“The Saturn is not our future” Sales, Costs and the Demise of the Sega Saturn

Among Sega fans there are few figures hated as much as Bernie Solar. Appointed as the head of Sega of America after maverick leader Tom Kalinske left the company, Bernie Stolar is widely credited with being behind a number of key mistakes – not least killing off the Sega Saturn in the West.

In an excellent episode of their podcast (and not just because it name checks my good self. Ha.) The Sega Guys make the argument for why it was wrong to abandon the Saturn almost too well: though far from threatening the now-rampant PlayStation, the Saturn had seen an increase in sales through the end of 1996 and 1997. The likes Panzer Dragoon Saga and Shining Force 3 were well placed to tap into Final Fantasy fever had they received the correct marketing nudge. Potential system sellers were cancelled, moved to the Dreamcast or – in the case of third parties – moved to the PlayStationsphere exclusively.

The more I thought about this the more I thought we were collectively missing something. I think the Sega Guys are right to suggest the decision must have been bigger than Bernie Stolar: even if we remove the complicated politics between Sega’s Japanese and American arms, the decision to effectively remove a company from its main marketplace for two years is so self evidently suicidal that there must be something else going on, surely? Logically the conclusion would have to be that, as Sega were a business, they thought that simply releasing games for the Saturn would cost them more than killing the platform and starting fresh. That couldn’t be the case though, surely?

Hardware Production

On the cost of building the Saturn – Next Generation, August 1995

One of the things said about the Saturn is that it was an expensive bit of kit to build, so expensive in fact that Sega ended up selling it significantly below cost. Surprisingly there isn’t a tremendous amount of information available today on how much the individual components cost. I think there’s enough anecdotal evidence to draw on to suggest this was almost certainly is the case. Back in 1995, Next Generation magazine was reporting that it cost Sega $380 to produce the Saturn. This seems to be confirmed by Sega’s head of consumer R&D at the time Hideki Sato. In a 2018 interview he said:

So we released the Saturn in 1994, and as I said before, there were two SH-2s. In addition, memory was expensive at this time, and we were using a large amount, so costs were very high. For each Saturn sold, we lost about 10,000 yen ($100)

https://www.sega-16.com/forum/showthread.php?33506-Hideki-Sato-on-the-Sega-Saturn-(incredible-new-interview)

Though the Saturn originally launched at $400 dollars in the US, it was rapidly reduced to $300 dollars to compete with Sony’s PlayStation, making these sources seemingly in alignment: Sega ended up selling the Saturn for less than it cost them to produce.

This wasn’t necessarily a bad thing of course. Not only could Sega recoup the losses via software sales, but over time hardware tends to become cheaper to produce. Thanks to Moore’s Law you can theorise that, over the space of two years, you should be able to cram the same chip into around half the physical space. Half the space means that the same size ‘wafer’ should be able to produce double the number of chips, reducing the cost overall.

How quickly could Sega cut the cost of producing the Saturn? Thankfully, the community have created a thorough list of the differences between each of the Saturn motherboard revisions, along with a database of the serial numbers you can use to check what motherboard revision your Saturn has (the format is manufacturer code, single digit year, single digit for the motherboard variation then the console’s individual serial number. You can find a huge database of serial numbers here)

From these lists, we can see the earliest cost reductions occurred with the very first revision (VA1) which was first manufactured in 1995. From an electrical design perspective, the original launch Saturn (VA0) was a bit of a dog’s dinner. Not only was the the entire cd block on a completely separate ‘daughter’ circuit board, but the the buttons and the sensor used to check whether the drive was open were all separate units and the console, as a whole, used a lot of off the shelf components.

From first revision (VA1) onwards, things were a lot more streamlined – all of the components listed above were included in a single PCB, with only the controller ports being separate from the rest. It’s difficult to say when they started to be manufactured, but given that a large number of VA0 boards date from 1995 and the machines used for VA1 quality assurance also date from 1995, I think its safe to say that these probably didn’t enter mass production until Sega were a good chunk through the year.

