Conference 7.286::digital

    Re .0
    
    The owners of a company do not measure goodness.  They measure profits.
    What are the other measurements you feel Digital should be using to
    track their return to good health?
    
    If I could run a company with one employee, outsource all of my work,
    and make a profit (and by that I mean one that would allow me to
    vacation anywhere in the world :-), I would do it.
    
    As demand for a company's product grows, you have to decide whether
    you are going to satisfy that demand or restrict the availabilty
    of your product (a small mom-and-pop operation may want to stay small).
    If you decide to grow, then you should determine whether it makes more
    sense to hire and buy additional resources, or if you should outsource
    some of it.  It is when you think you can do it all by yourself (that is, 
    no outsourcing), that you are probably going to run into problems.  That
    was part of Digital's problem.
    
    When our profit margins were high, we could afford(?) to do things
    inefficiently.  But now, our profit margins are small.  And all the
    problems that always existed are now in plain view.  
    
    Can you give some examples of outsourcing that would be a bad idea?
    I'm not saying outsourcing is always good, I'm just interested in
    hearing some of your examples.

    
    
    Printing documentation and software media replication....
    	
    	excellent candidates for outsourcing!
    
    Ever try dealing with SSB?  Process, Process and more Process 8*(
    
    Now call another one stop shopping supplier and see how your treated.
    
    
    		Like a customer!!!!
    
    I do remember being a customer... it was nice.
    
    gb
    

    
     I look back at all the things DEC used to make as a hardware 
    vendor but didn't know how to invest in, manage, cost effectively
    engineer and push volume on.
    
     We no longer design or manufacture: power supplies, disk drives,
    cabinets, mice, keyboards, floppy drives, pcb etch, monitors, cable/
    cable assemblies, and drams (yes, DEC made drams some time ago).
    
     Each in it own right could be a nice little business. So considering
    the amount of VA left in system design these days, we should be up
    around Compaq's $600K. 
    
     

re: Note 3317.3 by PCBUOA::ROGICH

>    cable assemblies, and drams (yes, DEC made drams some time ago).

How long ago and where did Digital make DRAM's.  I'm kind of positive that
we never made a DRAM product in Hudson.  SRAM's yes, but no DRAM's.

I disagree with your 600k/employee.  As long as we have SI, consulting and
field service people, you can't compare our revenue/employee with Compaq,
Dell and Gateway.  

-Bruce

re: .1 (how else...)

Certainly profit per employee would be an interesting metric.  Also market
share for generic markets (for instance, i don't care if Digital is the
leader in services for the VAX installed base...just tell me about the
midrange/server installed base).  Any metric that could show the ratio of
employees directly involved in generating revenue (sales, service,
manufacturing, engineering) vs. non-revenue generating employees (admin,
logistics, finance, personnel, IM&T, marketing) would be very useful too.

In general i think it's better to have consistent profitability than high
profitability on a variable basis.  And have that coupled with high market
share and "mindshare".  Those seem to be the things that will give you
long term staying power.  We seemed to be achieving some decent mindshare
in the early '80s, before the Smith-Shields-Hindle-Grainger-Zereski-Sims
clique got thoroughly entrenched.  Palmer cleared out that crew but didn't
seem to come up with credible replacements (Bill Johnson as head of
Marketing??) until it was too late to save the 80K jobs he said he wanted
to save.  Now i expect we'll end up with about 45K jobs--but much smaller
revenues too.  It will be interesting to see how it all plays out, and
whether the revenue per employee really improves all that much.

							- paul

    AARGH !
    
    If we are going to use Revenue per employee, then we should also note
    Total Cost per employee. (A simple subtraction then follows..)
    
    re the 1 employee company - you still have to do the complicated subtraction
    outlined above.
    
