How to get a process improvement project approved

I just posted this on LinkedIn in response to someone asking how to get a process improvement project justified and approved.  Hope it helps...

One of the challenges senior executives face is trying to chose where to deploy limited resources ($) in order to maximize the return to the stakeholders.  And it is not an easy task when every project proposal promises spectacular "results."  Your first step in trying to differentiate yourself from all of your peers who are competing for the same funding is to put yourself in the shoes of the person making the decision.  And the best way to do that is to ask them "what results are you looking for and how will you measure them?"  I will use a labor example as they typically involve the hardest decisions.  Your Decision Maker may tell you "our labor costs are too high for this process.  I want you to find ways to improve our efficiency." 

Once you understand what result they are looking for your next visit needs to be to Finance to have them show you exactly the line item(s) on the P&L that correspond to the labor your Decision Maker is talking about.  And you need them to fully explain to you what makes up the labor cost.  That is you need to know specifically who is being charged to that account, how much they earn (salary or hourly rate), and how many hours they work.  And you need to understand all other costs that go into the rate such as overtime, benefits, and so on.  In many cases the people in Finance will have trouble figuring this out for you because they are used to dealing with totals.  Most organizations leave the detailed labor calculations to complicated computer systems.  But you must persevere until you can exactly match the weekly payroll using your own spreadsheet model.  You would be amazed at how many opportunities for improvement you will find just by making sure everything is being charged to the right account.  I have found individuals still being charged to an account years after they moved on to other areas in the company.  Ideally you want to go back 3 - 5 years by month when collecting this information.

The next question you have to understand are any changes in workload for your process.  Again you want to look at 3 - 5 years history where you compare "total units produced" (wigits, POs, quality audits, employee evaluations, projects, earned hours, quotes, etc.).  And you want to find out from your Decision Maker whether or not they expect volumes to go up, go down, or stay the same.  Often when they complain about costs being too high it is because volumes have gone down but they are still paying the same wages for a particular process.

After you understand the volumes and the hours you are now ready to calculate the efficiency of your process.  Using the information above you want to graph out 3 - 5 years of historical "units/paid hour" as well as the "labor rate per hour."  Here again if you are not dealing with a traditional production environment you will be amazed at the opportunity that will present itself.  I once did this analysis for a computer data center where I looked at "transactions per paid hour."  My reasoning was that the gross number of transactions in and out of the data center more or less drove the level of work inside the data center.  The more transactions you have the more equipment you need.  The more equipment you have the more technical resources you need to maintain and upgrade the equipment. And so on.  I calculated that they averaged 99 transactions per paid hour.  But their efficiency spiked in the summer at 145 transactions per paid hour.  And dipped as low as 60 transactions per paid hour in the winter.  No factory manager in the world would accept that kind of process variability and neither should someone managing resources working in an office.

What you will more than likely see are huge (30 - 50%) fluctuations in efficiency.  Again the further away from Production you get the less likely the resources will be managed against volumes.  In fact it is always amazing to me how efficient departments get in the summer months when people take vacation.  Again these variations are huge opportunities to look for process improvements.  Start to ask yourself "how do they manage to get the work done in the summer with 20% fewer people?"  hmmm....

The last step is to try and figure out how people spend their time.  This is where the real process improvement comes in.  In the case of a Procurement department for example if their Key Volume Indicator is a Purchase Order you may calculate that it takes on average 4 paid hours per PO (or more).  As part of your project proposal you want to identify ways to reduce non value added time from that 4 hours.  You include some estimates for how many hours you plan to eliminate and when you plan to have them eliminated.  Then you back-calculate using your Decision Maker's forecast how many people will be eliminated and use that as your Measure for Success in your proposal.

Of course this is not as easy as it sounds.  Whether you are an internal or an external consultant you still have the challenge of getting the area manager on board.  And if they are dealing with a 'broken' process and all the associated issues they will likely not welcome the idea of downsizing staff.  How well you engineer the situation to ultimately generate results for your company or client will determine where you fall on the continuum between Report Writer and Implementation Expert.  I've often said that process improvement projects are 10% technical and 90% tactical.

Best of luck.

Part 2: Don't Just Clear The Snow...Make It Not Snow In The First Place

Ok so we talked a little bit about Value.  And in particular the Value that this group of snowplow maintenance workers provides.  Namely fixing snowplows and performing preventive maintenance on snowplows.  And we also conducted a survey with one of your customers to find out just how much Value she thought you provided, which wasn't very much.  So tell me, what do you think so far?

"I think it's a bunch of crap.  What does she think?  The plows just fix themselves?"

general grumbling in agreement

"I'm not sure, but she did seem to understand a surprising amount for someone her age.  Anyone else?"

"So according to you nothing we do here is of any Value?"

"No, according to one of your customers.  Anyone else?"

"So I guess we all just go home then.  Let's see how happy Mrs. Smith is next winter when none of the plows are working."

"Ok let's continue.  The first step in Lean Manufacturing is to understand the Value you provide from your customer's perspective.  I think we've covered that.  The next thing we need to do is talk about our Value Map.  Does anyone know what that is?"

