This is written for the Professional - software
Developer
Pre-Replication Testing & Tweaking: Make money while saving your
customers headaches.
© 1999, By Terry E. Mercer, Pacific Buyer’s Group
Published by Replication News, 1999
If you create pre-masters or deal with full replication of CD-ROM’s for other
companies, you have a "captive audience." People are coming to you and your
company to get their CD-ROM’s replicated. How many of those CD-ROM’s have
technical problems? I am not talking about a complicated learning curve or difficulty
figuring out how to use the program. I am talking about major problems installing,
uninstalling, or just plain working the basics. Do you care? The client sends you a Gold
– you make a stamper out of it, and hand the customer back either spindled product or
finished goods within the time frame they want. You have done your job, you have serviced
the customer. Or have you?
Have you optimized their chance for success? The more helpful you are, the more
successful they are the more CD’s they will need, the more money they will spend with
you. Have you really capitalized on the profit potential available, by not
fulfilling another very real and existing need? This pre-replication testing service is
something more than 50% of the existing client base (of most pre-mastering contractors and
replicator) need and could immediately benefit from. If even only 10% of a reasonable
customer base can be convinced to use this new and unique service you could see additional
profits within the first quarter, definitely within the first year. Why should this
service exist?
In real life, most programmers and software developers have three serious handicaps:
they are:
1) "too close to the forest to see the trees,"
2) often pushing to meet deadlines, and
3) resources limited to one or two computers & sets of eyes & brains to find
potential problems.
Most large companies have setup different divisions of people for design, graphics
(with program and packaging graphics are usually further defined), development, testing,
and technical support. Out of the thousands of different developers in the world, very
few companies have the luxury of having multiple people testing their products.
The small companies are often so rushed to get the product finished and to market that
they haven’t tested different hardware the program will be run on, or the unorthodox
way end users may try to use the program.
A real life example was an extremely educational "kid proof" program with
incredible graphics that required an adult to execute the complicated exiting routine, by
holding down the Control+Shift+X+F1 keys at the same time. Worse yet, the on-disc
instructions for exiting were inaccessible while the program was running. The instructions
(available only after a hard reboot – and not restarting the program first) further
stated that the purpose of this exit feature was to keep the kids from messing up the
computer. This is a unique concept, but extreme and far too difficult to figure out.
Unhappy parents had local technicians berating the company for their lack of foresight and
for not thinking about the customer. The company lost its reputation and now the retailers
are scared of product returns and are hesitant to touch any of the company’s products
present or future.
Ok, you might think this is an extreme example… but it is not. What about the
high-end game that runs in DOS protected mode, a very common environment, that has
problems with the OS Upgraded versions (runs perfectly on full versions) or the game that
does not allow the sound card to be reconfigured to a different IRQ? How about the mixed
mode CD that will not operate on CD-ROM drives that run through a sound card? Or the
program that eliminates the ability for the Windows 98 to be shut down properly without
reformatting the hard drive, because it over wrote an important DLL or registry entry that
Windows requires for shut down? There are many programs developed that do not take some of
the most simple and obvious things into account when you have to use the program on a
different computer.
Most small companies have only one or two programmers that often eat and drink the
"code" for months at a time. Most developers have a flow chart on paper or in
their head showing how their product will operate. They get in the habit of executing only
a few set test routines on a limited number of system configurations. Eventually, this can
result in missing obvious problems… problems that will ultimately cost them.
Sure, if only 100 or even 1,000 discs are released the problem can be minimized. If the
price is low enough (under $8.00 retail) then most people will not bother to call about
technical issues (especially if the number is long distance). However, that
company’s reputation is now tainted and the customer will think twice and
possibly not purchase another product from that company again if there are serious
technical problems. A prime example is Word Perfect Corporation’s first Windows
(v3.0) version called Word Perfect v6.0. Did you notice how long version 6 was on the
market before version 7 came out? We are talking YEARS – with a wide variety of bug
fixes in between. With a quickly diminishing market share, mounting technical support
problems, and an inability to turn things around, and Corel bought the company for cents
on the dollar. Their problems were extreme, and the best alpha and beta testers
couldn’t get many of the problems fixed – as the program code (and programmers)
couldn’t deal with the shear volume of problems in the early Windows versions. In the
event the customers data integrity is compromised, the customer generally won't ignore the
problem, they will likely complain to anyone and everyone who will listen. These
factors will only make it more difficult for the developer to fund or release the next
project; it will ultimately costs you – the replicator – money, and
very possibly a customer.
I look at about 50 to 200 different programs per year. Some are pretty mundane and
"normal" having a very limited use or are not quite as good as the competition,
while others have a great deal of potential. My job, usually isn’t to do a
competitive analysis of the program or CD, or to develop a sales strategy… it’s
to make sure all of the basics work. The purpose is to reduce
the technical support once the product is released and in the hands
of an end-user.
