Voice
Communication in Business Volume 1
Essays on telecommunications,
1969-1980
Chapter 11
Those Awful PBX Proposals:
Station & Maintenance Features
As I sit here writing
this chapter on what you may or may not find in PBX proposals, I
note that we have slogged together through some 15,000 words on what
you ought to know about PBXs and we have only just arrived at
station features. Chances are, station features make up a very large
proportion of your proposals. This is just another clue about what
proposals do not contain.
STATION FEATURES
But what about station
features? Why are they important? There are two reasons. Station
features can reduce the work-load on the console attendant, saving
money on operating salaries, and they can improve the utility of the
telephone system by giving station users new capabilities. It is
tempting to think that single-line instruments with advanced PBX
features can replace key telephone systems. If this were true, we'd
have utopia.
Unfortunately, it isn't.
Secretarial screening is almost impossible without key telephones
and, as we shall see, even when the new features provide useful
functions, they place a heavy burden on the station user.
Fortunately, there is some hope. New telephone sets are becoming
available which insure customer convenience, simple station wiring,
and full access to the capabilities of modern PBX equipment. But to
appreciate these new sets, we have to look first at the basic
station features listed in Table 5, page 42.
Station transfer, hold and conference
This triumvirate, the
backbone of the old "Series 300" feature package, started out to
ease the load on the console and attendant by letting station users
transfer calls themselves. The general method of operation called
for the station user to flash the switch-hook to put the connected
party on hold and get dial tone; another extension number could then
be dialed into the PBX control and a new connection set up for
private consultation. A second flip of the switch-hook would tie all
three parties together for a three-way conference, and then the
transferring party could drop off to convert the conference to a
transfer.
Because these features
were built into the incoming trunk circuits in electromechanical
PBXs, they were only available on incoming calls from the central
office. They were not available on outgoing calls, tie trunk calls,
intra-PBX calls, etc. When system designers suggested that they
could be used to replace key telephone features which work with a
dispassionate uniformity on all calls, many of us laughed. Many of
us are still laughing, but sadly.
Most descriptions of
features tell how the features work. They do not tell how they do
not work. Suppose, for instance, you are trying to do the
consultation hold bit, and the new party you are trying to contact
is busy or doesn't answer. What should happen? How should you get
rid of the consultation connection that obviously isn't going
anywhere?
The answer, you say, is
obvious. Just flip the switch-hook to terminate ringing or busy tone
and rejoin the held party. Great. But note what this implies. If you
want to transfer a call, you have to wait for answer. You cannot
dial the party who is to get the call and hang up. You'll dump the
call. And if you get audible ringing and try to go to three-way
prior to answer, a flash will terminate ringing immediately. Thus
you are stuck until the called party answers.
It is easy to understand
why this sort of design might be done with all-electromechanical
systems such as the 770, but there is no possible reason for the No.
101 ESS, a completely electronic stored program system, to be just
as dumb. And certainly we have a much smarter breed of PBXs today.
In many cases, you flash for dial tone, dial the number to which you
wish to transfer the call, and hang up. That's all there is to
transfer. If you want a three-way conference, just hang on until
somebody answers. If you want to consult privately, however, you
have to put the other party on hold first, usually by flashing and
dialing the "hold" code. When you get the other party on the line,
you can flash and dial the "conference" code. Note that the
assumption here is that transfer and conference are more important
than private consultation; it is easy to transfer and conference,
but a little harder to get a private consultation going. The method
of operation described here applies to the Womack and the Danray, as
examples. But the Dimension is different.
The Dimension still uses
the switch-hook flash to put one line on hold and give you dial
tone. You can then dial up a new extension and consult privately.
However, if you dial a new connection and hang up, the call will
transfer; it will not be dropped as in older machines. If nobody
answers, the calling party will pop up at the console after an
interval of timed but fruitless ringing. All PBXs make some effort
to rescue a call that is ringing unanswered. But many different
designers and programmers have many different ways of doing things.
All PBXs these days offer some form of station transfer and related
features, but few of them do the job the same way. If you prefer one
approach over another, be sure you know exactly how the thing works,
and how it reacts to busy and no answer.
