|★ PEOPLES ★ CPC STAFF ★ The birth of a notion: the CPC ★|
|The birth of a notion: the CPC (New Computer Express)||Peoples Cpc Staff|
Part III: William Poel relates the inside story of Amstrad's move from hi-fis to micros
Amstrad was a successful lowbrow purveyor of audio equipment with more knobs and flashing lights than critical accolades by early 1983. Yet within months it was transformed, and had even won its first Which? Best Buy (for a computer of all things!). How?
A brief resume of the product line after record player tops reveals the firm adopting the Japanese technique of spotting a trend and then distilling the salient features of an expensive and up-market implementation in a product that the lorry driver and his wife could aspire to, and afford.
This is no mean feat, and displays that Amstrad's greatest strength was from the very inception of the business to recognize what the mass market was asking for, rather than striking some statuesque pose behind the veils of "High Fidelity", and ending up in a respected but niche (i.e. small) marketplace.
There was a period around 10 years ago when the hi-fi marketplace burst forth with the "separates" business. The matched tuner, amplifier, record deck and speakers became all the rage, and remained the rage for quite a while as the product evolved through the valve and on to the transistor and integrated circuit. Amstrad produced a range known as the Integra, which employed the features of the desirable but unaffordable, and managed to assemble them together in a box that looked the part.
However, Alan Sugar wasn't really after the accolades and best buy labels from such noble organs as The Gramophone. As ever, he wanted the mass market. So what's a little hum amongst friends? The rustle of Yorkie bars being unwrapped and rumble of lorry engines outside the council flat would probably drown them anyway. And the volume control would only get a bit noisy if you kept fiddling with it...
In other words, Amstrad played the percentages, and quite rightly realised that most users would be willing to put up with a few rough edges when the product was basically good value for money.
Such an approach to business has gradually been made extinct by tightening consumer laws and, more crucially, the public's attitude towards cheap and cheerful is waning, as the yuppie economy emphasises the "real" thing. The main retailers have also obliged their suppliers to tighten up considerably, as the costs of dealing with a returned product from a customer are absolutely vast compared to the margins being made in the sharp end of brown goods retailing.
The market switch from separates to the Tower system was started, as usual, in Japan, where the boxes connected with plates of spaghetti approach (remember the John Cleese ads?) had serious repercussions for the shoebox living style of the average Japanese family. Sugar spotted this on one of his regular spying missions to the Far East, and set about applying the principle.
However, rather than simply tucking co-ordinat-ed separates together in a very tidy stacked audio approach, Amstrad slapped the whole lot together in a single unit, made it look as if it were a set of perfectly aligned separates.
The cost advantage was dramatic: a single power unit, and a single chassis carrying all the electronics. The move coincided with increased automation in the assembly plants, and so a £199 "tower" system did the job of £500-600 worth of separates. This was the product that provided Amstrad with substantial growth around 1982,
The Shoebury plant was also a hedge against possible import restrictions that the possibility of a switch of government might have meant. Also, as Amstrad's remarkable success (relatively) in France has shown, it was a useful card to play in the EEC "where's it made?" game.
However, quality was not a strong sales feature of the Tower family. Frankly, most things made in the Shoebury plant could have benefited from the attention to manufacturing detail that Amstrad's far East partners like Orion and Funai brought to bear on the goods that they made for Amstrad under the Amstrad label. Staff from the Far East were indeed seconded from time to time. But as most British businesses have discovered over the years, it is not possible to expect a UK workforce to display the productivity and reliability of Japanese workforce without total Japanese control.
Sony and Nissan have plants in this country that purport to have proved that the UK workforce can be as good as any. But the management regime applied is definitely alien to British techniques and requires commitment and capital investment that would send the average UK industrialist and his bankers down the club for a few stiff pink gins to recover their composure.
And whilst Sugar is definitely as scornful of the pink gin brigade as the most ardent Trotskyite, he isn't too keen on strolling around the factory dressed in the company jumpsuit, doing the company knee-benders.
