WIP and TameFlow

Kanban tells us to limit WIP and we all know that having too much WIP is bad due to time lost in context switching, etc. When looking at WIP from the TameFlow approach, how does this affect the financial throughput and the operational throughput?

P. S. I’m a mathematical guy, so I prefer an explanation through the equations :slight_smile:

1 Like

Hi Savio

As you know the fundamental equations of Throughput Accounting are:

  • Throughput : TH = Revenue - Totally Variable Expenses
  • Net Profit : NP = Throughput - Operating Expense
  • Return on Investment : ROI = Net Profit / Investment

So the question really is, how does reducing WIP impact the above.

There are two ways to answer.

One is to consider the immediate “investment” you make when acquiring information so that you know what to do. Call it Story Writing, Requirements Elicitation, Analysis, etc. This increases the Investment denominator in the ROI. So reduce any upfront work, and you have better ROI. But it can get ugly quickly because you will be tempted to do allocations, and fall back into the world of cost accounting.

Another way is to reason in terms of Flow Time.

If your Financial Throughput (TH) is proportional to your Operational Throughput (TP), it is easy. And if it is not (because maybe you have a subscription based SaaS-like service, then you need to work out with marketing what the connection is between sales funnels, life-time-revenue, and other such variable, with TP).

With Little’s Law we now that: TP = WIP / FT

So if you can produce an X% reduction on FT, you will get an improvement on TP that is given by:

1/(1-X%)

And how do you reduce FT… As per usual TameFlow Approach: First reduce Wait Time and second improve on the constraint. To reduce the Wait Time the only recipe is to reduce WIP, and you minimize WIP by knowing where Herbie is an applying Drum-Buffer-Rope scheduling, and get as close as possible to One Piece Flow.

1 Like

Makes sense. Just to close the loop, my reasoning is as follows:

  1. The “calculating the investment” angle doesn’t excite me. Our analysis process can be quite lengthy and hard to calculate. I would prefer to include that it the whole process, as part of FT, so that it can be measured and improved through the 5 focussing steps, like every other stage in the system. Treating it as investment just risks hiding it.

  2. The part I could wrap my head around was that using TP = WIP/FT, increasing WIP is good as it would result in increased TP. This is indeed true, but only if it does not increase FT. I guess the only situation this would happen is if you had excess capacity in the system, at your constraint. Then again, even in this case it is most probably easier to focus on decreasing FT, rather than increasing WIP.

Hi @savio

Regarding…

That is true only on paper - the equation holds in theory. The real killer here is not that you need to keep the same FT, which would imply “working harder.” The killer is that as you increase WIP while keeping the same resources then you will induce Multitasking with the cascading effect that Wait Time will increase, Flow Efficiency will go down, and your Flow Time will grow - and typically it will grow at a rate that is greater than your increase in WIP. And to increase WIP you would have to use the… ehm… whip and push work into the system; but we know that pushing is a performance killer compared to any kind of pulling mechanism.

If you want to increase WIP and (possibly and hypothetically) keep the same Flow Time, then you need more resources. But that means more Investment (hire people, new computers, training, onboarding, etc.). And in the fundamental Throughput Accounting equation, increasing Investment is contrary to your objectives; ideally you would want to decrease your Investment.

So focus on: decrease Wait Time, increase Flow Efficiency, increase Throughput at the Constraint by applying the 5FS all the while you Limit WIP by using Drum-Buffer-Rope Scheduling.

(Besides, have you seen the new book where all this is explained in depth?)

1 Like

Yes totally agree with your logic.

(No, I have not seen the new book yet. I have heard about it through this forum but do not have a copy yet.)

1 Like

It is good to think individually in terms of each variable in the equation. But the real math must be based on the differential SUM of all 3 variables : TH - OE - I

As for OE, a simple heuristic comes to mind. The longer you stay in the system, the bigger the accumulation of OE. So reducing WIP has a beneficial impact on reducing FT, which has a beneficial impact on reducing OE.

1 Like