Pick-up
Pick-up is one of the
best of the "new" features. In the guise of the Universal Night
Answer, pick-up has been around for quite a while. But regular
pick-up of one ringing station by another is a little different.
There are two general approaches to pick-up. In the first, various
lines are assigned, via class-marks, to a pick-up group. When any
line in the group is ringing unanswered, any other phone in the
group can come off hook and dial .the pick-up code (often the *);
the call will then be moved to the pick-up phone immediately.
In the second version, a
phone ringing unanswered can be picked up only by a user dialing the
pick-up code followed by the ringing extension number. This means
you have to know the ringing extension number. Clearly this approach
leaves something to be desired.
Some systems seem to do
both "group" and "individual" or "directed" pick-up. Naturally, you
have to have a different code to tell the system what you are doing.
This is particularly necessary when two phones are ringing at once.
If you want to use group pick-up, the system will decide which call
you get when you hit the pick-up code. Thus you must have a
different code if you want to grab a specific ringing line out of
several.
Another variation uses a
dummy pick-up number for a group; this number can, for instance, be
the pilot number of a hunt group. Thus, if you know that some phone
in a hunt group is ringing but you don't know which one, you can
dial the pick-up code followed by the group's pilot number.
Personally, I like group
pick up, and I try to set up systems in such a way that, if you can
hear a phone ring, you can answer it. But different applications
have different needs. The important thing to note is that call
pick-up, as a feature included in the endless alphabetical list in
your proposals, means different things to different designers.
Paging and parking
Paging and parking have
already been discussed as console features. The interaction with
pick-up is vital. Note that when responding to a page, group pickup
may not be as effective as directed pick-up.
Call forwarding
This is one of the most
advertised and also one of the most dangerous of the modern PBX
features. And to complicate matters, there are three kinds of call
forwarding. To understand the implications, let's move ahead
cautiously.
Call forwarding, all
calls (CFAC) lets you divert all your calls to some other extension
on the PBX. To put it into force, you come off-hook, dial the CFAC
code and then dial the extension that is to receive your calls. The
system will sometimes give you a "confirmation" tone to let you know
it understands. To cancel CFAC (and all the other types of call
forwarding), there is another code to dial into the system. Some
systems, however, require you to dial the CFAC code followed by your
own extension number.
There are two basic
problems here. First, the user may forget that CFAC is in effect and
wonder why he isn't getting any calls. Someone else is getting them
and, in most systems, when that someone else tries to contact the
forwarding party, he hears busy tone since his call, like the
others, is forwarded to his own phone which is, naturally, busy.
Sometimes, of course, he may get his own secretary if his line is in
hunt. Life is full of little surprises.
The second problem is
that many people may choose to forward their calls to the same
number at the same time: a conference room, for instance. You can
imagine what happens to the conference, and also to the overloading
of the switch. It is even worse to encourage everyone to CFAC to the
console attendant during the lunch hour.
All this is a far cry
from the rosy picture painted by the salesman about a person having
his calls follow him into an associate's office. The salesman never
explains the annoyance the associate will feel when he gets a "wrong
number" prior to the call-forwarder's arrival.
There are, of course
times when call forwarding can be used effectively. When a person is
away from his phone for the entire day, he might as well have his
calls forwarded to someone authorized to take messages. Similarly, a
person temporarily assigned to another department or area serviced
by the same PBX may simply have his calls forwarded rather than
change his telephone number. But forgetting to cancel forwarding, or
having forwarding gang up on one particular number, can make the
benefits of this feature open to question.
Several PBXs have
modified the basic call forwarding pattern in a most interesting
way. Danray and Digital/Executone let a station user put CFAC into
effect, but only for calls originated by every source other than the
line to which calls are forwarded. This line, and this line alone,
can call the forwarding phone. It all works out nicely. The boss,
who normally answers his own calls, forwards his calls to his
secretary when he doesn't want to be disturbed. She answers and
takes messages. If an important call must be put through, she alone
can call the boss and, if necessary, transfer the call to him.