    Focus on margin please. Our problem has been that in our *very core*
    business, ie making H/W, S/W, we don't make $$. And I guess it would be
    hard to quit your core business. So we need (in our core business) to
    increase revenue (assuming cost stays the same), decrease cost, or
    both. I think we are doing both. Oh, and it would a jolly good idea if
    we didn't discard too many of the bits of the company that currently
    have a positive margin while we go about it. Otherwise when we have
    finished, the sums still won't add up.
    
    [Yes, I know that our current systems don't make it easy for us to tell
     where, and precisely how much, we make/lose $$, but the simple
    equation above remains valid...]
    
    AW

    Re: Note 3317.4 by YIELD::HARRIS
    
}   How long ago and where did Digital make DRAM's.  I'm kind of positive
}   that we never made a DRAM product in Hudson.  SRAM's yes, but no
}   DRAM's.
    
    Actually we were making 4K DRAMs off a design we had licensed from
    MOSTEK in 1979-1980. The DRAM manufacturing started in Worcester (WX),
    the space we had rented in the Sprague Electric facility, and the
    design was one of the test vehicles for the early Hudson plant.
    
    We got in the business during the 1978-1980 DRAM shortage, when it was
    difficult to get DRAMs. Our yields were never commercial. 
    
    /jim

    I'm glad to see "revenue/employee" get its own note. I'd like to see
    someone explain how this ratio is a true indicator of goodness. Seems
    to me that *the* measure of success of a for-profit business is simply
    profit.
    
    If you're going to invest in a company, why should you care how many
    employees they have if they generate the profit you want? What
    difference does it make if Company A generates $1M profit with 10,000
    people whereas Company B generates $1M profit with 20,000 people? (In
    fact, if I were altruistic I'd be inclined towards Company B since
    they're providing jobs for twice as many people :-)
    
    Seems that the conventional wisdom is that Company B is in worse shape.
    But who's to say that if we tried stripping Company B down to 10,000
    people that it would continue to generate the $1M profit? (In fact, I
    still believe this is true with Digital. We've practically cut our
    employee population in half, and our profits have continued to
    decline.)
    
    I'm eager to learn, if someone can educate me. Thanks.
    
    	BD�

    	The fact that anyone is still talking about revenue (or even cost
    or profit) "per employee" I find encouraging. It means we are still 
    in a service industry where the employees still count. Imagine that we
    were a manufacturing company, and that with improved automation we
    could cut down to a single employee whose only job was to call in some
    third party when something went wrong. In that sort of environment we
    would be talking about revenue (and profits) per kilowatt-hour since
    the employee costs could be ignored.

    
    	Surely the only relevant measure is indeed profitability, but
    qualified as NET profitability as a PERCENTAGE of GROSS TURNOVER?
    
    				Malcolm.

    As *one* way to measure, I think revenue/employee is relevant iff you
    are comparing peer companies.  For example, it *may* not be relevant to
    compare r/e between a company of 10 and a company of 100,000.  Or, it
    *may* not be relevant to compare r/e between a company in a commodity
    business with a company in a customized solutions business.
    
    I think a problem lies in having a deficient r/e ratio compared to
    peers, then taking steps to improve the r/e which, in reality, places
    you among different peers and new but relatively unimproved r/e ratio.  
    Lather, rinse and repeat until bald ...
    
    Steve

    re .8 and others
    I agree that profitability is the important measure, but not measured in 
    $, but rather in the return on investment to shareholders (profit per $
    invested; smaller $ profits may still yield greater % return to the
    shareholders).  In addition to operating costs, there is usually a direct 
    relationship between # employees and required capital investment (office 
    facilities, etc).  
    
    I suspect the theory is that by reducing employees by a greater factor 
    than revenues (possible?), Digital will reduce both operating costs 
    (thereby increasing $ profits) and reduce capitalization (to provide a 
    greater return to shareholders - the ones that really decide our future).  
    One measure of how you're doing (relative to peer companies) is R/E.  
    
    I'm not saying the approach is right or wrong or will work, but 
    maybe it's understandable (?).
    