"It's our process."

"That's right.  So what's a process?"

"It's what we do.  The steps we go through."

"Yes that's right.  Has anyone here ever seen a process?"

"Yeah every year or so they bring in some guy like you to draw it out for us.  It's supposed to make us more efficient.  Funny thing is none of you ever pick up a wrench while you're here."

general chuckling and nodding of heads

"Trust me even if I knew what a wrench looked like you wouldn't want me to pick one up, much less try and use it.  You'd end up with a lot more snowplows to fix.  But I'm glad you are familiar with processes. 

So since you're all such experts can someone quickly tell me what your process is here?"

"Anyone?"

"They drive 'em, they break 'em, and we fix 'em"


lots of chuckles.  one or two high-fives

"Wow, that's probably the best description of a process I've ever heard.  Well done.  Thank you for that.  If you don't mind I'm going to write that out on the board here...

1) They Drive 'em
2) They Brake 'em
3) We Fix 'em

"Did I get that right?"

heads nodding

"Good.  But wait a minute.  I see 'We fix'em' on there.  But where is the preventive maintenance?  How does that fit in?"

"That's a separate process."

"Is it?  Let me make a suggestion.  What if I used a Decision Point and re-wrote the process like this:"

1) We Perform preventive maintenance
2) They Drive snowplows
     ??? Decision: If broken then...3), If not broken then 1)
3) We Fix snowplows

"Do you see what I've done?  I've reflected the fact that you perform preventive maintenance in order to reduce the number of snowplows that break.  Does that make sense?"

a few people leaning forward and nodding

"Ok I think we're getting somewhere.  We've talked about Value with our customer.  And we've drawn our Value Map.  The next thing Lean Manufacturing tells us to do is identify our Value Added activities and our Non-Value Added activities or Waste.  Anyone want to take a shot?  Or do I need to give Mrs. Smith a call?"

no response, not really liking where this is going

"Ok let's start with what we know.  Step 1); according to our customer that's "Waste."  Step 3); again according to our customer more "Waste."  So that just leaves Step 2).  Any guesses?

"Value Added?"

"Are you asking me or telling me?"

"Value Added"

"You're sure?  Is that what Mrs. Smith would say?"

nodding of heads

"I think so too.  No need to bother her again.  Let me re-write our process including Value Added steps and Waste:

1) We Perform preventive maintenance (Waste)
2) They Drive snowplows (Value Added)

     ???Decision: If broken then...4), If not broken then 1)

3) We Fix snowplows (Waste)

"Before we talk about how to improve our process I can tell that the label 'Waste' isn't sitting too well with some of you.  Is that fair?"

nodding of heads

"Ok let's think back for a second.  Who told us it was Waste?"

"Mrs. Smith did.  But she doesn't understand our process."

"Maybe not.  But do you think she understands the Value that having snowplows and drivers to drive them provides to her?"

nodding of heads

"Ok let's look at it another way.  Would you agree that to improve a process you want to spend more time doing Value Added tasks and less time (or no time) doing Waste?"

nodding of heads

"And since you feel that what you do here should be considered Value Added let me ask you this?  Should you be doing more preventive maintenance?"

"Of course.  We never have enough time to go through everything, especially in the winter.  That's half the reason why we end up fixing these things over and over again.  That and the fact that those guys somehow think a blade can magically go through a curb without getting bent all to ....."

chuckling and nodding of heads

"How long does it take to do the preventive maintenance on a typical snowplow?"

"It depends.  Maybe 3 days to do it right.  More if you find problems to fix."

"Why don't you take longer?"

"What do you mean longer?"

"Why don't you completely strip the machine down and inspect each and every tiny part under a microscope to look for signs of wear.  Then give everything a fresh coat of paint and lovingly re-assemble it.  Wouldn't that help reduce the number of breakdowns?"

"That's stupid.  That would be a waste of ...."

"Go on."

"You only need so much inspection to find the most common problems."

"Oh I see.  So someone sat down and figured out the fastest way to perform preventive maintenance without taking too much time.  Is that it?"

slight nodding of heads, not liking where this is going again.

"So tell me.  We all agreed that we wanted to spend as much time as possible doing Value Added tasks.  And as little time as possible doing Non Value Added tasks.  Kind of sounds to me like you are trying to do as little preventive maintenance as possible but still finding the major problems.  Is that right?"

no response

"Kind of sounds like something you would do to address Waste in a process, not Value Add. 

Look, one way to decide if a task is Value Add or Waste is to ask your customer like we did earlier with Mrs. Smith.  Another test you can do is ask yourself 'should we be doing this as much as possible?'  In the case of driving snowplows I think it's pretty easy to agree that if it's snowing your customers would want you Driving Snowplows as much as possible.  That passes the test for Value Add. 

But in the case of fixing snowplows, or even performing preventive maintenance, the less time you can spend the better.  In fact I might go so far as to suggest that if there were such a thing as magical snowplows that never broke down don't you think the taxpayers would all be in favor of buying those instead?"

no response

"So now what do we do?"