Call forward on busy
differs from hunting in that the station user can invoke and cancel
the feature at will. This saves programming charges, but once again
depends on the user (and the system itself) remembering that the
feature is or is not in force.
Call forwarding on no
answer is potentially a very useful feature; if a call is not
answered after N rings, either by the boss or the secretary, it can
be moved to the receptionist, another secretary, or a message-taking
facility. In a large system, the use of console attendants for
taking messages is often not practical, particularly if there are no
message-waiting lamps on the phones. Thus someone nearby, other than
the secretary who can pick up the line on key, is a more suitable
target for call forwarding on no answer.
Again, however, we run
into problems. Dimension, for instance, combines call forwarding on
busy and call forwarding on no answer. Under the assumption that
everybody has a single line phone and there is no hunting, combining
the features isn't too bad an idea. But with key equipment to
facilitate the boss-secretary relationship, and with call forwarding
on busy overriding hunting, we get into this sort of thing: The boss
wants to arrange for the message center to get calls when both he
and his secretary are away from their phones at the same time. Thus,
after three rings, the call should be forwarded to the message
center. However, when both the boss and secretary are at their desks
and the boss's line is busy, his calls should hunt to his
secretary's line. No way. You can't have both. Thus call forwarding
on no answer cannot be put into the system and left there, because
the locked-in call-forwarding-on-busy kills hunting. To use the
feature, the victims must program call forwarding on busy and don't
answer when leaving the office, and kill it each time they return.
Separating the features or arranging for hunting to override call
forwarding on busy would be a great improvement. Since it is only a
matter of program, someday it may be done. Then again, it may not.*
[* FOOTNOTE: Call forward on no answer
was finally offered as a separate Dimension feature]
In any event, the mere
presence of call forwarding in a feature list raises more questions
than it answers. And we haven't even considered what happens if a
line forwards its calls to another line which is also in the
call-forwarding mode. Or what happens to call-back queuing when you
forward your calls to somebody else's office. Or what happens to a
hunt group of which you are a member and you invoke call forwarding.
Do hunting calls follow you, or jump around you to the next person
in the hunting list? If you make assumptions about how these
features work, you are almost certain to be wrong.
Executive override
This is a feature that
lets the boss crash in on a connection that is already up so that,
even though the phone of his subordinate is busy, he can get
through. Obviously, only a few people must be given this feature
(preferably none), and the system must know what to do when they
call each other. Further, some sort of warning must be given to the
call already in progress before the new call barges in.
Call waiting
Call waiting is a much
gentler form of executive override. It is similar to camp-on by a
console, but it usually works with intra-PBX calls and sometimes DID
and tie trunk calls. The called party hears a tone in the earpiece
of his telephone set. He then excuses himself and flips the
switch-hook to put the first party on hold and let the new call in.
He can then alternate between the two calls by successive
switch-hook flashes.
This, at least, is one
way of handling call waiting. But it has two problems. First, it
does not allow all three lines to be tied together for a conference
call which may, sometimes, be desirable. And second, it does not let
the called party transfer either call to someone else. Usually you
flip the switch-hook to get dial tone to instruct the system
further. Here, you don't get dial tone, just the other party.
To eliminate these
problems some systems require that the user flash and dial a "hold"
code to put the first call on hold and let the second in. Then by
flashing and dialing 3, for instance, he can alternate between the
two calls. Or, if he flashes and dials 2, he gets a three-way
conference. Or, if he flashes and dials 4 plus an extension, he can
transfer the call with which he is associated at the moment. As you
can see, the user must be quite adroit, and must have memorized many
interesting rules for operating the system.
There are a few other
cautions about executive override and call waiting. If data, fax or
similar nonhuman conversations are being pumped through a
connection, you clearly do not want either executive override or a
call waiting tone to add unnecessary information to that already
being transmitted. Further, there is still the need to be sure that
the warning or call-waiting tone goes only to the person for whom
the new call is intended. And finally, some consideration needs to
be given to the application of call-waiting to a line on hold. If
the system knows the line is on hold, it shouldn't try to apply call
waiting. But with conventional key systems, it has no way to tell.
