The worst technology products that succeeded
Mark Pickavance examines classic technological products that succeeded in spite of themselves
When a product succeeds, most pundits like to pick apart the design and focus on those things that enables it to rise above its fellow products and soar to new commercial levels.
Some products just tick all the boxes, and in doing so become ingrained in our society. Those brand names that become synonymous with the devices they make, like Hoover and Aspirin.
Yet, I'd contest that some devices achieve this notoriety by other than being good products, either by a quirk of fate or the failure of others. Some of these are massively successful, to the extent that most homes contained them at one time or another. Here are my top technologies that didn't actually deserve the success they got, be prepared to strongly disagree with my very personal choices.
The Compact Disc
Most standards are the invention of a single company, who then try to convince others to use it, but not the Compact Disc. It was actually a cooperative venture by both Sony and Phillips, who'd dabbled with digital laser disc technology in the seventies and first showed off the Compact Disc or CD in 1982, although they'd demonstrated similar technologies at least six years earlier, and most of the concepts contained within it were patented in 1965.
So why is it a poor product, considering that it entirely dominated the world of distributed audio for at least 20 years?
Well, and this is a personal view, the trouble with the CD is that it was designed not based on what could be technically done, but on some entirely arbitrary choices by those wishing to market it.
They dictated the physical size and the amount of music stored on the disc, the disc could easily have been smaller and the running time of 73 minutes is just 'odd'. I've read various versions of why it is 73 minutes long, including the old chestnut that it's the length of Beethoven's 9th, except that's about 65 minutes long.
Whatever the real reason, it seemed a sufficiently awkward length to make double albums a possibility and the price premium that they demanded. CD was also meant to be better than Vinyl, although in terms of sound quality it isn't, and more robust.
That last point might have some credence if CD makers understood the chemistry of placing aluminium inside plastic and the detrimental impact of the metal oxidising had on the audio they stored. Minor scratches are supposed to be error correctable, but often aren't and early players could often be picky about which discs they'd play.
But for all these faults the one that really gets me is the layout of data on the disc, where no thought whatsoever was actually put into how the disc should be organised. As such they never even considered putting a block of bytes at the start of the data to indicate the name of the disc or the artist, or anything about the manufacturer. That was stupid.
Even if they'd done this just to help disc makers make a simple machine that could sort media or print the right design onto the right disc it would of made sense, but the audio data is basically in one giant block per track, and it only supports 99 of those per disc. What's really annoying is that on the lead-in area of the disc there is a 5K block of unused space, but they never considered putting anything useful there until much later.
They added the capability to address this point with CD-Text, but that took until 1996 to appear, and relatively few discs used the specification and even fewer players supported it. By then the CD with all its obvious faults was here to stay, until the digital download eventually started to undermine it.
In retrospect the CD didn't really demonstrate what technology could actually do in 1982, was built primarily as a marketing tool and succeed despite the extra the music industry wanted for technological 'development' (despite them not actually contributing to that cost), and lacked any strategic thinking about what people might actually do with digital music ultimately.
Ironically, Sony came up with a format that addressed many of these issues, Mini-Disc, but people hated those and refused to give up their beloved Compact Discs.
Iomega Zip Drive
There was a time, in the mid-nineties when every PC owner I knew had and used the Iomega Zip. On the face of it these little blue boxes seemed a remarkably good idea, as PC owners had outgrown the pathetic amount that a floppy disk held some years previously.
The Zip came in both internal and external versions, and it could connect to the Parallel port, SCSI, IDE and later USB, depending which one you bought. A Zip disk could hold 100MB, a good sixty times the capacity of a floppy and Iomega already had a reasonably good reputation after they'd made a cost effective Bernoulli Box that many people liked. They launched the Zip in 1994, and within a year it was a huge success. The drive cost about $200, and each 100mb cart about $20, which considering the hard drive prices at the time, seemed like a bargain.
It was only at the pinnacle of its success that suddenly people started to notice how unreliable a technology in the Zip Drive was. The drives died, the disks failed, they were very slow on a parallel port. But the really annoying failure was the 'click of death'. The technical explanation of this was that the drives could easily develop misaligned heads, usually caused by either the build-up of oxides from use or just from being handled roughly. The heads also weren't magnetically shielded, so coming near any sort of strong field could also do for them.
If for whatever reason the special low level formatting on the disks, called 'Z tracks' wasn't readable, then the disk was toast and so was any data on it. The ability of the drive to go wrong in this way and take some of the data stored on the disks with it was massively infuriating, and eventually lead to a class action lawsuit in the US against Iomega. The case was eventually settled, and all those affected got a rebate they could use against another Iomega product. Sadly, Iomega then went on to make the Jazz drive, with a bigger capacity, but exactly the same point of failure.
What ultimately did it for the Zip was a combination of bad press, poor customer experience and the rise of the inexpensive CDR media. When it was launched it very effectively filled a niche that had been created by the growth in storage, but the success of it was rapidly curtailed by lack of reliability.
Iomega went on to launch other removable media products like the Clip, but they never approached the success of the Zip, which by 2003 had shipped some 50 million units. These days Iomega concentrates on hard drive and SSD storage, where the click of death is a much less likely scenario.