A Quality Assurance unit produced for the VA1 Saturn run

From there, most of the changes from VA2-5 are mostly tinkering. As the remodeled ‘Type-2’ Saturn case design no longer needed the memory reset button and the access LED (and its accompanying resistor) these were naturally removed from the design, but there were also other changes to accommodate different memory and newer (21-pin) CD drives. The initial 21-pin CD drives don’t seem to represent a cost saving (they represent similar build quality and were from the same manufacturer as the originals) but the memory might: Up to VA10, even numbered Saturn revisions used SG RAM (Synchronous Graphics Ram) while odd numbered revisions used SD RAM (Synchronous Dynamic RAM.) As graphics RAM is generally more expensive it’s unclear whether this represents Sega getting a great deal on job lot of SG RAM (thus a cost saving), or whether they had to spend extra money due to an inability to acquire SD RAM in sufficient quantities (which might represent a cost increase.)

Still, we can say with certainty that a degree certainty that larger cost savings appeared from VA6 onwards. From Variation 6 all of the components, including the controller ports, were produced on a single board. There was a further cost saving in the CD block: in earlier models the design had been split between two different integrated circuits but in Variation 6 it was all condensed into one. It also looks like Sega were manufacturing a ew different variations simultaneously to use up supplies of older parts. We can see that one Variation 6 was manufactured in Malaysia in September 1996, for example, while one variation 8 (which featured two ICs on the CD block) was manufactured the month before.

Further cost reductions appear to have occurred in early 1997. The Saturn’s sound unit was made up of multiple chips: a custom sound chip from Yamaha, a dedicated supply of ram and a cost-reduced Motorola 68000 CPU to control them. From variation 10, these were condensed into one integrated circuit from two, which should have produced further cost savings – though this revision seems to have produced some sort of error, as variation 11 included a small daughter board to fix it.

Most likely the single biggest cost saving, however, didn’t occur till the end of the Saturn’s life – round about or just after Bernie Stolar’s fateful E3 appearance. In VA15 – the final Saturn variation – both of the Saturn’s H2 processors were condensed into a single integrated circuit. By this point, the condensing of the CD , Sound and processor blocks into single units – combined with Sanyo CD drives (which definitely seem to be of lower quality compared to the originals) and falling RAM prices – mean that the components for VA15 Saturns should have been significantly cheaper to source than the ones used in the launch models.

One of the final Sega Saturn variants

Did these cost reductions mean that later Saturns weren’t as onerous to manufacture as legend suggests? Unfortunately there are a couple of caveats that suggest this wasn’t the case. The first was the large number of manufacturers Sega had to rely on. It’s undoubtedly true that combining the two expensive simple SH-2s into a single unit would have reduced the manufacturing cost, but an article about the development of the H2 chip from the September 1997 issue of Nikkei Electronics magazine suggests that this was the culmination of three years at work at Hitachi. This begs the question of just how large a proportion of any savings generated would have eventually found their way to Sega:

By March 1997, a total of 7.56 million Saturn units had been manufactured. In other words, approximately 15 million SH-2 chips had been made for the Saturn. With this, the SH had suddenly risen to the second-most-selling high performance RISC microcontroller, and its name became known throughout the world.

However, in terms of sales, the outcome was by no means favorable. Even though the Saturn was equipped with two SH-2s, that did not mean that profits simply doubled. In fact, the profit per unit decreased. Due to being the first sale, they were unable to increase the profits to a satisfactory level.

The only solution was to decrease costs as much as possible. From April 1994, before the start of mass production, work had begun on reducing the size of the SH-2. After that, a special chip was made for the Saturn that combined two SH-2s into one. This effort paid off, and profits finally rose to a healthy level

https://www.sega-16.com/forum/showthread.php?33527-The-Story-of-the-Hitachi-SH-2-and-the-Sega-Saturn

The second important caveat was the cut-throat nature of the 32-bit market. Sega are often criticised for the early launch of the Saturn, but in fairness to them I think their decade of struggle in the industry had taught them the initiative early. The Master System, for example, was a demonstrably superior piece of kit to the Nintendo NES/Famicom, but despite this fact Sega was never able to claw back mind share among the general public outside of Europe and Brazil. Likewise, Nintendo’s SNES was pushing out impressive 3d titles as early as 1993, but it was only really able to approach the lead of Sega’s Megadrive/Genesis in the West after Sega had begun looking towards the next generation of consoles.

