[sdiy] Radio Shack catalogs

[Paul Schreiber]

> Interesting article on the re-branding of Radio Shack:

I probably should not read that if I want to keep my blood pressure down :(

This is what caused the death of Tandy (the design 'arm') and of Radio Shack 
(the retail 'arm') in a nutshell:

#1 - refusal to drop gross margins for computers and related equipment

Every item (and I do mean EVERY item) that was in a RS store was approved by 
a single person: Bernie Appel. If Bernie sais no, it was NO and end of 
discussion (he like to refer to them as "my stores"). And, the average gross 
profit margin in a RS store was 56 points. This is about 4 TIMES what a 
high-end retail store like Nordstrums gets. High-volume retail stores get 
around 4 points. The average of the Fortune 500 WITHOUT Intel and Microsoft 
is like 8 points (these 2 skew it up a bit, they get around 26 points). In 
order to get 56 points (and I think in 1983 they hit like 63 points) you 
have to mark things up 5 or 6 TIMES the cost (it's not 1.58 times the cost).

Bernie didn't care if it was a 486 computer of a blank audio cassette. It 
was 56 points or "Not in MY store!!!".

#2 - a combination of arrogance and stupidity

I have a 500 page book on this alone. A few examples, all 100% true (no way 
I could invent them):

a) we had a "summer intern" that was a Dec grad writing diagnostic code. He 
was VERY bright and a cool guy (he restored Porsche 912Ts). After he 
graduated, he applied for a full-time job. He had interviewed at (of all 
places) a donut factory in AZ but he was from Ft Worth and basically said 
"just match my offer and I'll gladly work here'. The offer was $19K. Tandy 
wanted to pay him $16K. He was like Come on, you know my family's here, I 
like it here, just match it and they didn't. He went to the donut guys.

The guy they did hire was the son-in-law (oh boy) of a factory manager. He 
was from the same graduating class. He was in charge of writing hard drive 
seek and read/write endurance tests (these tests ran for like 3 weeks, 
logging errors). He spent about 3 months on the code (all in Z-80 assembly) 
and it was time for the CDR (Critical Design Review, where in from of the 
entire EE staff you presented the work). He ran the demo, it works 
flawlessly, the screen was logging all these seek requests in nice columns 
with a stripchart showing progress and the hard drive heads were flailing 
all over the platter. When he asked if there were any questions, the head 
EE, without saying a word, reached across the table and yanked the ribbon 
cable completely off the hard drive. His software NEVER BLINKED OR 
REGISTERED AN ERROR. Everyone quietly filed out of the room and the VP was 
trying to be calm said "I think you have a bit more work to do before beta 
testing".

b) we hired an engineer from Texas Instruments that was recently laid off 
(oh boy). He was in charge of a cost-reduced CPU card for out Model 3 
computer (moving to the Model 4 which was the newer low-cost version). He 
did the schematic and because to time we didn't have a CDR, they went 
straight to pcb. When the boards came back, it was totally dead. He spent 
like 2 weeks of 18hr days until we formed an "at your bench" review, where 
someone asked "where are you buss buffers" and he replied "I took them all 
out to save cost" to which we replied "Errr...how do you separate the memory 
access from the video RAM accesses" and there was silence.

c) I designed a video graphics card (pre VGA ICs) that used a lot of 74Sxxx 
logic and very fast DRAMS. The DRAMS had critical drive current and waveform 
compliance specs and had terminated lines. This was to be a 6-layer pc 
board, the FIRST one every at Tandy (they were having kittens over it). I 
spent HOURS writing a detailed parts placement document for the CAD guys, 
with detailed descriptions about trace length, width and such. I left 
nothing to chance.

I got back the prototype boards (all nicely wavesoldered) and noticed all 
the DRAMS, instead of being in  1 long row close together, were randomly 
scattered all over the pc board. Of course th board failed the diagnostics 
(which I had running on wire-wrap). I stormed over to the CAD manager's 
office and he stated that their CAD tool had "minimized all the connections" 
and that "this was the preferred layout". I asked about my 12 page document 
and he said "we didn't have time to read it". So they had to pay a guy to 
lay it out BY HAND using Bishop tape on Mylar.

Paul S.