These days it is hard to imagine not having a PC with USB ports, because without them you'd have a tough time plugging anything into them. But back into the mists of time, the PC had a vast range of connection options, most of which involved horribly complicated cables, unreliable and proprietary technologies.
Printers connected with parallel or serial cables, sometimes with odd wiring, and sharing the device to a machine alongside required a nasty switchbox that resembled the hairstyle of Medusa inside.
When USB appeared it looked amazing, not least because you could plug things in quickly and easily, and systems came with multiple ports. Except what's easy to forget now was the first few years USB was a nightmare.
USB 1.0 was the release version, and that had bugs that made it almost unusable. The first one that actually worked for a fashion was USB 1.1, and it was the release that most people encountered. It had two speed modes 1.5MBit/s 'low' and 12MBit/s 'high'. These days we'd wonder what slow speeds would be useful for, but then they only considered USB as a connectivity system for mice, keyboards and the like, and not really for external drives. That subject was covered by 1394 Firewire, which Apple had made so expensive that people mostly chose to forget the idea.
However, the real problems with USB started when devices started drawing power from it, something that was allowed to a degree within the specification.
As I recall some motherboards reacted very badly to some devices, as connecting them by USB could easily make the PC reset. This was because the USB power was 5 volts, and the CPUs of the era also required this line to function. A sudden drop in the level of the 5v line and everything would crash spontaneously.
It didn't help that the early USB drivers were also poor, and the implementation of the port on some systems equally so. The now defunct ABIT made some Athlon boards that would fall over just at the suggestion of a USB connection, and when you combined this with Windows 95 Service release 2.1 or 2.2 you were asking for trouble. It wasn't until Windows 98 and USB 2.0 that this port became the connection that works reliably.
I can't entirely blame those that brought us USB, because many of the problems associated with the early versions were due to the port being used in ways that it just wasn't intended. If it had been restricted to mice and keyboards then USB 1.1 would have been fine, but once the connectors became cheap many people started to exploit the port in increasingly unrealistic ways.
USB eventually grew up and matured into a technology most people like, but given how it started out it is very fortunate indeed that it ever got so far.
The Sinclair ZX Spectrum
I can see a few raised eyebrows at this inclusion, but let's be honest here as important as the Spectrum was to the history of British computing, can we really suggest it was a 'good' product?
I'll defend myself from the inevitable Letter's Page onslaught by saying that I owned a very, very early spectrum (board ID number 1023), and it didn't live long. That was a 16K light grey keyed model, and Sinclair kindly replaced it with the 48K model that did last a bit longer. But at the same time I also had a BBC Micro B, and the difference in both design and construction quality between these devices was chasm-sized.
But then the BBC B cost the best part of £400 I recall, where the 16K Spectrum was just £125. And, in the cold light of day, that was what the ZX Spectrum had, it was cheap. The list of failings were many, however, including a keyboard that defied you to type at any reasonable speed, a character based graphics system that limited the number of colours in any 8x8 block and nasty sound that required processing power to operate.
And, with the exception of the ZX81 compatible rear edge connector, it was by no means simple to attach other hardware. No parallel port, no serial port, no joystick ports, and only tape cassette as a means of loading and saving data.
In many respects it wasn't much of an evolution from the ZX81, and it's wobbly RAM pack, better if only because it could support a display and compute at the same time. For those who think the Spectrum was entirely new it's worth decoding the firmware, which you can find online, and discovering many old ZX81 routines, many of which aren't even called, still taking space.
Sinclair did make a reasonably solid choice of processor for the Spectrum, picking the excellent Zilog Z80, and they could market their product well, it should be said.
But the truth, if such a thing exists, is that if the Commodore 64 or BBC Micro had been a little bit cheaper, the ZX Spectrum would have not sold the 5 million units that Sinclair made (or Timex, actually), and the topology of British computing would be radically revised.
Some of the later models sought to fix issues like the keyboard, but they never actually achieved the success of the original bare-bones design. The ZX Spectrum was a remarkable success, despite not being well made or especially well conceived.
It's hard to imagine now, but before the Internet existed many people used dial-up bulletin boards. This is sort of how America Online (AOL) started, but it soon progressed to be an Internet portal where users were shepherded like small children through the vagaries of the world wide web.
For the enjoyment of doing this users paid both a flat monthly rate and an hourly fee, though AOL didn't actually have the infrastructure to actually support the number of customers they'd attracted causing massive service disruption.
Despite very poor levels of user satisfaction AOL still managed to attract huge numbers of customers, enticed by widely distributed installation discs, and by 1996 they had a user base in excess of 10 million people.
Having grown so rapidly AOL became a take-over target, and in the year 2000 Time Warner offered protection in the form of a merger, and between them created AOL Time Warner Inc. The peak of 27 million subscribers occurred in the middle of 2002, and from this point onwards it collapsed. At the time of writing AOL still has 3 million customers, but most people moved away from the quirky interface and middle-man application with the onset of broadband.