Automatic call-back
This feature works only
for intra-PBX calls, and its interaction with hunting can sometimes
be a little amusing. Again, it seems to be based on the idea that
everybody answers his own single line telephone, there is no
secretarial screening, and that the person placing the call will be
available when the other call is completed. One may wonder how many
call-back calls can be kept in force by one impatient individual at
the same time, or how long a person may want to wait for an existing
call to be terminated. But the alphabetical lists simply announce
the presence of the feature name, usually under "A" for automatic.
Again the user must be able to inform the system that the feature is
to be put into effect when he finds the called extension busy. This
may require a new attempt with the feature code dialed followed by
the called extension.
Feature codes
We have seen that most
station features require the user to know, and know how to use, a
number of feature codes. Further, the user also has be able to
recognize a number of call-progress tones in most systems so that he
can be sure the system understands what he wants. This conversion of
the innocent telephone user into a telegraph operator poses no
problems whatsoever for system designers, but it offers a certain
amount of difficulty for the sales force. About a week after
cutover, angry users, stripped of their key telephone sets and
confused and annoyed by the system errors often traced to incorrect
selection of the 20 or more feature codes, may be ready for physical
violence. The cheery face of a salesman just may attract a fist.
To minimize these
problems, Rolm began putting the feature codes on the telephone sets
adjacent to the Touch-tone keys, Womack, since marketed by ITT, has
followed suit, and other systems are tending to do something along
these lines. The approach is a good one, and makes use of the system
features much easier. But the lack of visual feedback leaves much to
be desired.
Feature keys
Northern Telecom's SL-1
was the first system to recognize the place of key telephone
features in PBX operation and to try to make a realistic solution to
the problem. Dimension's Custom Telephone Sets have quickly followed
suit, and several other electronic key telephone sets are now
available.
In electronic key
telephone sets (EKTS), there is a voice path and a data path to the
PBX. The sets have a number of buttons with associated lamps.
Depression of the button sends a data message (saying "Button X was
depressed") to the PBX; the PBX may send back a data message saying
"Light lamp Y."
Some of the buttons or
keys on the telephone sets may be associated with extension numbers,
as in key equipment. Others may be associated with features. The
function of each key is identified in the system program. Complete
flexibility is possible and user satisfaction seems to be assured.
Only two or three pairs are required to each telephone set, even if
the set looks like a 30-button call director.
In any event, a station
user can put a call on hold by depressing the hold button. He can
pick up a call in his pick-up group by hitting the pick-up button,
which may be flashing to let everybody in the group know that there
is an unanswered call somewhere, even if they can't hear it. He can
select the kind of call forwarding he wants to put in force and, by
having a visual indication, be reminded that his calls are going
somewhere else.
Line pick-up keys
In addition to a general
pick-up key, EKTS have individual line pick-up keys. Lamps
associated provide distinctive signals so that a busy, idle, ringing
or held line can be identified. Learning to use an EKTS poses no
problems for anyone who has ever had standard key equipment. And the
added features at the touch of a clearly labeled button make full
use of the PBX system easy for the first time.
EKTS are still quite
new, and their full capabilities have not even begun to be explored.
For instance, the system can be programmed to permit the phone to
select the main extension of a user automatically when he comes off
hook, eliminating the need to select a line by pushing a line key.
Or, if the phone is ringing, picking up the receiver will deliver
the call, again without making a selection. Or, the old standby can
be provided: you get the line you used last.
Other possibilities are
available. For instance, the system can send a digital signal to
tell the boss's ringer to sound but not the secretary's, even though
the key associated with the boss's number flashes on both phones.
Then, after N rings, if the boss does not answer, the secretary's
phone will receive instructions to sound off and the boss's phone
will become silent.