Some interesting numbers were thrown around a few years ago at a software conference I
attended. They estimated that over 30,000 new programs are created each year, of which
less than 250 new programs actually go to any type of real distribution, and less than 50
ever actually sell more than 500 units. Another interesting set of numbers, while I’m
throwing around numbers, is that less than 1 program in 50 are released to
distribution with any type of third-party, pessimistic testing done… and at
least 30 of those 50 programs have some major problems – whether it is obvious errors
in the program itself or an incompatibility with either an operating system or hardware. Most of these problems can be fixed or avoided quickly and easily
– if they are found prior to release. The cost of software
testing can directly reduce the cost of technical support, returns, and will more than
cover the expense of a lost reputation.
The budget CD-ROM distribution giants, most of which have either died and faded into
distant memories (and nightmares in many cases), helped reduce the price of CD-ROM’s.
This has increased the demand to have the drive as a standard part of every computer
system. Larger, more robust products, required the additional space. As the replication
prices dropped, distributing products on CD-ROM made sense. It didn’t matter if the
program was 650MB in size, or 200K. Why? Because the distribution channels couldn’t
easily erase the disc (like they could with floppies by shipping the product too close to
a magnetic field, like the one found in large speakers). Furthermore, the customer
couldn’t accidentally erase or over-write the disc. A virus couldn’t
mysteriously appear on the disc, with the company (often wrongfully) being accused of
shipping the virus. The move to CD-ROM’s makes sense. The value is assumed. The
problems are reduced. However, with the added size and the more complex programs, we have
to remember that the average human being cannot think of everything, cannot catch every
problem, and will not try to run a program in every conceivable fashion with the same type
of hardware or software running in the back ground.
Personally, I try to have the software tested on every operating system, with a wide
variety of hardware differences, that the packaging (and the program) says it is
compatible with. No matter the subject matter, all software has four primary areas that
must be considered and analyzed. 1) installation, 2) how it is intended to be used, 3) how
it will may be used, and 4) how to uninstall it. During the answer gathering stages, it
would be prudent for the developer to consider how the product compares with the
competition, and the best methods for marketing, packaging, and distribution.
The installation. How will the program function on a system? Where will the files be
placed on the hard drive? Is there a danger that files (mostly DLL’s) will be
overwritten with the program on installation, thus causing another program to have
problems? Having a self-contained program is always the safest… and usually the best.
How will the installation program function under different operating systems? There are
different versions of Windows 95 and at least six different versions of Windows 98. There
are still some people running Windows v3.1, v3.11, v32s, and WFWG. Who will likely be
purchasing and trying to use the program? Will the packaging be adequately and accurately
marked?
If you want more details, have specific questions, or have a need us a help you setup
and establish a testing process, you can email me at: terry@helpus.com
The budget CD-ROM distribution giants, most of which have either died and faded into
distant memories (and nightmares in many cases), helped reduce the price of CD-ROM’s.
This has increased the demand to have the drive as a standard part of every computer
system. Larger, more robust products, required the additional space. As the replication
prices dropped, distributing products on CD-ROM made sense. It didn’t matter if the
program was 650MB in size, or 200K. Why? Because the distribution channels couldn’t
easily erase the disc (like they could with floppies by shipping the product too close to
a magnetic field, like the one found in large speakers). Furthermore, the customer
couldn’t accidentally erase or over-write the disc. A virus couldn’t
mysteriously appear on the disc, with the company (often wrongfully) being accused of
shipping the virus. The move to CD-ROM’s makes sense. The value is assumed. The
problems are reduced. However, with the added size and the more complex programs, we have
to remember that the average human being cannot think of everything, cannot catch every
problem, and will not try to run a program in every conceivable fashion with the same type
of hardware or software running in the back ground.
Personally, I try to have the software tested on every operating system, with a wide
variety of hardware differences, that the packaging (and the program) says it is
compatible with. No matter the subject matter, all software has four primary areas that
must be considered and analyzed. 1) installation, 2) how it is intended to be used, 3) how
it will may be used, and 4) how to uninstall it. During the answer gathering stages, it
would be prudent for the developer to consider how the product compares with the
competition, and the best methods for marketing, packaging, and distribution.
The installation. How will the program function on a system? Where will the files be
placed on the hard drive? Is there a danger that files (mostly DLL’s) will be
overwritten with the program on installation, thus causing another program to have
problems? Having a self-contained program is always the safest… and usually the best.
How will the installation program function under different operating systems? There are
different versions of Windows 95 and at least six different versions of Windows 98. There
are still some people running Windows v3.1, v3.11, v32s, and WFWG. Who will likely be
purchasing and trying to use the program? Will the packaging be adequately and accurately
marked?
If you want more details, have specific questions, or have a need us a help you setup
and establish a testing process, you can email me at: terry@helpus.com
|