Unfortunately for Sega, when it came to the 32-bit market they ended up exactly where they didn’t want to be: fighting an aggressive campaign against a rampant PlayStation without a whole lot of momentum to back them up. Consequently they had no choice to cut the price of the system: By October 1995, Sega was forced to slash prices in both the US and Japan by about $100 in each region to compete with the PlayStation, with more to follow shortly after. By May 1996 – four months before Sega produced the example VA6 we saw earlier – the Saturn was selling for $250 dollars in the US, £250 in the UK (£50 off the original RRP) and Sega had launched a redesigned white Saturn in Japan with a retail price of 20,000 yen – about $185 at the time.

The timing of these price cuts demonstrates that, if Sega did start to generate cost savings from Saturn motherboard variation 6 and onwards, they were playing a desperate game of catchup with the price cuts they had already made. They most likely weren’t eradicating that crucial $100 deficit that had been created by matching the PlayStation’s price in late 1995.

Worse, these price cuts continued in the months that followed. In June 1997, Sega were forced to follow Sony and Nintendo in cutting the price of their machine to $150 in the US – less than double of the official $80 dollar price of the aged Sega Genesis at the time. Had the Saturn officially stayed in the business, they likely would have had to cut prices further: Sony reduced the the price of the PlayStation to $130 in 1998 before breaching the $100 dollar mark the following year. In order for the Saturn to break even, Sega would have had to remove around $280 from the cost of the manufacturing process by June 1997 and then somehow squeeze another $20 dollars the following year. The Saturn seems to be hardware destined to lose money then.

To be fair to Sega once again, that had essentially been part of the strategy from the beginning. To extend the Hideki Sato quote we saw earlier:

For each Saturn sold, we lost about 10,000 yen ($100). That’s how the hardware business works. But the goal was to recoup the losses from software royalties. If there are lots of third parties, lots of games sold, and we get 2,000 yen for each, it’s possible. However, if software sales are weak, and for each console sold, we’re ultimately losing 5,000 – 6,000 yen, what’s going to happen from the business perspective? We’re going to stop selling consoles. This later became a huge problem.

https://www.sega-16.com/forum/showthread.php?33506-Hideki-Sato-on-the-Sega-Saturn-(incredible-new-interview)

Much of the discussion of the Saturn’s failure focuses on the hardware (which definitely reduced Sega’s wiggle room) but when it came to the bottom line software was arguably just as big a problem for Sega.

Software and Sales

As we’ve seen, the Saturn was never going to make Sega money on hardware sales alone. Sega needed to shift a large amount of software in order to balance the books. Where better to start looking, then, then the Saturn’s totem region – its home country of Japan.

In Japan, the Saturn was unquestionably Sega’s biggest success. Historically, Nintendo had the region pretty much sewn up with their Famicom and Super Famicom, completely dominating console sales all the way back to 1983. Games for the Super Famicom would, in fact, still regularly brake into the top 100 well into 1996. The Saturn, however, managed to break this trend. Though eventually outsold by a factor of around 6-1, the Saturn managed to take an early lead over Sony’s behemoth and successfully beat the N64 permanently into third.

And this is generally where the narrative about the Saturn in Japan ends. Our focus in the ‘console wars’ is often too much on the hardware we play the games on and not the actual games themselves. When we look at software rather than hardware, Sega’s success with the Saturn becomes a distinctly more qualified one. The Saturn launched late in 1994 and had just two titles to go at launch. Consequently, it shouldn’t surprise anyone that the iconic Virtua Fighter sold through at a rate of about 1-1.

Now, It’s worth noting that the Famitsu Weekly sales data we’re about to take a look at (courtesy of the utterly fantastic Game Data Library) isn’t 100% accurate, but I hope you’ll agree that the margins are big enough for us to get a reasonably clear picture of how things stood in the land of the rising sun.