    ...jack

    I think that revenue/employee is only relevant if cost/employee is
    consistent.  A top-heavy company could distort those numbers. 
    Incidentally, I wonder what % of total company payroll goes to the SLT
    compared to 5 or 10 years ago?

    re: .5
    
    I like the way you think.  But then that is the way I think.
    
    re: .12
    
    You make a very valid point.  ROI is important, either through
    dividends or stock growth.  I am not sure how that figure help the
    routine (day to day) management of a company.  I see that ratio more as
    an indication of the type of investment the company is.
    
    I also would susgest that profit / emplyee should be done across the
    various business units of Digital as well.  This would help compare
    apples to apples since we do have quite a collection of different
    industries within Digital.
    
    rww
    

    re::All...
    
    I suspect all of you are correct,,, Rev/Emp is simply ONE of the
    ratios companies (Digital included) must be aware of and MANAGE.
    
    Each financial indicator is important, however as all have said;
    we must monitor all as a whole.
    
    
    Need more be said ???
    
    Marty

    
        Revenue/employee, profits/employee, whatever .....
    
        You can use any of these if you use them properly, if you 
        understand their basis and act accordingly.
    
        The main problem (or stupidity) in revenue/employee as an 
        argument for downsizing is that company leadership and 
        analysts seem to think that you can cut the denominator 
        (employees) while the numerator (revenue) stays constant. 
        No one seems to understand that revenue is also a function 
        of employees, and probably not a simple one.
    
        So you start on the slippery slope:  cut employees - oops, 
        revenues go down, so cut employees - oops, revenue goes 
        down ......    in a semi-infinite loop
    
        On the other hand, profits/employee is probably too volatile
        a parameter to interpret; revenue/employee is more stable.
    
    
       Peter
    

    re: .1
    
    The owners of a company are important.  So are the employees,
    customers, and even the suppliers.  Hell, even government
    regulators are important!
    
    Good management isn't going to ignore any of these.
    Profit is important to measure the return to health.  Probably
    more central than revenue/employee.  But I'll stand by my original
    claim:  there is no single metric of goodness.
    
    
    Regarding outsourcing:
    
I don't think I'm smart enough to come up with a bulletproof example of
bad outsourcing.  Here's an example that would give me the willies, though:

   Outsourcing the production of Alpha chips

If the potential suppliers require a minimum commitment, I've got
a risk on the downside.  Contracts can be hard and expensive to
break.

The upside worries me even more.  Let's say the following three
things come together:

1.  The computer industry, which is cyclic, goes into its up cycle.

2.  Through careful ads and PR, backed up by some reality, the
    name "Digital" becomes increasingly known and trusted.

3.  Some industry, like say, the video imaging industry, hits the
    wall with 32 bit processing, and really needs to go to 64.


Now I've got my choice between an availability risk and a quality
risk.  Other suppliers, or returning to insourcing impose either
a risk of bad quality or a risk of long delay.  With my existing
suppliers, I'm competing with their other customers, who also
want to eat my lunch with my customers.

I'm probably going to end up leaving a lot of money on the table.
I don't really know that any of this is true, but I'd sure
worry about it if it were my call.

Dave

    
    
    

    IMHO, revenue/employee is only valid with like (VERY like)
    industries and is a competive measurement.
    
    I dare say EDS's revenue/empolyee is very different from Compaq's. 
    True, both are in the "computer industry" but not the same area at all.
    
    Also, consider counsulting.  To reach the 220K/employee, you would need
    6 to 7 people at 250K/year to support 1 non-revenue person.  However;
    at 250K/year, 1 person can support 2 or 3 peole based on margin with
    some profit leftover.
    
    In the commodities area, the number are completely different yet again.
    