The upshot of this is that Mr S. could see that the end of the Tower System boom was nigh when the world and his dog began to pile in with the products that cut his all-important margins tc the less interesting end of the acceptable range.
So what next?
The failings of the rivals
The CB boom was always rightly regarded with suspicion by Amstrad. By this time, it was the most practical UK company as a result of its observations of other markets, and was mindful of the stern advice from its Far Eastern suppliers.
However, as the CB boom loomed (and waned!) the personal computer market became suddenly interesting.
Amstrad noticed that here was a business where the price paid by the customer was about six to ten times the cost of the manufactured product. This implied two possibilities. The charitable option is that the people making the product enjoyed some amazing control over the complex issues of the intellectual property (like Apple managed). Alternatively, they were simply inefficient, needed huge margins to pay for the cost of returns and had a less than perfectly honed approach to consumer goods manufacturing.
The Commodore Vic 20 was at that time the epitome of getting it about right, and so the Amstrad approach was conceived. But bearing in mind the strengths of the Tower HiFi approach, Mr S. decided to lob in the monitor and cassette recorder with a single power supply. And so he could tout a complete design that could deliver far more than the competition who did not have the
Monitor and box were duly tooled up (to a standard that spawned the now famous "pregnant calculator" jibe aimed by Sugar at the hapless Sinclair Spectrum).
Meanwhile, Sinclair managed to ruin the QL with a stylised presentation that was beginning to wear thin even with the Sinclair faithful. What they wanted was a real keyboard, please. Amstrad in its relentlessly practical way, noticed that most of the reported troubles of the competition were due to keyboards and connection of power supply and cassette recorder.
Hang about chaps - what about the innards of this box? Well, that can't be difficult, let's get a bloke to knock one up. It's got to be 8 inches wide, by 22 long with fixing holes at....
Yes. Amstrad had misjudged this end of the project ever so slightly. When the board of the right dimensions appeared around April 1983 (and singularly failed to work) to give Amstrad and technical director Bob Watkins their due, they spotted fairly readily that the person notionally responsible was not capable of the task, and that alternatives should be sought with some urgency.
A bid to take on the task from a components distributor and electronics design house, Ambit International, led to the formation of the new famous Arnold project team, resulting in the CPC464.
Ambit had no direct experience of micro computers, although it had spent a lot of time trying to hawk components to the likes of Acorn and Sinclair, and had identified that in any case this was the business of the blind. But Roland Perry, then of Ambit, knew plenty of men who had experience of micros at a more serious level. Moreover, he could see a constructive role emerging as broker between the down to earth approach of Amstrad, and the esoteric approach to the "posher" end of the office micro business.
In particular, he knew a group of skilled people who had recently left office technology firm, Data Recall - then famed for the Diamond word processing system. These people were Locomotive software and MEJ Electronics: respectively the software and hardware brains behind all of Amstrad's better computer offerings ever since.
The ill-fated microcomputer design was based on 6502 technology ("cause that's wot's in the Vic 20, innit?") and 'parts had ever been ordered against a shopping list.
However, time was marching on, and it was August 1983. A brief tour of other potential talent revealed that the choice of Locomotive and MEJ was likely to produce the results within the time and (outrageously modest) budget.
So a switchover to Z80 was duly made having exhausted the 6502 option, and prototypes appeared in November, after starting the project in August. This was no mean feat, especially since the prototype hardware and software largely worked. The first progress session was quite memorable, as about the only thing that was working available was an ability to drive the cursor around the screen using the arrow keys.
But even this was a whole lot more impressive than anything that had gone before, and so Mr S. and his people gave nodding approval, and heaved a sign of relief that maybe their plastic mouldings were going to be of some use after all.
"Arnold" received the go ahead to be completed with a view to a launch in April 1984. The rest of the world, of course, would know it as the CPC...
The pat leap forward
Having launched the CPC amidst the disarray of his rivals, Alan Sugar had a bigger, better battle to fight. A certain notion of an all-in-one word processing machine was looming...
New Computer Express #5 (12/1988)