Heading into 1995 then, we can clearly see the year was the last hurrah for the previous generation. Super Famicom sales accounted for an amazing 43% of all game sales in 1995 with the Saturn and PlayStation accounting for about 25% each. Given the accuracy issue noted above, it becomes difficult to call for one system or the other given the closeness of the race (Famitsu puts it at less than 200,000 sales) but we can say without much doubt that Sega was by far the largest publisher of the year. At over 6 million unit sales they almost sold more than Namco (2) and Squaresoft (3) combined. Mind you, this victory came at a cost. Sega may have sold more than double the number of units of Namco, but at 49 titles they released more than three times as many games as either of the other two – meaning the number of sales they made per title were far lower. The same holds true for the Saturn itself. Though software sales between the Saturn and PlayStation were roughly equal at this point, Sega’s new console had 11 (13%) more titles.

In 1996, things get more concerning. Though Super Famicom sales began their steep, inevitable decline, the Saturn didn’t really benefit: with around 18% of the software sales market up for grabs, almost all of it seems to have gone to the PlayStation, with the Saturn gaining just 1.3 million extra game sales (around 2% of the market) compared to the PlayStation’s 5.3. (around 12.7% of the market.) In terms of Sega’s own software sales the picture is a little more complicated: Though they sold less titles year-on-year (5.3 million compared to 6.7) and finished behind an N64-powered Nintendo, they released 12 fewer titles than in 1995. Consequently, the number of sales per title actually increased slightly year-on-year – 145,078 in 1996 to compares to 137,154 in 1995.

1997, however, was a year with no saving graces (except, maybe, for N64 demand falling off a cliff.) While Final Fantasy-fueled rocketed the PS1’s share of the market to over 60%, Saturn Sales declined dramatically to just 8,042,930 – down from 11,500,000 the year before. Sega couldn’t take solace from their own titles either: though they released less titles year-on-year for a second year running, their 33 titles shifted just 2,923,294. Finishing behind Final Fantasy publisher Square was probably a given, but they also sold less copies than Nintendo, Bandai and SCE. Sales per title dropped from 145,078 in 1996 to just 88584 in 1997.

Looking at the overall software attach rate in Japan for this period then, the Saturn’s 2.5 million hardware sales by the end of 1996 meant that Sega sold around 4.5 games per Saturn in in 1995. This rose ever so slightly to around 4.6 games/Saturn in 1996 before falling hardware sales allowed the rate to climbing up to around 5.4 games per Saturn in 1998.

Now, none of this means that the Saturn did poorly in Japan by any means. Aside from the early period that saw it capture 30% of the hardware market, over the course of its life it outsold the PC Engine, N64, GameCube, Wii U and the Neo Geo AGS. The individual games were no slouches either – Virtual Fighter 2 shifted 1.7 million copies in Japan, meaning that by the end of the consoles’ Japanese life it was owned by just under one in three Saturn owners.

However, as well as the Saturn performed in Japan, it’s quite clear that it wasn’t doing well enough to carry regions where the console wasn’t putting in such a strong performance. Indeed, Even if we presume that all the software Sega sold in 1995 and 1996 were full price Saturn games, Sega titles only accounted for around 66% of the Saturn games sold in 1995, 47% of Saturn titles sold ’96 and just 36% in 1997. This would mean that, for the 4-5 titles your average Japanese Saturn owner purchased between 95 and 97, there’s a good chance that 2-3 of them would have been third party titles that only have generated around 2000 yen ($20) for Sega in royalty payments (an amount implied by Hideki Sato.) Considering the timing of the price cuts we saw earlier, by 1997 there’s a good chance many of Sega’s Japanese Saturns may have still technically been sold at a loss.

