Conference 7.286::digital

  Sometimes, customers just want to see a friendly face for their
  $n,000 per month. It helps if they can associate a smiling face
  with that big maintenance bill.

  Other times, even though you've asked <some specific technical
  question> ten times over the phone, you just need to go out and
  see it for yourself. Here are several examples from my past:

    o Over the phone: "Did you measure the +5?" "Of course we did!"

      On site: "Did you notice that the +5 is +4.75?"


    o Over the phone: Are you *SURE* that the Razzeframizz L pin
      is grounded?" "Of course! I told you that already!"

      On site: "Say, what's this ground wire doing hanging here?"


    o Over the phone: "Are all the LAT parameters the same between
      these two systems?" "Of course! We checked!"

      On site: Have you noticed how much longer the "transmit" LED
      stays lit on the older system? And how it's set to send LAT
      packets that are ten times as big as the newer system?"


  In each case, somebody spent a kilobuck or so to fly me out to the
  site to discover, umm, "an innaccuracy" in the phone report of the
  situation. In none of these cases were any parts replaced.

                                   Atlant

    
    I can't answer your direct questions, but I can tell you why we
    turn some calls over to the field.
    
    a)customer is unreliable or unable to work the issue and has no dial
      in access. At times like that, someone you can believe in front of
      the system is necessary. Aside from flying out there yourself, that
      leaves the field server person (or you can ask the customer if the
      system manager is around and he can get really pissed off and tell
      you that he IS the system manager.....been there...done that...;-)
    
    b)there is no expertise available off site (in the center). How can 
      this happen? No training. Not enough people. Too many calls in the
      queue. Too many products for nn people to know all of them. No
      training. No access to the equipment that the customer is using to
      try and reproduce the problem (the world of Multi Vendor support and
      all that hooey..)
    
    In the Digital fantasy world, all customers would have dial in or
    internet access, and specialists would have the necessary skills
    and tools to work the calls. 
    Aside from flat out parts requirements or hard down
    systems, we would NEVER have to log a call to the field.... 
    
    cpitt
    

    Re: .1
    
    Gotta love doing over the phone diagnostics... "You're sure
    it's set up exactly the same, right down to the bit level?"
    "Yes."
    "You're really, really sure it's exactly the same?"
    "Yes."
    ....
    "Why are all these fields zero instead of containing status
     information?"
    "Oh, those fields.  Nothing would ever depend on those."
                

re: .3 and .1

>    Gotta love doing over the phone diagnostics... "You're sure
>    it's set up exactly the same, right down to the bit level?"
>    "Yes."
>    "You're really, really sure it's exactly the same?"
>    "Yes."
>    ....
>    "Why are all these fields zero instead of containing status
>     information?"
>    "Oh, those fields.  Nothing would ever depend on those."
                
My favorite from the past (this wasn't at DEC):

   Support person:  What does it say on the screen?

   Customer: Nothing.

   Support person: Are you sure?

   Customer:  Yes, there's nothing on the screen.

   Support person: Do you see the blinking cursor?

   Customer: Yes.

   Support person:  What does it say above the cursor?

   Customer: It says "blah blah blah...."  (I don't remember the exact words)

Ever try to analyze a dump over the phone with the 'new' system manager, 
you know the one they put in charge, after the last one quit, because that 
person knew how to login and access the application menu?  Seems to happen 
about once a week during these days of dehiring.

Jilly

    There are many parts that are pre-consumed due to the low $ dollar
    amount associated to that part. examples are fuses, cables, fans,
    springs, belts ect... I guess a perception that no parts were used
    could happen if these part were not consumed on that log #. These
    parts are not inventoried, so do they really exist? 
    
    Don't forget the power of waving a rubber chicken.
    
    I guess the list of reasons why could be as long as you want
    
    Trying to fix things over the phone can be tough and nothing is beeter
    then a knowledgable pair of hands & eyes on site.
    
    David
    MCS - NorCal
    

    
    It depends on what's wrong and, like previous replies suggest, who you
    are dealing with. At my favorite <large Aerospace company> I have a
    certain rapport with most customers, and have confidence in what they
    are telling me. And! They *DO* what I tell them! I know that if I give
    Frank Dober a mission, he will report back accurate information. But
    it's taken me 10 years to train him! If I could depend on the guys at
    <large cement plant> to do what I ask, I'd never have to stop in to pop
    open some covers and blow out all the cement dust (There 'ya go, it
    prints again!) Tonight was another good one in this light, Who is going
    to find that pesky SDI cable that keeps the RA81 (you know, the one
    that someone has an exclusive Lock on?) from being used by the rest of
    the cluster? Sure, I can find it, but you sure as hell aren't going to
    get Mr. Customer to drag them cables out of that nightmare under the
    floor.
    
    And the comment about the customers wanting to see a friendly face?
    That's what made DEC Field Service (note use of non-PC
    nondoubleplusgood words) renowned the world over. We took in a
    customer, learned what he used this stuff for, gained a rapport with
    him/her/she/it/they/them and made them feel GOOD that they had some
    broken stuff that we could work on. That is the attitude that is dying
    off today, "Here's your new one, gotta GO!" I've known FSE's who
    couldn't fix much of anything, yet their customers threatened to drop
    service if they lost "their" engineer. Now, with the "new" digital, the
    emphasis is one of "Do more with less, FASTER", and "Why does it take
    this guy 2 hours to fix a <widget> when that guy only take 45 minutes?"
    
    This fixing stuff is not really Rocket Science, but making the customer
    feel good about his maintenance dollar output is. It's a lot like a
    being a salesman, except you also have make something function again.
    
    Enough Raving
    .mike.

    	The question is too ambiguous. We are looking through the worng
    end of the binoculars.....
    