    This is why I favor profit/employee or at least margin/employee
    measured for business unit.
    
    rww

    
    	The best measurement of all is margin/employee. It becomes very
    easy at that point to then look at "winners" and "losers" in a
    multiproduct company, and their growth over time.
    	Since margin is both product and revenue sensitive, it acts as
    the best rule of thumb for senior managers over time. It also allows
    for the "tweaking" of sales/marketing promotions and programs for
    additional margin dollars (although at a higher "cost").
    	That will also explain why we love the software "giveaways" so
    much - no real cost, but margin dollars for manufacturing go up, up,
    up.
    
    		the Greyhawk

    No No No...it is not Margin per EMPLOYEE but rather Margin $ per Labor
    Dollar expended. Actually it should be Margin $ per Expense Dollar
    expended.  What is important is how much do you spend for each dollar
    of profit you make.  Hopefully it is less than a dollar.
    
    I am tired of competing against districts who have a higher Revenue or
    Margin per employee metric, while they fail to include the 25- 40
    Non-Headcount Temps/Contract Workers into their equation.  It also
    doesn't take into account a cost structure which may be under control. 
    If I can afford 175 people for the same labor cost as another district,
    which supports only 150 people, I am at a disadvantage on Revenue /
    Person Metrics.    

    Re: .20
    
    >>I am tired of competing against districts who have a higher Revenue or
    >>Margin per employee metric, while they fail to include the 25- 40
    >>Non-Headcount Temps/Contract Workers into their equation.  It also
    
    Let me generalize that.  I am tired of hearing about districts
    competing against other districts when the same effort would have
    been better expended competing against IBM, HP, SUN, etc. within
    the district.  Let's be clear:  internal competition is an inducement
    to work harder or work smarter.  Your real competitors are the people
    standing between you and your chips.
    
    Dave
    

It's possible to "monkeywrench" with any statistic, by using it
in a strange context for an unorthodox purpose.
Several previous replies say the same thing, only more eloquently.

It's even possible to monkeywrench with the ultimate statistic:
profit, the famous "bottom line".  Here's one way:

    quietly decimate engineering

Huh?  I thought engineering was the backbone of our product stream,
and therefore of our revenue stream?  Won't you ruin revenue if you
apply a monkeywrench to engineering?  It was, it is, and you will.
When technology advances again, you will no longer be able to keep
pace using your own expertise.  Eventually, you will no longer
be able to introduce leadership products. If you're lucky you might end up
like Tandy, pushing lots of schlock.  If you're not lucky, you
might end up like Wang.  So how does the monkeywrench work?

The key word is 'eventually'.  There's a long delay time between
the time a good piece of engineering work is performed and the time
that work is reflected in the revenue stream.  My guess is that it's
highly variable, but goes anywhere from 3 quarters to 8 quarters.
The delay between work and revenue is much shorter for field service,
manufacturing, and sales.  So you are able to cut *current* costs
at the expense of impacting *future* revenues.  For a few quarters,
you could look like a hero.  In conjunction with other efforts, you
could even put the company in the black.

You need to be quiet about it, so that not too many customers will
stampede.  They'll stampede anyway, if they hear "Chapter 11" too
often.

BTW, I don't think RP is doing this monkeywrench, or any other so
far as I know.  If anything, people are frustrated with RP because
he hasn't come up with a laundry list of 'simple, obvious things
that everybody agrees should have been done years ago'.

Hell, if it were *that* simple, I'd apply for the job.

Dave

    	The delay between work and revenue can be much longer for research
    work. It took around 30 years for the field effect transistor to go
    from something demonstrable in the laboratory (1937) to something
    commercially viable. When I started work, just out of university, the
    company I worked for had about half the patents in fibre optics, but
    had fired all the engineers and were getting out of the field because
    there was clearly no future in it.
    
    	I can produce other examples of the time scale. In 1972 Prime
    computers were amazing, but expensive. They got a new managing director
    who got rid of most engineering. By 1975 they were *very* serious
    competition for DEC, and in 1978 I was involved in a bid where we lost
    trying to sell VAX machines against them, but eventually their lack of
    new engineering cut in. Note that it took 5 years or more.

    HOW ABOUT REVENUE/VP? WE MUST REALLY BE ON THE BOTTOM.