There are other fun
things. An extension number in use can be made to provide a busy
indication at all appearances, as in existing key telephones, or it
can be made to appear busy only at the phone using it. What good is
this? Suppose you have DID into your PBX, and you want to have six
people handle questions about your new stock offering. You list only
one telephone number in your advertising, and have it appear on each
of the six phones. When a call comes in, all six flash and sound
off. One person answers, and the ringing signal stops at all the
other phones. A second call comes in, and the five remaining phones
show a visual indication of ringing along with the sound. Six calls,
in this illustration, have to be in progress to the given number
before calling users will receive a busy signal. Why not use
hunting? The system would require a different number for each phone,
and it would decide who got the next call. Most hunting schemes do
not deliver calls uniformly. With the multi-appearance approach, all
the answering crew members can see a call is coming in and anyone
can select it. This prevents a call from ringing unanswered at a
momentarily unoccupied station.
Note that this approach
isn't good for call screening. A second party cannot pick up the
call by depressing the line button, since the call is on one phone
only.
Observe that the
Dimension Custom Telephone Sets and SL-1 sets are quite different.
Some of the features described above come from one, and some from
the other. But the fundamental differences are as follows. SL-1 uses
digital signals for all signaling information between sets and PBX.
This includes the dialing key-pad that looks like a Touch-Tone pad.
Digital signaling is good in that it is faster than Touch-Tone (a
skilled piano player or typist can easily out-dial Touch-Tone on
familiar numbers), but, in the event of a power failure, a digital
set cannot be connected into a conventional central office. On the
other hand, use of traditional stupid/smart supervision, as with
Dimension, means that Touch-Tone digit receivers must be provided,
and must be associated with a given line before any numerical
information can be entered. Dimension sends digital signals from its
feature/line-select keys, but regular Touch-Tone from its key pad.
The "recall" button, the functional equivalent of a switch-hook
flash, is also digital. But you have to get dial tone and then key
in numbers.
Dimension allows some
features to be placed on keys and others to be accessed by hitting
the recall button and dialing a feature code. SL-1 lets you have
only the features on the feature keys. But the whole field is brand
new. Whatever comes along in the future can easily be met by program
changes in most instances. Hopefully, program changes rather than
wiring changes, and simple, uniform wiring to each telephone,
regardless of the number of buttons it contains, will improve
service and reduce costs for everybody.
Abbreviated dialing
Abbreviated dialing is
another feature that is often publicized beyond its utility. With
abbreviated dialing, the system stores telephone numbers in its
memory, and users invoke these numbers by dialing much shorter
private codes. "Hot line" is sort of an abbreviated form of
abbreviated dialing. One called number is stored with a hot line,
and when the phone comes off-hook, the system pretends that stored
number has just been dialed. Hot lines often connect the caller to
the console, but they can be arranged to connect to any internal
extension or external telephone number. That is, they should do this
when they are provided at all. Note that fixed night answer,
discussed earlier, is a form of hot-line that connects a specific
incoming trunk to a specific extension. "Direct in" lines work the
same way.
True abbreviated dialing
comes in two forms: private list and common list.* With a private
list of numbers for one extension, that user can set up his own
frequently called numbers. He has his own private section of the
system memory which he can use as he pleases. He may, as a matter of
fact, have numbers in his list that many others have in their lists.
[* FOOTNOTE: Sometimes it does not
come at all as in the Dimension 400 without Feature Package 4.]
Common lists save a lot
of memory, and in many ways are more desirable. Frequently called
customers, suppliers, or other contacts can be given numbers that
look just like internal extensions, and all users can dial them
quickly and easily. The first digit is an access code, and the two
remaining digits (matching a three-digit PBX) select one of a
hundred numbers.
With private lists, the
user has to remember two numbers: one for use at his own phone, and
another for use at any other phone. With common lists, the
abbreviated dialing numbers are the same, at least from every phone
that has access to the list.
As mentioned in an
earlier chapter, it has always seemed to me that abbreviated dialing
common lists should work very much like trunk groups when it comes
to restriction. Giving restricted station users access to the
abbreviated dialing lists would insure only business calls from
these extensions, even though such calls might have to go to
specific phones all over the country.
Swapping one number for
another, even if one is much shorter, is not always a good buy. When
a relatively small list is required, feature keys on EKTS can be
used. To reach a given number, simply hit the feature key that tells
the system to make a connection to the desired number. This
"Touch-A-Matic" approach may well be quite widely used at some time
in the future. It may not be substitute for a large common list, but
it beats small individual lists in terms of convenience.