    What percentage of these calls are PL31? What percent is PL07? What
    are the percentages by device type? What percentage of calls were never
    screened? This information is available for the entire US.
    
    PL31:
    
    All of the equipment in this category will have a certain percentage of
    service calls that don't require material. These calls are predicated
    by the design of the equipment rather than our ability to avoid a
    dispatch to remedy an operator error, especially on equipment that
    requires human intervention to operate i.e., tape drives.
    
    Our larger customers demand more from their account reps i.e., customer
    meetings, site audits, etc. All of this time is usually tracked on a
    LARS and closed without parts because engineers need a way to account
    for their time, otherwise it may appear (in some other statistic), that
    engineers aren't being utilized.
    
    PL07:
    
    Engineers are still being paged to replace keyboards, mice, and
    monitors. What is the 'hit rate' of the CRU program? When an engineer
    receives a call for a keyboard, he/she can route the call back through
    the CRU program and we close the call in the field with, guess what, no
    parts used. In many instances, customers weren't even asked if they
    would like to participate in the CRU program not to mention the
    additional burden and cost to the field replace a mouse, etc.
    
    The percentage of service calls closed with no parts is probably higher
    in PL07 because problems are usually the result of a software setup or
    configuation issues. There are a few things the CSC could do to reduce
    these calls......
    
    	1) Measure call screeners (and give them additional tools) on their
    	   ability to avoid sending a service request to the field.  
    
    	2) Change the CRU program to include anything on a PC, including
    	   the motherboard, memory simms, disk controllers, etc. A lot of
    	   customers are capable of replacing these themselves and with the
    	   average one hour of travel time we add to each service call, we
    	   would save money by avoiding a call to the field.
    
    	3) If idea #2 seems like too much of a paradigm shift for this
    	   company, then consider installing removal / replacment
    	   video's in the form of AVI's on all Digital PC's to help 
    	   customers repair their own. In addition, include instructional
    	   literature that simplistically displays removal procedures.
    
    	4) Design our PC's so that everything can be removed without tools   
    
                                                                 
    Regards,
    
    
    GHS
    

>    What percentage of these calls are PL31? What percent is PL07? 
>    LARS
>    What is the 'hit rate' of the CRU program? 

I've been in MCS/Field Service for around 7 or 8 years and I have no idea what
these TLA's are, let alone other non-service types in this conference.  I have
been hearing people mention PL31 and PL07 for years and *still* can't remember
what they mean.

>    GHS
    
gee, you even signed your note with a TLA   8-)

-mark

    
  Internal customers don't typically need to talk to the CSC before 
logging a call.  (That was tried for a while (by edict)  but in general,
internal customers are pretty good at diagnosing problems on their 
own, and calling the CSC for an obviously dead power supply etc.  was
simply a waste of both customer and CSC time...)

  From an internal perspective,  these types of calls generally get
closed with no parts consumed:

  -- Noisy fans/disks,  and fuzzy monitors.   In these cases, the 
	resolution is to log another call when something breaks.

  --  CALLS WHERE DEC DOESN'T STOCK COMMON REPAIR PARTS ON SITE, and the
	field engineer needs to "borrow" one from a less critical 
	system.  (The spare parts situation in ZKO is abysmal. We're 
	starting to keep "spare" (junk) equipment  around, to allow us
	to minimize downtime.)  THIS IS NOT THE MCS ENGINEERS FAULT.

  -- The famous RA82 "reformatting" edict  which required Field Service
     	to try reformatting a failing HDA before replacing it.  (this
	simply guaranteed that the HDA would fail a month later at a 
	less opportune time.)

  -- Memory Errors which disappeared after a reboot and never came back.

  -- Memory Errors where field service swapped the suspected board to
	another array, and requested that a new call be logged when
  	the problem reappeared.

  -- BI Bus flakiness that could be cured by reseating the boards in the
    	backplane.

  -- Calls where the failure isn't deemed severe, and the system manager 
	is unable to give the Field Service engineer the required downtime.  
	(And it is mutually agreed to re-log the call at a more convenient 
	time.)  

  --  Calls where the repair part is extremely expensive, or otherwise
	difficult to obtain,  where the system manager will let the Field
	service engineer experiment and exhaust all alternative workarounds
  	before consuming an expensive part.


  -- Old systems once  MCS diagnoses the problem as non-trivial, 
 	the system manager electes to simply scrap the 
	system (or cannabalize a part from another system) rather than
	demand an expensive repair.  


  In virtually all cases,  SAVING DEC MONEY figures strongly in the
quality of care recieved.  (This is not a slam at MCS -- usually the
remedy is agreed on by both parties.  Internal MCS is very good at
recommending creative ways to "Do The Right Thing" while keeping an
aging fleet of hardware running...)

						-al

Spinning off on Mike Foley's point:

> This fixing stuff is not really Rocket Science, but making the customer
> feel good about his maintenance dollar output is. It's a lot like a
> being a salesman, except you also have make something function again.

  And often, what's most seriously broken on site is the relationship
  between Digital and the Customer. Parts aren't what's needed to fix
  that. *PEOPLE* (and "people skills") are what's needed to fix that.

                                   Atlant

Some recent calls closed with noparts used.

LA424	Had bad ribbon, Customer had just replaced the ribbon.
	I know the printer is in a dirty enviorment, assume that
	the rails need cleaning, so didn't think to ask customer about
	the ribbon

LG02	Intermentt  fault codes, Cleaned contacts on Power Amp Board.

LN06	Had a Paper Jam that the customer could not clear.

LPS20	had a bad OPC Drum, Its a customer replacable Supply

LPS20	Had a bad Paper Jam, customer could not clear

PTX60	VMS needs software patch.

LPS20	More Paper Jams, Customer has a supply of bad paper.

