Re: Financial data

["Vic Forte" <vic@vicf.com>]

> I need to take another step with the CFO, before he pushes
> the trend line button in Excel and predicts bliss. What
> is the step?

I have been doing something similar for some time now. Here's the approach I
take.

I concentrate on the main aggregate data from the P&L, and I use control
charts not run charts:

Sales (I break sales down into a volume component and a prices component)
Direct costs (with direct costs I also maintain a percentage-of-sales
control chart)
Distribution costs
Management costs
Office / Admin
Production Overheads

I.e. the main summaries presented in the P&L

I also maintain control charts of the more detailed data, but keep these to
myself.

I look out for special causes on the main aggregate data, and if I find one,
I bring this to the attention of the team as a kind of alarm signal. We then
all "drill down" into the more detailed figures to see where the underlying
causes are.

We then discuss the special causes together, and plan improvement actions
where appropriate, or ask the question "what happened here".

You may find that the financial data you have may not be sufficiently
detailed to properly identify the special cause. You will probably get lots
of different opinions from people and also find that you have to
de-aggregate the data. The finance department may already have some of this,
but in the end you will probably have to move closer to the process itself.
This is where leadership comes in, and working on the process to improve it
with the help fo the work force.

I find that by and large the data exhibit controlled, but high, variation.
One of the issues is variation caused by the way that the data are collected
for financial purposes. Work needs to be done on presenting the financial
data in a way that minimises variation caused by the collection process
itself. For this you will need a financial director (CFO?) who understands
and supports the principles of variation that you are working to - or else
you may need to do some intermediate work on making sure that the data fit
the appropriate time periods.

Where the range of variation is unacceptably high, work needs to be done
with the top team on asking the question "why such high variation" and "can
you see how this level of variation makes the business more difficult to
control". Discussing causes of variation in this way may lead to a strategic
re-think, if you are lucky.

So you have two distinct processes going on, the first one is the
identification of, and learning from special cause variation. This you can
usually do effectively with the help of the work force. The second one is
studying the range of variation itself and taking steps to reduce it. This
is more likely to be a matter for the senior management team because it is
going to be a system wide phenomenon and will bring in issues of purpose -
system design and so-on.

Good luck.