The AOL Time Warner merger is now seen as one of the worst business decisions of the modern era and AOL as a concept that was outgrown by its customers. For those that used the product it was expensive, difficult to manage and a massive drain on computer resources. It's success entirely predicated on the customers not understanding the technology, and believing foolhardily that AOL was there to help them and not gouge their pockets. That it succeeded at all is a testament to their marketing skills and not their technical abilities or customer service record.
Internet Explorer 6
When Microsoft woke up to the idea that the Internet might be useful, it set about attacking any company with important software to exploit it, most notably Netscape.
Internet Explorer was the browser that came with Windows, and by version 6, released just after Windows XP in 2001, they entirely dominated the browser market.
At its peak it's been calculated that IE6 had 95% of the entire browser market, even more amazing when you consider that it's now considered one of the worst pieces of software ever written.
What's also rather astounding that a full ten years after Microsoft released it they were still issuing vulnerability patches for the code which had more points of potential exploit than a Swiss cheese. What really annoyed people however, was the total lack of immediacy with which Microsoft went about fixing known exploits.
They argued that with 234 distinct releases of IE to contend with the testing phase of any update was excessive, but users just noticed their PC being overrun with malware. Somewhat undermining Microsoft's view is their resistance to replacing IE6 entirely, as with more than 90% of the market they decided to invest effort elsewhere. It took IE7 at least five years to appear after IE6, and as a result the browsing public began to explore other alternatives.
In many respects, the true value of IE6 was to vastly accelerate the take-up of Firefox and Chrome, as people thought they'd avoid another Microsoft browser. In the end Microsoft withdrew support for IE6 while sovereign nations told their citizens not to use the browser. The only government that refused to stop using it was the UK, when it was petitioned by the populous to do so in 2010.
In short, IE6 was the most successful browser in the history of computing and you'd be mad to still be using it.
Here are some of the other poor products that sold well;
- Syncronys SoftRAM
This company managed to convince 700,000 people to buy a piece of software that theoretically doubled available RAM, yet did very little that couldn't be done by adjusting the cache size manually. When this became public knowledge bankruptcy beckoned for Syncronys, but for a short while they basked in having a big selling product.
- Windows ME
After Windows 3.x was released Microsoft was on something of a roll, delivering better and better versions of Windows with exciting new interfaces that got the users excited at upgrading. And then the end of the century came along, and for no good reason Microsoft took Windows 98 and butchered it into Windows Millennium Edition, and managed to upset the huge number of people who paid for it. Software that wouldn't install, wouldn't run or crashed were the signatures of the first big version of Windows people rejected.
- Ashton-Tate dBASE IV
When most computers used DOS, companies like Ashton-Tate ruled the software roost with revolutionary products like dBASE. They controlled more than 70% of the PC database market with dBASE III, and then they launched dBASE IV and sold lots of copies of that too. Except it didn't take people long to realise it was a complete dud, and they moved in their droves to rival Foxbase. It was the end of Ashton-Tate, and they eventually got absorbed into another rival, Borland.
- IBM Deskstar 75GXP
At the point in 200 where IBM launched the Deskstar 75GXP they has a very high reputation for drive reliability with the Deskstar brand. They sold millions of these drives only for them to start suddenly dying in huge numbers. This ended with a class action lawsuit and IBM giving $100 every customer with a dead drive and probably convinced IBM to ultimately leave the hard drive business and sell that division to Hitachi. Although ultimately made reliable, in computing lore the 75GXP will always be remembered as the 'Deathstar'.
- Nokia N-Gage
Before Nokia entirely lost the phone plot they tried all manner of radical ideas with their phones, one of the oddest being the gaming phone, the N-GAGE. Initially reviews seemed positive, and Nokia claimed that it had shipped 400,000 devices in the first few weeks of launch. Later these numbers got massively revised, when it was admitted that these were the number sent to retailers and not actually sold. The lack of real success over imagined was put down to some of the many obviously flaws in the N-GAGE design. One of these was that to change the game cartridge it was necessary to dismantle the phone. Nokia later redesigned the N-GAGE with the QD model, and actually shipped some 3 million units, but it never endangered the Nintendo Gameboy as they'd intended, and public perception was probably to blame.
While I was working on this article I kept thinking about old cars for some reason, and those that invest heavily in maintaining them. If you know someone with a 'classic' vehicle they're often honest enough to accept that the brakes are abysmal, or the wiring atrocious, or the roof leaks in the rain. These are known faults that the owner chooses to overlook, to get the rest of the experience.
So many of the examples here fit into that category, because despite having a truck-load of faults we were willing overlook them for one critical advantage, usually. Or, they were skilfully marketed to avoid buyers focusing too much on the negatives while promoting what positives it had.
Often previous products can influence people to buy new models without actually experiencing the new version, and this can lead to some consternation when the new version isn't actually as good as the one it replaced. In most of the examples I've given here even if these products were initially successful eventually gravity overtook them once people realised they'd bought a lemon.
In some cases, like IE6, this actually damaged the longer term prospects, as people decided to avoid that suppliers subsequent releases. In many cases having a successful product go sour in this way can be terminal for a company, though some seem able to take what money they made and reinvent themselves. The best products in the market aren't always the ones that win, as other factors can alter the public perception such that those with less merit succeed.