Profit per VP would also be an interesting metric ...  I'm sure that some VPs
such as John Rando and Enrico Pesatori run profitable operations. 

>================================================================================
>Note 3317.24               Beware of Revenue/Employee                   24 of 24
>PCBUOA::MAY                                            1 line  15-AUG-1994 20:13
>--------------------------------------------------------------------------------
>                              -< REVENUE PER VP? >-
>
>    HOW ABOUT REVENUE/VP? WE MUST REALLY BE ON THE BOTTOM.
>

    re:.25
    
    	Revenue/employee is total revenue divided into total number of
    employees. To have an identical metric you would have to take the
    *TOTAL* number of VP's divided into the total revenue (or profit if you
    prefer).
    
    	Ray

    
    The thing that bothers me about a metric like revenue/employee is we
    treat it like a cause when it's really an effect.  I.e., successful
    companies have high revenue/employee because they're successful; the
    success comes from other things, like efficiency, having products
    people want, etc..
    
    If I dress like a doctor, does that mean I can cure people?
    
    len.
    

    No, but if you cut off a leg, you have a wonderful cure for corns.

    Re: .27
    
    Excellent!
    

    This is a brain-dead discussion! Why?
    
    
    
    This company is not making a profit right now!!! Once we're back writing 
    black numbers then we might discsuss profit per whoever. :-$
    
    In the good old days (~14 years ago) European Operations out of Geneva 
    defined an EXPENSES-over-INCOME factor. Started out at the HQ as 0.9
    (means for the 90� spend you need $1 income) which yields at a profit
    margin of 10%. Whenver this E-over-I was passed down one organisation
    level it dropped a bit 0.8...0.7... to pay for the overHEAD. For the
    lowest cost-center level in a Digital trainings department it was 0.48.
    Could you imagine what we had to charge a customer for a trainings
    class? I never understood where all the money for the 0.9-0.48
    difference went.
    
    Guenther

    The stat that I would like to see is the average salary today as
    compared to four years ago. I believe that even after all of the dumb
    sizing, that the average salary has risen. This due to the chiefs
    cuting the indians and then filling the indians positions. JMO. RKL. 

    
    
    		I found this phenomenon very interesting as well.
    	I know of an Indian that lost their tech job and had to
    	accept a assembly job or have no job at all. The decrease
    	in wages was substantial. Yet, I have seen managers become
    	supervisors and I can only wonder if they lost pay due to
    	job change or not. You don't think they'd tell me do you,
    	I was an Indian , once.
    
    		But if there are any Management types out there 
    	experienceing this cut in pay , please let me know so
    	I can breathe a little easier and get more sleep at night.
    
    				(%^)
    
    				Pablo
    
    	

    RE: Indians to Chiefs Ratio...
    
    Personaly, I'd like to see a 'realistic' breakdown of this company by
    job code/type/class (whatever you wish to call it) and be able to see
    just waht the Indian/Chief ratio really is.
    
    My personal observations of this mess so far point to the good old boy
    club protecting their own at the expense of those they do not like. I
    feel that there are far too many management/overhead types that produce
    nothing and consume way too much. What's that old saying? "Those who
    can, do - those who can't, manage." I'm sure I just stepped on a lot of
    toes with that one, but look around, how many non-revenue producers can
    you see right now? Enough said.
    
    AS far as figures go, I has a manager once who typified the statement
    "Figures don't lie, but liars sure can figure." Until there is an open
    and non-distorted accounting of what comes in vs. what goes out, I
    remain unconvinced of this companys future. I sure hope it survives,
    because I really like what I do, and I like being looked to as "The DEC
    guy". We CAN pull this out, but a reboot will be neccesary.
    
    Of Course, all this is my opinion, and I'm keeping it. 8-)
    
    .mike.

    The other trouble with a revenue per employee number is that when years
    of bad decisions and no decisions by higher level management finally
    show up in falling sales, the rank-and-file employee gets blamed for
    the loss.  
    
