technology
Over the past two or three decades, technology arrived. Indeed, computer systems (note the word system) supplant many level (III) functions. MRP and ERP software systems, in their algorithms, require very specific steps in specific sequences, level (IV). The algorithms were created by some very smart teams who created systems and system integration in a variety of disciplines.

However, with effective technology implementation, the managerial work changed. So, let me pose this question. If we have a technology platform that serves to move data between multiple functions in the company, integrating those functions together, a level (IV) role, then what is the work of the COO?

Here is a hint. Work is decision making and problem solving. In the presence of an effective ERP system, what decisions are left to be made and what problems are left to be solved by the COO? There is an answer to that question.

Your thoughts? -Tom

The Struggle for Emerging S-IV

From the Ask Tom mailbag –

It took a long time, but our company has grown. Our business model is a distributor, it’s all about supply chain for our customers. Because our business model is driven by the logistics of incoming and outgoing material supply, we recently promoted our warehouse manager to VP-Inventory Control. For us, it was more than just a change of role title. Our warehouse manager took us through re-binning our inventory, bar coding SKUs, RFIDs on serialized product. He is a really bright guy. But his promotion to VP-Inventory Control seems to have gone to his head. With his new-found power, he has emerged as a prima-donna. In our executive team meetings, he believes that inventory control should be the deciding factor in every business decision for the company. If he keeps this up, he is going to get fired.

Indeed, the move from a Stratum III (S-III) inventory manager to an (S-IV) is a dramatic change in level of work.

  • S-III – System (creates the system, monitors the system and improves the system)
  • S-IV – Integration of multiple systems and sub-sytems (attention to dependent systems, interdependent systems, contingent systems and bottlenecks)

The focus at S-III system level is internal. We demand each of our systems be efficient, profitably leveraging its resources for maximum output. Your inventory manager did just that with a bin system, bar codes and RFIDs. Kudos.

The focus at S-IV is integration. With an internal focus on inventory management, his new role is to assist in the integration of inventory with all the other systems in the company. It is no longer a matter of profitably leveraging resources for maximum output, but optimizing output with the other systems in the company. It is a matter of how one system’s output (reinforcing system) is impacted by another system’s output (balancing system).

This requires the focus for the new S-IV to transition from internal to external. You don’t have a prima donna personality conflict. You have not clearly defined and communicated the new role, nor its differences from the prior role.

You also skipped a step. How did you know if the inventory manager was ready for these new accountabilities? You didn’t. You blindly promoted and now you have a bit of a chocolate mess. The step you missed, prior to the promotion, was assigning S-IV project work, coaching and evaluating the output. Team members should NEVER get a promotion. They earn promotions by successful completion of project work similar or identical to the work in their new role. -Tom

Finger Pointing Between Functional Departments

From the Ask Tom mailbag –

Any advice as to how to align all members from multiple cross-functional departments into one purpose; create an efficient, streamlined process that assures that communication, documentation and actual product flow is executed efficiently?  We are specifically a design firm, but our revenue comes from the products manufactured from our exclusive designs.  So, we have mature internal systems in each department, but the transitions of work flow from one department to the other sometimes break down, so there are logjams, finger-pointing and sometimes, general chaos.

Much of the answer is in your question.  Let me pick out your key words.

  • Purpose
  • Efficiency
  • Flow

Let me add three more.

  • Balance and optimization
  • Authority
  • Accountability

What you have described is the classic transition from Stratum III systems to Stratum IV system integration.  It sounds like you have done an adequate job of creating multiple internal systems, that are efficient in each of your workflow disciplines.  It is the integration of these systems that is giving you fits.  Let me take a stab at listing some typical systems in this flow.

  • Market research system
  • Design system
  • Prototyping system
  • Approval system
  • Production system
  • Finished goods inventory system
  • Marketing system
  • Distribution and logistics system

Each of your internal systems likely works well within itself, but now you are experiencing balance problems between your internal systems.  It is not sufficient to have a great design system and a great production system.  If you have a weak prototyping system, your designs will get stuck on paper and never make it to production.  You may have a great marketing system that creates consumer demand, but if you have a weak finished goods inventory system, your products will never find their way to distribution.  Your weak systems will be doing their best and your strong systems will be finger-pointing.

So, that’s the problem.  What is the solution?  This is a Stratum IV issue, where someone needs to have end-to-end accountability.  Some companies attempt to solve this problem by creating a role called product manager.  The product manager would be accountable for tracking each step, likely creating a Gant chart of product progress from one function to another.  While this role gathers necessary data about the status of a single product in the chain, it still might only document that the product is stuck.

That is why this is a Stratum IV issue, one of balance and integration.  The S-IV manager (likely a VP) would be accountable for examining each system for capacity and handoff.  This is not looking internally at the mechanics of a single system, but the interaction of each reinforcing system to each balancing system.  It is not a matter of having one or two high performing functions, but having all functions able to keep up with each other, optimized for capacity.  No single system manager will have the authority, nor likely the capability, to do this work.

And somewhere in this integrated whole system, there will be a constraint.  There will be some limitation in a single system which will drive the cadence of all the systems working together.  The hat trick is identifying and placing that constraint strategically.  Typically, this strategic constraint will be an expensive resource, too expensive to duplicate (which would double the capacity of that system).  The identification, selection and placement of the strategic constraint, and then subordination of all other systems to the strategic constraint is the work of the S-IV manager.

With this integrated system design, then the work of documentation, handoffs, communication and feedback loops begins.  Most companies get this backward and have a communication seminar without balancing the systems for total throughput.  You can imagine that this communication seminar makes everyone feel good, but nothing changes in throughput, the finger-pointing continues.

For more reading, start with Eli Goldratt’s The Goal and Peter Senge’s Fifth Discipline.