43JR1/4000 model 300	system booted ok when I got on site

952AA	Software, Too many programs loaded in Memory, Needed to go on
	site to see problem. the customer said they were getting
	Memory errors.

RZ73	Bad power connection form the power supply


This is 10 calls out of the last 30 that I've taken so that comes out to
33% with no parts used. 

     Just a few that come to mind.
    
    1) Reseating a module corrects the problem.
    
    2) Install new firmware to fix a problem.
    
    3) Install/Deinstall.
    
    4) PM's.
    
    5) Found to be a software problem.
    
    6) Set-ups are wrong and makes it look like they have a problem.
    
    7) Adjustment fixes the problem.
    
    8) Cleaning fixes the problem.
    
    9) Fix a broken wire or cable.
    
      Mike

Do "expensed" parts get figured into the equation John Rando used to come 
up with his percentage or was he just working off "consumed" parts.  i.e. 
parts with Digital part numbers assigned?

It is my understanding that an expensed part does NOT show up in his 
equation.  the only place that part shows up is in the negative cash flow 
equation that M.L. Krakauer is a bit concerned about.

i.e.  it's a lot easier for the SDU to purchase many parts locally and 
"eat" the expense of those parts.  Since they're not tractable, the call 
appears to be "no parts used".  Is this the case?

Are we actually using non-trackable parts?  I KNOW this company is paying 
large sums of money for these non-trackable parts.  They represent a large 
chunk of that negative cash flow that's recently started to get worse. 
(critical, is probably a better word)

Quite often, the parts sourced locally are in stock (having been purchased 
by Digital at volume-discount prices) but the SDU is looking for a vendor's 
model number (Deskjet) 500 instead of C2106A.  They go out and buy a 
replacement printer and the defective one's future is in doubt.  (it 
usually gets scrapped)

If they'd sourced the part through MCS logistics, they'd have gotten a new 
printer and would have had a repair policy for the defective and it'd show 
up on John Rando's report.  I doubt that the "off the street" one will but 
it'll certainly impact MCS's negative cash flow.

    How about "consumption reduction"? We are supposed to do everything
    we can to save money. Using parts costs money. We in the field have
    become very creative in repairing equipment with no parts consumed...
    from logistics.
    
                                     Steve J.

>    PL07:
>    
>    Engineers are still being paged to replace keyboards, mice, and
>    monitors. What is the 'hit rate' of the CRU program? When an engineer
>    receives a call for a keyboard, he/she can route the call back through
>    the CRU program and we close the call in the field with, guess what, no
>    parts used. 

If the PC is under on-site warranty and the product (monitor, keyboard, mouse,
etc) is a CRUable item, the CSC support specialist contacts the CRU (Customer
Replaceable Unit) group and determine if the item is in stock.  If so, the item 
is shipped overnight to the customer; if not in stock, a service call is logged
to the field.  

I have always speculated that logging a call to the field in the event the item
was not in stock was to  a) provide proper notification to the local office that
the customer is experiencing hardware problems (the field is obligated to know
this), and  b) source the product in question from a local or regional stockroom
(if possible).  What looks as though is happening, however, is that the FE
calling the CRU folks again (mind you, probably a couple of days later), finding
the product now in stock and following the process attempted previously.

What is the CRU 'hit rate'?  Frankly, since we do not receive feedback when
either a customer or a field engineer 'fixes' a problem, I would not even
venture a guess.


> In many instances, customers weren't even asked if they
>    would like to participate in the CRU program not to mention the
>    additional burden and cost to the field replace a mouse, etc.

If logged as a contract call, perhaps (I have no knowledge of this one way or
the other).  But, if logged as a warranty call, not so.  Our PC Warranty
explicitly states what they are entitled to in terms of warranty service and the
PC Warranty Support group here at the CSC goes to great lengths to identify
whether a caller is entitled to CRU.  

While it is in everyone's best interest to 'CRU' a call, there are sometimes
exceptions -- the customer (and the product) may sometimes dictate whether the
CRU process is followed. A customer *can* refuse to participate in the CRU
process, but I believe they are then responsible for Per-Call charges since they
are no longer following the terms and conditions of the PC warranty.


>    The percentage of service calls closed with no parts is probably higher
>    in PL07 because problems are usually the result of a software setup or
>    configuation issues. There are a few things the CSC could do to reduce
>    these calls......

At one time in the past, there was a project between the CSC(s) and the field
whereby FEs would "switch roles" with their CSC counterparts -- FEs would do
support in the CSC, and CSC support engineers would shag calls in the field --
for a period of about 1-week. Since the vast majority of those who participated
had never worked in the other's environment, it was an invariably eye-opening
experience.  

Having worked in both environments during my tenure with this company, I feel at
least semi-qualified to suggest that you have no solid basis for suggesting what
the CSC could do to reduce "no parts" calls. Frankly, it is quite easy to come
up with an idea for how to improve a process -- it is quite different, however,
when it comes to a functioning implementation, especially when it involves
factors not even discussed here (call volume, talk time, avg speed of answer,
etc).  But you can bet-your-paycheck that this sort of thing is, and has been,
monitored quite closely and is constantly being evaluated for improvement.  


>    	1) Measure call screeners (and give them additional tools) on their
>    	   ability to avoid sending a service request to the field.  

You are assuming here that Joe gets an equal number of "no parts" calls as Mary;
that a uniform measurement can be applied so that, if Mary avoids sending more
calls to the field than Joe, she must be doing a better job.  This logic is
faulty.


>    	2) Change the CRU program to include anything on a PC, including
>    	   the motherboard, memory simms, disk controllers, etc. A lot of
>    	   customers are capable of replacing these themselves and with the
>    	   average one hour of travel time we add to each service call, we
>    	   would save money by avoiding a call to the field.

That is ludicrous.  We get calls from people who cannot even tell us the name of
their system (you would be surprised at how many different ways "Starion" is
pronounced :-) much less expect them to begin changing out modules.