It should be noted that
no law says feature keys have to send digital signals, and no laws
say that abbreviated dialing numbers have to be telephone numbers.
With inexpensive memory in each telephone set, depression of a
button can cause a feature code to be outpulsed as easily as a
ten-digit telephone number. The set would have to first send a
switch-hook flash, then detect dial tone, and then outpulse. This is
fairly easy to do and, if made into a package, electronic key
telephone sets could be available for regular PBXs. Maybe this will
happen someday.*
Comments
There are quite a few
more station features, but these are some of the more important. We
have seen that all have some utility and some booby traps, but that
future developments giving consideration to modern technology and
human factors can do wonders. My feeling is that the system should
be programmed, not the users. I cannot take seriously systems that
require users to do all the work while the system seems to lie in
wait to deride their goofs. We are just beginning to explore
possibilities. In a year or two, user-related features will be
something quite exceptional, and much easier to use than the
electromechanically limited features presently available.
MAINTENANCE FEATURES
One of the best things
about step-by-step systems was the way you could track down troubles
manually. Hot relays, stuck selectors and the like were easy to
find. But one of the worst things about step-by-step was the
frequency with which such troubles turned up. Modern electronic
equipment is much more reliable than the older systems with moving
parts, but troubleshooting means must be available since visual
observation is, in general, fruitless. See Table 6, page 42.
Trouble alarms
Major and minor alarms,
often part of the console, have been mentioned. Most telephone
systems from time immemorial have provided such alarms, and they are
the minimum one should expect. Some sort of audible signal for major
alarms is also quite desirable.
Many switching systems
for PBX use, built around standard computers, have input-output
buses that can serve teletypewriters, CRT terminals, etc. Often an
RS-232 interface is provided. Such instruments permit not only alarm
information to be displayed, but allow diagnostic information to be
entered as well. Teletypewriters may be old fashioned, but they do
leave a permanent record that is particularly useful after a system
failure.
Maintenance consoles
Although teletypewriters
and CRT data terminals are used for maintenance access, program
changes, system update, etc., some systems use special consoles with
specialized input keys and limited but adequate output. The MAAP
panel on Dimension is typical, and the MAP panel for the Siemens
SD-192 is similar. The Siemens system is particularly nice in that
it has enough alphanumeric capability to cue the maintenance man
through a great many routine operations in the system in plain
English.
Internal diagnostics
When you can't see what
has gone wrong, trouble- shooting requires special means for giving
insight. Most PBXs have some sort of trouble-shooting programs built
in. That is, they are stored on a magnetic tape where they are
available for use, but they are not generally kept on line in the
system memory. Off-line memory, which can be loaded for testing in a
matter of moments, allows some fairly sophisticated testing to be
done. Since maintenance is usually carried out by the vendor under
contract, the nature of the maintenance programs is often of little
interest to the customer. However, companies planning to provide
their own maintenance should go into this area much more thoroughly.
Testing connecting circuits
Means should be
available for testing connecting circuits: tie trunks, FX lines,
WATS lines and central office trunks. Making outgoing tests isn't
particularly hard. To test central office trunks, the trunk should
be seized, dial tone detected, a digit sent, and removal of dial
tone verified. If tone outpulsing is used, a different digit should
be used each time to make sure that all frequencies are being sent
out properly.
With tie trunks,
outgoing tests can be analogous, but it helps if there is a standard
tone at the connected PBX so that loss in the return direction can
be measured. To test both directions of transmission, the far end
should be able to test all its trunks outgoing, too.
Testing incoming
circuits is something of a problem. As discussed earlier, if you can
busy out all circuits in an incoming group but one, the far end must
seize that one when you place a test call. Such tests, obviously,
can only be placed after hours when traffic is minimal.
Note that automatic
testing of connecting circuits is not something that many PBXs (or,
for that matter, central offices) do very well, if at all. Such
testing is highly desirable, however. When you're paying $17,000 a
month for each of three WATS lines, you should have some means of
being assured regularly that they are all working.