    --bonnie

    As has been stated by other, revenue/employee is only one way to
    measure a company.  I agree with this completely.
    
    What concerns me is:
    
    1.  The goal is to return to profitability.
    2.  The method is to reduce costs.
    3.  The metric is revenue / employee.
    
    The metric being used does _not_ tell you _anything_ about how well the
    method is doing in reaching the goal.
    
    The metric used does not help me make decisions that will further us
    toward our goal.
    
    Now I realize that profit and cost figure are closely guarded and not
    for public consumion, except in the most general terms, i.e. quarter
    ending reports however, I would really like to have addition metrics so
    the people in the field can assist in the turn around.  After all, the
    bulk of the work (and success) is going to occur there.
    
    rww
    

Digital's antiquated business systems do not allow an accurate measure of
profitability by business segment or product lines.

Thats why I get a big kick out of the individual groups claiming to be
profitable.   Its all in how you measure it.

I can play with numbers and make any individual group look profitable.

But when outside people get a look at the numbers (e.g. CSC looking at the
books for Digital Consulting) then the truth comes through loud and clear.

Revenue or net divided by the number of people in management would be a very
interesting number.  Because of the hijinx that have been going on the last
couple of years with TFSOing ICs and giving managers their jobs, i would want
to calculate the management population so as to include all individuals who
have had a management job code at any time in the last 18 months who are still
with the company, plus anyone with a job description of "program manager".
And maybe NOT count anyone who has a job title of just "supervisor" for ICs.

								- paul

    re: .36
    
    You have voiced my worst fears.  I hope you are wrong.
    
    rww

    RE .35
    
    How very true and how very sad.   The metric does not measure if the
    method to reach the goal is working.   And, since metrics drive behavior,
    until the metric does reflect *PROFIT PER SALE*, we will continue to give
    away the store.                        
    
    I do agree with .36 too.   Until Digital can get a handle on what it
    really costs to produce every item, before overhead, and then
    accurately figures in overhead, we will not know if sales are
    profitable down to the part number level.    Until we know that, then
    we cannot put the right metric into place!   
    
    Milo Minderbender (sp?) would just love this.
    

    RE: .38
    	The bad news is that the remarks made in .36 are absolutely true.
    There is no doubt!

    RE: .36, .40 
    
    If only you were right - the fact is that CSC did not find out that 
    DC was performing poorly - they couldn't make any determination because
    of the conflicting data collection systems and organizational metrics.
    
    				----------
    
    Re: Revenue Per Employee
        Cost Per Employee
        Profit Per .....
        Margin Per .....
        Etc.
    
    A single metric or set of metrics can not be applied as a measure of
    success or failure to a company whose operations are as complex as 
    Digital's.  All metrics and measurments must be "balanced" by contra 
    metrics or measurments to insure that the effects on the organization 
    are as planned.  Selection and assignment of metrics may be the second 
    most important task of senior managment (after charter definition), and 
    it should be undertaken with lots of soul searching and care. 
    
    There are fundamental differences between operations and yields of
    "marketing" companies, "manufacturing" companies, and "service"
    companies.  At a lower level, these differences also exist when you
    look at "structured" service organizations (MCS) and "professional"
    service organizations like DC/IS.  If you agree with this, you know
    that past business metrics and measurments processes at Digital did 
    not lead to success.  
    
    KO once espoused that one could not properly manage from a "control
    panel".  At the time, I took that to mean that management by numbers
    without understanding underlying operations and success factors couold
    not succeed.  I still believe that is true.  I only hope that the
    divestiture and divisionalization being undertaken will:
    
    	1.  Not throw out the baby with the bath water 
    	2.  Allow operating units to finally understand whether they 
    	    are successful or not.
    
    One thing for sure, this process will still NOT allow senior management 
    to apply a single set of "simple" metrics and financial measurments to
    all businesses - each stands alone and has unique success factors.
    
    Managers SHOULD manage, not measure!
    
    	Bill