>    	3) If idea #2 seems like too much of a paradigm shift for this
>    	   company, then consider installing removal / replacment
>    	   video's in the form of AVI's on all Digital PC's to help 
>    	   customers repair their own. In addition, include instructional
>    	   literature that simplistically displays removal procedures.

Digital won't allow AWSP (Authorized Warranty Service Providers) engineers to
service our equipment without being A+ certified -- you think providing a video
for our customers will suddenly make them qualified to work on them?  This logic
is laughable.


>    	4) Design our PC's so that everything can be removed without tools   

We did this once:  PC100-, PC200- and PC300-series systems.


Sorry to have gotten on my soapbox.  I just felt an obligation to provide an
alternate point of view.

Regards,

-Andy

    
    
    Hi MCS folks,
    
    
    Murphy on MCS:
    
    Improving any operational metric will allways screw at least one other.
    
    By the time you have MTTR under control and reduced avarage repair
    times to 0.2 minutes, you are asked to get Utilization under 
    control.
    Once you have done that, management starts asking for travel expense
    reduction. Why so many miles?  What's wrong with your dispatcher?
    Then it's calls per man per day.
    Average travel times in Canada compared to Denmark seem out of balance.
    Then it's overtime cost.
    Then it's training expenses
    Then you are made to reduce response times
    As soon as you have done just that you are asked why your response 
    times for F-calls (non-contract) are so good. Probably because of 
    too many engineers, right.
    The next thing will be to use only one part per call. How can two parts
    or more fail at the same time. So we invent 'single FRU dispatching'.
    Then you will be asked to investigate why you get returned parts 
    with no problem found in the repair centre. Do your engineers swap just
    for fun? Do you know what that costs?
    Then they tell you that in your branch the avarage repair times are 
    worse then in South Africa. Can you explain why?
    Then off-course your customer survey is not what it should be.
    Then the cost per call or the cost per whatever becomes an issue.
    Then the next quarter you will be asked to reduce inventory and improve
    level of service from the stock-room.
    Then you are asked to improve PM performance (old days)
    Then your MTBF on RZ28-VA's is worse then in Bombay. Why the hell is that?
    Then it becomes very fashionable to focus on escalation and support.
    Then your contract retention and contract penetration is not ok.
    Then it seems that in your team the engineers produce less sales leads
    then the engineers in ISRAEL!
    Then it seems that in your area we see more discounts and a bad DSO.
    
    After a while this whole sequence of questions pops up again and again 
    and again. It has been around for over 25 years and it's here to stay
    for ever to drive you all nuts.
    
    Our metrics are designed so that improving one parameter will allways
    screw another one. This is an invention of a group of highly
    skilled data-base jockeys. 
    
    In short, you are always the looser as a field manager.
    Staff, of which we still seem to have endless amounts, can ask more 
    questions then you will ever be able to answer. As a field manager you 
    are handicapped because you try to run an operation, satisfy customers
    and employees. In the same amount of time you try to answer all these 
    questions and prove they are wrong about your performance because 
    it's all different in your area. 
    You are allways in the defence because somewhere on this
    globe there is a manager who manages that one parameter better then you
    do.....and we will rub it in.
    They (staff) spend all their time in the office watching your 
    performance while you are asked to run an operation with the resources 
    you were promised. Staff is never hindered by customers.
    In the end our overhead will allways prove you are over-staffed,
    over-spending, not in control and don't know the first thing about 
    customers and service.
    
    Take it from me I've been there for many years and gave up.
    
    
    Aad Kooijman
    Motto; Life is hell then you die. 
    
    
    

      Eric, (author of .0) is any information gleaned from this string
    likely to be fed back to John Rando, or other senior MCS managers? Were
    you asked to investigate the problem, or is this something off your own
    bat?
    
      Either way, he ought to see the replies in this topic. It might give
    him a little insight into what life is like in the real world, as opposed
    to the one he sees in stats, reports etc..
    
    Ken

    Bottom line... there is an associated cost to dispatching a call
    without using parts and at 40 to 60 percent of our current call volume,
    this becomes a *huge* expense. These percentages for other companies
    are much smaller (we are viewed as a 'dispatch happy' company). If
    you know someone who works for other companies (especially ones that
    subcontract their service business), ask them how they deal with this.
    
    We also have an opportunity to reduce the amount of 'costly' calls we
    send to the field by expanding the roll of the CSC's.
    
    There will always be a certain percentage of calls that don't require
    material, it's the nature of the business, but there are things we can do
    to reduce that expense in lieu of *other* expense reduction activities.
    At the same time reduce the call load of our engineers.
    
    In response to -16's soapbox reply....
    
    	Spending time as a engineer and unit manager I can guarantee you 
    	that engineers do not typically wait a couple of days to re-route
    	calls to the CRU folks, especially when they sign-in with 5 calls
    	behind their name and one is for a mouse. If the CRU folks find 
    	that the material is not in stock, why isn't that information
    	noted on the call so we don't send it back to them?
    
    	I'm sure the PC Warranty Support group works very hard at
    	determining whether a caller is entitled to CRU. We still have
    	warranty customers who say they weren't asked if they would
    	like to participate in the CRU program (is this because they
    	couldn't pronounce "Starion"?). 
    
    	Why would it matter if a mouse is part of a contract or
        warranty to determine if we can utilize the CRU program? We had
    	a service call where it took two trips to replace a mouse (the
    	first one was DOA). With an average of 1 hour travel per trip,
    	WE LOST MONEY on that call!
    