No-test and no-hunt connections
When testing both
extensions and trunks, it is often necessary to get one particular
circuit and test it. If that circuit is a "permanent signal" (one
that has been off-hook for some time with no other circuit
connected), it must still be possible to connect to it to see what
is going on. "No-test" refers to making a connection without testing
for busy. A no-hunt connection lets you pick a specific trunk or
line in a hunt group and connect to it without hunting to some
adjacent circuit. Many systems have provision for such connections
on a manual basis, and at least give you a chance to track down bad
central office trunks and off-hook station lines.
Remote testing capability
PBXs using CRT and TTY
peripherals for maintenance naturally bring up the obvious question:
can such devices be located remotely? The answer is yes, and the
possibilities are very important. If a modem associated with a
dedicated CO trunk can be dialed up from a remote test bureau for
routine maintenance, station, feature and class mark changes and the
like, a member of a centrally located corporate telecommunications
group or vendor maintenance team could provide most of the
customer's service needs instantly without having to drive to his
location. And even when a trip to the location is necessary, a
complete diagnosis before leaving the test bureau is possible so
that the right parts and test equipment can be taken along.
Going one step further,
remote access allows the manufacturer to reach any system he can
dial up, making the designers and other factory experts available
for help if the going should really get rough. Several levels of
back-up via instantaneous connection through the public network can
be a potent sales argument, and can do a lot to insure proper
operation of a system.
Note that such
capability, although it poses no particular problems to implement,
was not generally available when this article was first published.
By 1980, it was almost commonplace. It is something to demand when
writing specifications, and something to look for when going over
proposals. This feature alone may be the most important of any you
look at.
Customer test access
In any telephone bill,
there is a line named "Other Charges or Credits," or OCC for short.
OCC represents station moves and changes, and various other one-time
charges for keeping the system running. Maintenance itself is
something quite different and comes with the service, but OCC can be
startlingly large.
Vendors of
interconnected systems are beginning to standardize their rates for
the equivalent of OCC. With computer control of the system, feature
changes, number changes and the like can be carried out quite
easily. If a charge is involved for such relatively routine items,
the question arises as to whether or not the customer's staff cannot
do the job.
There are many things
having to do with the structure of the operating programs that the
customer probably should not touch unless he has a very well-trained
staff indeed. But most of the items equivalent to moves and changes,
which simply alter parameters and not programs, might well be put
under customer control.
This is another item to
demand in specifications, and to study carefully when proposals come
in. What can you do with your own staff, and without payment for
each item? When you do your financial analysis, the equivalent of
the OCC item must be considered very carefully, and a good proposal
will give you some idea of what you can expect to save. There are
few good proposals today, but at some time in the future, things may
be different.
CONCLUSIONS
Now that we have gone
once lightly over the material a customer should know something
about to make an intelligent choice in PBXs, it is likely that
others have joined me in my annoyance at the incomplete, inaccurate
and generally unsatisfactory nature of the PBX proposals available
today. When you know what to look for, you become painfully aware of
what you don't see. And when you see all the buzz words thrown
around in casual disarray in place of hard facts, you know that
somebody doesn't know his business.
There are, of course,
many other points that may need to be discussed in special
applications. Obviously, some people will have different ideas about
what is important, and many will feel that some of the things I have
stressed are not as important as I indicate. However, the reader who
has stuck with me all the way through is surely in a better position
to make up his own mind than he was before he started.
These chapters have no
intention of telling you which PBX you should buy. Their purpose has
been to alert you to some of the more important questions you should
ask before you sign on the dotted line. Once the machine is yours,
you have to live with it. In such a marriage, love is not enough.
***
The
original articles have been updated somewhat, but a lot is being
done in the PBX field these days. For those who want up-to-date
information, I suggest the use of the BCR Manual of PBXs. The
manual is based on the seminars I have done over the years for BCR,
and includes the current version of the data sheets prepared
originally for the same seminars.
[
Top ] [ Next ] [
Table of Contents
] |