    	I don't recall any special certification needed to maintain a Digital 
    	PC, so to say that giving the customer the option to replace a memory 
    	simm is ludicrous or creating an AVI is laughable logic is the same
    	mind set that prevents us from being more profitable (have you seen
    	where Digital stock went today?). Believe it or not, there are many 
    	customers capable and willing to do this themselves than to wait
    	for an engineer to schedule service. When you purchase a Micron
    	PC, you get AVI's that show you how to replace anything on the PC.
    	With Windows 95's ability to 'autoconfig', this is an easier task.
    	We cannot stay competetive in the PC business if we maintain this mind
    	set that customers are not capable. Technology and designs are 
    	going to change so that it *will* be feasible and cost effective
    	for a customers to maintain their own PC's (look at some of the
    	latest plug-and-play notebooks). 
    
    	AWSP engineers *can* work on our PC's without having an A+
    	certification. Just because a person is certified doesn't mean
    	he/she is capable of servicing equipment. In some cases it only
    	means that they've taken the practice test enough times to pass
    	the real one. We look at more than A+ in determining AWSP
    	status.
    	
       
    The CSC does an outstanding job in support of customers and the field.
    It's my opinion that those roles could be expanded, given additional
    tools and changing some systemic procedures and paradigms, to help
    reduce our cost associated with taking a service call.
    
    
    Regards,
    
    GHS 
    
   	
    
    
    	 

    Well, -1:
    
    	As someone that works back in the Server support organization,
    I must beg to differ.  Sure the field people are not required to
    be certified to repair our PC's.....I for one believe that this is
    a major mistake and will be attempting to rectify it.  When you see the
    calls coming in (helpdesk and IPMT) on our Server base4d systems, and
    read the notesfiles on all the other PC products you tend to get
    a different view of the world.....
    
    Soap box = on
    
    There are a vast number of questions/problems that glare out at us
    and make us wonder what the hell are these people doing?  I'm not
    stating this as a casual sweeping statement of the field, but more
    and more we see stuff being done that needs to stop if we are ever
    going to get back in the profitable stage again.  If MCS engineers
    are not trained on our products, then we need to make that a
    requirement to service the system......no if's ands or butts.  If
    Neither mMicrosoft, nor Compac nor anyone else that has digital as a
    primary service delivery organization......then why dies Digital not
    have the same  does not allow Digital to work on systems without being
    certified than what gives Digital MCS the right to not train on
    our own products.  I've seen this statement too many times in my
    22 year career....."if the customer can't get a warm and fuzzy about
    Digital fizing their OWN systems, how do you think they will feel about
    the rest of the equipment on contract".
    
    Case in point:
    A recent case that came out way.......ended up with an IPMT case.
    Server system crashing, replace the motherboard, CPU daughter card, Mylex
    controller, 3 RZ disks, a new network card, and a new adapter 2940 ONLY
    to find out that the customer didn't load the tape backup software and
    the software company states that "yup....it'll cause that problem fer
    sure! Here's the correct way to do it..." problem resolved.....
    
    HOW many parts were replaced?  To what cost to digital?
    
    Bottom line: When you send an untrained engineer to a site to fix
    something that causes the customer downtime....and that engineer is
    untrained......YOU (ie: the company sending him) are percieved as
    an idiot, and the engineer feels inadaquate and looses faith in
    the company that's setting him/her up to fail.....and it COSTS the
    company un-necessary dollars for the majority of cases.
    
    Soap box = off
    
    Chet

There are a number of "terms" being thrown around here.  

Consumption

consumption reduction

no parts used

expensed

cash flow

To my knowledge, parts used is based on consumption reports.  Parts used 
means break-fix is working (sort of)

Consumption-reduction breaks the break/fix system.  Perhaps the proper term 
should be "consumption COST reduction".  We still need to consume parts but 
not necessarily higher-cost major FRUs.  There is, however, a break even 
point where the cost to place an engineer on site outweighs the cost of the 
sub-FRU component repair.

Consumption reduction has resulted in a lot of parts being acquired locally 
through LVO as "expensed" parts.  To my knowledge these DO NOT show up on 
the consumption report.  These expensed parts DO, however, show up as a 
negative cash flow for the company.

Negative cash flow and consumption reduction seem to be two of the 
buzzwords we're hearing from management.  My question is:  "Does management 
realize that consumption reduction has adversely affected cash flow via the 
expensed part route?"

The SDUs are buying more and more parts (parts, by the way, for which there 
IS a Digital part number already assigned) locally under the guise of +L / 
expensed / LVO orders.

We're shooting ourselves in the foot, gang.

    RE: -20
    
    	The opportunity to save money is with desktop PC's, I don't think I
    mentioned Servers. 
    
    	I don't understand how a Server trained or certified engineer is
    going to be better prepared to determine that a vendors tape backup
    units software caused a problem. Your IPMT sounds like someone got 
    burned by some vendors software driver, especially when the software
    company confirms the symptom. These situations happen weekly and I hope
    we didn't consume all of those parts. 
    
    	In the real world we can't possibly send a trained engineer on
    every product we support, it would be nice, but it just doesn't happen.
    An engineers technical aptitude has to be something that helps get
    through those service calls. I'm sure many engineers have been in those
    situations and unfortunetly the first call always takes the longest
    amount of time. I'm sure we are not the only company that has these
    issues.....
    
    Regards,
    
    GHS 
       

    Hi agian,
    
    
    My point is that we should look at our metrics in a more holistic way.
    There should be a balance in a large number of parameters. And we should
    manage the real exceptions. After more then 25 years I still see
    questions popping up like "Why is it that your MTTR is worse then in
    Helsinki?". This is a stupid question because you can not compare
    Helsinki with Amsterdam (where I worked as a field manager).
    When you focus on only one metric for example MTTR you will tend to loose
    focus on things like utilization. When you focus on overtime you will
    see other metrics go sour. When you focus on inventory, people start to
    buy parts locally and we see cashflow problems.
    So again I would vote for a more holistic balanced aproach where
    similar SDU's should have similar results within a certain bandwith.
    The problem with these staff idiots is that they have been in the field
    when the PDP 5, 7, 9, 15 and 1120 were recieving hot iron awards. 
    They will allways refer to the days they ran the operation and will 
    never really help you to really solve an operational problem. 
    They will only point it out to you and tell guys in the old boys network
    also with badge numbers <20,000 how clever they are in finding 
    your exceptional performance problems.
    They have all the LARS data of the whole world on their screen while
    the field is solving the day to day problems of customers who pay our
    salary. Customers by the way that are increasingly becoming
    dependant on IT and the service they recieve.
    And only there is where the rubber meets the road in MCS.
    People to people communication.
    My vote goes to the thousands of engineers (remote and on-site) who 
    solve our customers problems and seem to be able 
    to do more with less every day.
    
    Regards again,
    
    
    Aad Kooijman.
    
      

    
    Quick answer: Digital MCSD engineers (and our customers) believe
    		  that the job is to fix the problem.  With today's
    		  technologies, most times the hardware won't be broken.
    
    Recent example:  I take a call on a Starion.  Many hours of customers
    	time spent with support.  Dozens of possible solutions tried.
        We eventually had to swap the system, to placate the customer.
        (as it turned out, we had to anyway, a stripped screw in the
         frame that holds the ISA cards in cannot be replaced in the
         field, according to PCBU).  Original problem?  A game's video
        was playing incorrectly.  Callouts were for me to replace CDROM
        and Telecommander.  Actual fix?  Changing the parameter in the
        Quicktime .ini file to specify driver optimization vs hardware.
        Oh, there were parts galore used on this one, but none needed
        to be (except for that darned stripped screw).
    
    Tell our managers that they have to send us to training.  They'll 
    tell you two things.  They don't have enough engineers now (and
    upstairs tells them they still have too many) and they can't aford
    it (because upstairs tells them to reduce costs).
    
    How might we improve things?
    
    1.  Implement Menu services. Strictly. NOW!  Only those customers that
        are paying for the "traditional" service should get it.  That means
        only a select few can count on having a engineer they can call
        their own and only they get problem fix-type service.  ANY OTHER
        CUSTOMER MUST BE TOLD WHEN THEY LOG THE CALL, THAT IF IT ISN'T
        A HARDWARE FIX, THEY PAY!  [PS - many of our "no parts used" fall
        into this category, see # 2 for possible eventual solution]
    
    2.  Implement a call-handling/problem resolution database with feedback
        and intelligence.  Use it to tie the the CSC's and the onsite engineers
        together, so that we work AS A TEAM!
    
    3.  We must have has a goal to NEVER send an engineer onsite unless 
        a hardware fix is expected to resolve the problem.  Exceptions
        are those services, of course, where we are providing system-level
        diagnosis/fix, such as our delivery of Microsoft's Service
        Advantage.  Note that for their large customers, Microsoft no
        longer makes a distinction between tech support and consulting.
        Currently, the norm seems to be - spend x amount of time trying
        to resolve over phone (if any), times up (gotta keep those call
        numbers that they measure me by up) log it to the field.  Maybe
        there's a FRU callout, maybe not, doesn't usually matter.  Oh how
        I fondly remember the times when our CSC folks were allowed to
        technically screen calls, they were good at it.  
    
    Bottom Line:  Our current state of operations is the price we pay for
        layoff-til-we're-profitable approach.  A better approach might be
        to staff appropriately, train appropriately, give those doing work
        the tools they need to be productive, offer clearly defined
        products and services that are competitively priced, and deliver
        them according to price points.  What good is selling a service
        that you know you are going to lose money on?  Why do we need such
        a large staff to tell us we aren't making our numbers?  Anyone know
        what the ratio of those-who-directly-work-with-customers vs those-
        who-don't in MCSD is?
    
    long winded tonight, sorry.
    Dennis
    

    Ditto -.1

    If we sell goods with 1 year warranty onsite, thats what we should
    provide.  If some things have to be returned then that should be
    explained in the warranty.
    
    later,
    robert

    
    
    re .19
    
    
    please don't expand the roll of the CSC without expanding the manpower,
    training, and tools in the CSC as well.  
    
    We've already 'expanded' ourselves into mediocrity at best and often
    times, abject failure.  
    
    

I seem to recall seeing software services and hardware services having been 
combined a few years back.  Did someone forget to tell Mr. Rando?

It's kinda tough to use parts on a software call.

.28> It's kinda tough to use parts on a software call.
    
    Nope! A couple of years ago a FS person swapped RA82 whenever the
    customer reported a disk related problem (like from ANALYZE/DISK).
    Out of the 100 HDAs swapped by this person only 5! showed a problem.
    Investigation revealed that this person was under a lot of pressure at
    this customer side, had too little training, wasn't told about a proper
    escalation procedure. The only thing this person surely knew was how to
    swap a Head-Disk-Assembly. Investigation showed that he was not to
    blame...but it had no consequences for his managers above him...
    
    Guenther

    
    
    re .29
    
    ...the more things change......

    Of course then there's the one about the FE's in one area only being
    allowed to use one part per call.  When I recommended three in order
    to eliminate obvious future breakdowns on an old machine, he told me
    he would have to log two more calls for the machine. 
    Well, he then makes his one part per call, his meantime to repair is
    great, travel is down per call, is everybody happy now?  What really
    happens is we're wasting his precious time playing the numbers game.
    

    Just for my own education in this space - could you please explain what
    types of parts you could recommend "to eliminate obvious future
    breakdowns"?   Are you talking consumables or real return to depot
    style parts that can be diagnosed in this way?
    
    /Chris.
    

Well, how about the cracked gas spring support that'll eventually break and 
make opening the printer cover difficult?

...or the broken cable strain relief on the printhead cable that'll result 
in a broken wire and a down machine waiting for service.

...or the cracked drum belt that's ready to break and result in a down 
machine waiting for service.

...or the defective slide on the 11/70 system box that'll allow the box to 
slide all the way out onto the floor.

...or...  I think you get the picture.  If you don't, you must be in 
software services, not hardware services  :-)

    I don't recall exactly, we found the printer had multiple problems, 
    the customer wasn't complaining about all of them, only one, but 
    the others would have sooner or later gotten worse, and caused him 
    to complain, why not make it 100% now, instead of waiting.
    
    We used to call this, "the total call concept", looking for other
    problems to prevent future calls.  I'm not sure what You'd call it.
    Times sure change.

    re .33
    
    OK thanks for the elucidation.  It was an honest question, not intended
    to doubt what you had suggested.  I'm not in Software Services, I'm in
    Electronic Services and I was seriously trying to understand some
    "industry benchmarks" that we are being faced with and their relevance
    when a "hardware" call is warranted.  Hence my question to you.
    
    Just to round this out a bit more let me explain.  There is a
    suggestion that as computer equipment goes more and more low-margin,
    & high-volume & commodity then service requests for this equipment will 
    be of the following categories ;
    
    	Configuration : Diagnose with remote support (either CSC or
    			customer self-service electronically or super-smart
    			"service robot").  Possible electronic (i.e. video
    			support) for customer configurable items such as
    			switches or jumpers.  Otherwise send engineer if
    			customer's contract covers this service.
    
    	Software fault :Diagnose as above, and either say "tough" (if it
    			ain't ours ) or schedule customer for next version
    			if it is stuff we are responsible for.  Deliver
    			this electronically if available.
    
    
    	Hardware Fault : Diagnose as above.  Ship 1 (i.e. ONE) predicted
    	(Warranty)	 part to customer if it is CRU, else ship 1 (ONE)
    			 part to engineer to visit customer. 
    
    	Hardware Fault  As per warranty (except the industry trends are
    	(Contract)	showing that customers will simply take the risks
    			here and be prepared to pay per-event and per part
    			consumed).  In this scenario customer will
    			determine acceptability of additional "preventive 
    			maintenance" parts replacement.
    
    
    The rationale for 1 part being that this is all that a remotely
    performed diagnosis can reliably predict.   The rationale for
    remote diagnosis being that it is too costly to have to travel to site
    to perform initial diagnosis for what will normally be a non-hardware
    type of problem.
    
    The rationale for trying to understand "no parts consumed" on-site
    calls being that this should be for calls where the customer is
    paying for the level of service we are delivering and that we should
    develop the smarts necessary so as not to diagnose parts needed for 
    software and configuration problems.
    
    Obviously this is a simple model and doesn't fit perfectly with some of
    the excellent premium level of service delivered by MCS field staff but 
    it just happens to correspond to the market dynamics we're seeing in the
    growth market segments (i.e. servicing PCs and PC servers).

    Maybe i missed it, but i haven't seen one of the very common
    reasons for a no-part call.
    [C] [M] symptom abated after reseating {module(s), cable(s), chip(s).}
    
    It still works with PC boxes.
    
    Sure there are customers who could try this.  There are others
    who have no business with their fingers inside anyplace where 
    electrons ever flow.  Who, other than their regular FE, is in
    a position to make a competant decision as to which type a 
    particular customer is?
    
    [dale]

    re .37 (reseat...)
    
    I charged the State University of New York $65 yesterday to reseat the
    simms in a 466LP box. Took me all of 3 minutes for the whole call.
    It didn't work before, it does now.
    
    .mike.

>    I charged the State University of New York $65 yesterday to reseat the
>    simms in a 466LP box. Took me all of 3 minutes for the whole call.
>    It didn't work before, it does now.
    
And how much did we lose on this call, taking into account how much it costs to
log the call, go through all the overhead of managing the call (dispatching
someone, closing the call, etc), the travel time, etc.

A number of years ago there was a model of the $500, $50 and $5 call.  $500 was
what it took to send someone on site, $50 to handle the call over the telephone
and $5 if the customer could answer the problem themself via tools like DSNlink
or DSIN.

I'm sure the numbers have changed, but even if we cut them by half we're still
looking at a hefty cost to travel to a customer's location.

-mark

    I have to say "Probably Not" to losing money on this call, since there
    was "No" travel time on 4 of the 5 calls I did - this is a Servicenter
    at a University, and I live 8 miles away, went there first thing,
    that's not "travel time". That call took about 3 minutes for the
    secretary (ours) to log, about 3 minutes for me to actually fix, and
    another 6 or 8 for me to close. (Love those slow dialup lines!)
    
    Now if I had been in Liverpool, and had to drive the 40+ miles to get
    to Oswego, then I would agree, that one would be a loser. That's why I go
    there twice a week, first thing.
 
    .mike.

    $65 is the flat rate Service Center fee. It's carry in. Not on-site.
    Also, the Service Center Engineer logs the call. A SIM reseat by
    carry-in is a money maker. Unless, the Service Center overhead that
    month is $4000, and it's the only call...
    
    Desktop per-call is $125/hr, including travel, 1 hour minimum (see the
    "U.S. Systems/Services Price List", January 2, 1996 - page MCS 7.8)
    
    Until the call screeners are goaled on successful diagnoses and not
    number of calls, and until they are required to follow a call to
    completetion after it gets sent to the field, this pipe dream of "remote 
    diagnoses finds problem - engineer goes on-site with one part - call 
    closed in one trip - no calls closed with zero parts" will always be
    just that. A pipe dream.
    
    BTW: I'm not picking on the call screeners - it's a difficult task to
    fix hardware from 500 miles away. 

    >Desktop per-call is $125/hr, including travel, 1 hour minimum (see the
    >"U.S. Systems/Services Price List", January 2, 1996 - page MCS 7.8)
    
    Very few customers get billed at this rate. Most get the $65 (or even
    less!) per hour rate so that the business doesn't go to a competitor.
    There are lots of smaller service vendors in the Houston area that only
    charge $45 per hour. 
    
    One salesperson in Houston even agreed to have MCS provide per-call
    support for a large desktop customer at a rate of $65 per hour (REAL
    hours, NOT 1 hour per PC. If you fix 4 in 1 hour, the total charge for
    labor for that hour would be $65), plus parts (at our cost plus 10%).
    But wait, it gets even better! They also agreed that Digital would NOT
    charge the customer for travel time, even though the customer was 55
    miles from the office... It makes me cringe to think how much money
    we're losing on that deal. :-( But sales gets rewarded for CERTS, not
    PROFIT.
    
    No one has even mentioned yet the cost of a) the time for the FE to
    BOTH fill out the online and the hard copy LARS report b) the time for
    the local admin person to then enter the same data into an online
    billing system (PEARS) c) the time for someone to actually send the
    bill to the customer, then keep track of when and if the bill actually
    gets paid. Plus who knows how many other non-value-added processes
    along the way.
    
    Harry
    

Examples of 'no parts' calls on printservers:
    
    Light print - BA bias's spring loaded contact was jammed in. (usually
    seen after a move or if the customer slammed the drawer shut real hard)
    light print - screw on the front of the drum hub is loose.
    light print areas - drum drive and developer drive aren't aligned,
    loosen developer drive to reseat the two.
    light print areas - customer spilled toner in the drawer, vacumed it
    but toner is packed onto the developer pressure spring area and needs
    to be picked out with a toothpick.
    dirty print - BB bias jammed in (same as ba)
    error 30, 32, - T/S grounding fingers bent, not making contact
    Duplex jams - modify (bend) jam removal plate so the rollers touch
    better and the page won't slip.
    Jamming - bad paper, might have to bring paper to the site to prove it.
    Cassette won't raise - elevate cassette lever is broke, small cheap
    part, many FE's order several and keep them handy in the toolbag.
    Many other small cheap parts are ordered in multiples and kept in the 
    kit (saves a trip to the stockroom).
        
    I could probably think of more, these are just off the top of my
    head, none would likely be fixed by the customer, no parts would 
    have been consumed.

    
    example from standby call last night:
    
    problem:
    
    TF86 would not load tape
    
    diagnosis:
    
    tape label underneath tape cartridge when tape was inserted into
    tape drive not allowing drive mechanism to engage tape cartridge.
    
    solution:
    
    remove cover from TF86, remove tape cartridge and label, re-install
    cover and return unit to service
    

    Three calls of relevance to this discussion which I have dealt with
    this week:-
    (1) On going problem with TA81 "...I swapped the head last week, still
    got the problem..I now need help to decode the tape errors". This
    should have been the other way around! After decoding the errors, this
    pointed to a software problem. Checked the patch level of the
    software...never patched!! Many outstanding patches to apply!!
    
    (2) Customer: "I've got some errors on a disk in a Storageworks array.
    	Me: "Do you have remote access to this system for me to analyse the
    errors?
    	Customer: "No. Our company policy won't allow any remote access to
    our systems. You will have to dispatch an engineer." 
    
    	(This company makes waterproof hunting jackets!)
    
    (3) Customer Call:	Errors on a 3rd-pty Tape autochanger.
    	Me:	"Can I dial in and look at the errorlog to decode the errors
    please?"
    	Customer: "I don't think the errorlogs are available anymore."
    	Me: "When did this problem happen then?"
    	Customer: "Last monday." (five days ago!)
    	Me: "Are you sure that the errorlogs are not still there? You don't
    usually delete them." (Past experience coming in to play here!!)
    	Customer: "I am pretty certain that those errorlogs got deleted."
    (I'm starting to get the message here..he doesn't want to be bothered
    setting up remote access for me, although the system is equipped with
    AES.)
    
    	I finally have to give up and request an engineer to be dispatched.
    
    	The sequel is that I've just had a call from a dispatcher asking if
    I can do a remote diagnosis for this problem as they don't have a trained
    engineer for the product!! I looked up the product support plan for this
    device and there's a pretty hairy commitment by DIGITAL on engineer
    training. As this is a developing situation, I don't yet know the outcome,
    but I can guess!!
    
    	Of these three calls, DIGITAL are about to, or have already
    swapped expensive parts for no good reason. Why?
    
    	Because Managers have failed to implement PRODUCT SUPPORT PLANS.
    
    	These are quite specific and determine the ratio of skilled
    engineers to installed base, the level of training, and the required
    resources.
    
    	There is no real excuse for not implementing these plans. In some
    cases where danger exists, it is a CRIMINAL offence for a manager to
    commit an untrained engineer to a problem situation, especially when
    the support plan highlights the danger. I can't see "ignorance" being a
    good defence!!
    
    			Dave Clark
    			UK-CSC remote support.

    
    Those "Product Support Plans" are developed by "those who know best".
    Then they are mailed (A1'd) out "to the Field".
    
    I get a copy of every one of 'em. One copy from "Back East". A
    forwarded copy from a "Support" Manager. A forwarded, forward from the
    SDE. A forwarded, forwarded forward my boss. At randomn times another
    forwarded * (n)forwards from some other source when the conditions
    haven't been met.
    
    I currently have about 30 or so to scan through, most don't apply to
    this geography, which makes deleting them easier and easier. The ones
    who are suppossed to read and understand merely forward. It IS after
    all, MY job to know all this, right?
    
    I figure that if your name shows up as a "forward" then I suspect that
    you add little value to the food-chain.
    
    .mike.