Unfortunately for Sega, It’s only goes down hill from here – even if the Saturn’s performance in the US isn’t as bad as reputation would have you think. The Saturn’s surprise early launch is often presented as one of the reasons for its failure, but it was successful to the point of the company lacking the machines to meet demand. The exact number of sales are hard to ascertain (thanks to Sony and Sega both apparently inflating early numbers for PR purposes), but by mid 1996 it looks like the Saturn had sold 500,000 consoles and 3,000,000 games. These figures didn’t look as impressive next to the 1 million consoles and 8 million games supposedly shifted by Sony, but they show that the console had steady sales with a reasonable attach rate.

For Sega, this chart in Gamepro June 1996 issue was presumably unpleasant reading

By the end of the year Sega had potentially boosted Saturn sales considerably to around 1.6 million consoles, but these sales came at a cost. Not only had the cost of the machine decreased in March 1996, but to boost sales in the pre-Christmas rush they bundled the system with three of their best games: Virtua Fighter 2, Virtua Cop and Daytona USA. The promotion worked to the extent that Sega claim to have sold 500,000 consoles in a month, but Sega had given away three titles they probably needed to sell – particularly as the bundle also contained a $15 money-off voucher for Saturn software. In a January 1997 press release Sega boasted of selling 2.5 million Sega-published and 2.6 million third party titles, but this would have given the Saturn an attach rate of just 3 games per console with only 1.5 being a first-party title. Given Sega and Sony’s habit of boosting sales numbers at the time, you also have to wonder if the 2.5 million first party titles figure included the 1.5 million games Sega had bundled with the console in December.

As time got closer to Stolar’s fateful announcement things got worse. In a Gamespot article that appears to originally date from 1997 (but has the a current publishing date of the year 2000) we see that any hubris Sega may have felt towards their end of year performance was probably short lived. Sega might have been slowly gathering momentum when it came to hardware sales, but they clearly weren’t selling software in the volumes needed to make the system profitable:

The top 25 games sold through to consumers in the first three months of 1997 (as opposed to raw numbers of copies sent to stores) reached 849,000 at #1 (Super Mario Kart 64), fell to 523,000 for #2 (Super Mario 64), and went down to 58,000 by #25 (Super Mario Kart for the SNES). The highest-rated PlayStation game, Tomb Raider, sold 143,000 in the first quarter of 1997 and was ranked #9, behind Donkey Kong Country 3 for the SNES (#8) and ahead of Donkey Kong Country 2 for the SNES (#10). Sega’s Saturn had no games in the top 25, meaning that even the best US Saturn game was purchased by fewer than 58,000 consumers in the first three months of 1997.

https://www.gamespot.com/articles/saturns-distant-orbit/1100-2466421/

Across the pond, the Story was much the same. In Europe the Saturn launches were initially limited but strong – with stocks quickly running out for both the console itself and the RF adaptors needed to run it on a surprisingly large amount of UK TVs. Once again, however, the Saturn was quickly overtaken by the PlayStation – Sony reportedly outspent Sega on marketing by a power of 4 – and Sega were once again forced to slash prices in order to remain competitive. Not only did the console have a rapid price slash to match that of PlayStation, but seven months after launch a special promotion saw it drop further to £250 (just 62% of the launch price) for the duration of April 1996 before dropping to £200 over the summer.

In terms of games, a November 1997 post mortem of the console/interview with Sega’s bold head of marketing featured in Edge reveals that there were 450,000 Sega Saturns sold in the UK by that point (the Megadrive, by contrast sold just 255,000 in its first calendar year.) When it came to software, Gallup charts put the Saturn at around 8 percent of a market – according to the government – worth £546 million. Dividing this £68 million total by the price of a Saturn game reveals sales of around 1.7 million full price games – an attach rate in 1997 of about 3.7 games sold per Saturn – broadly in line with the rate of 4.4 given at the time.

For its first full year on sale, the Megadrive only sold 255,000 units – and it sold for significantly less than the Saturn.

Europe is not just the UK of course, and back in the ’90s the non-existence of the Euro made it an even bigger headache for console manufacturers. The Saturn had prices in German Deutsche Marks, French Francs, Spanish Pesetas and Italian Lira. No matter where you look, however, the story is broadly the same as the UK and the US: Debuting at 79,900 Pts in Spain in July 1995, the console was priced was slashed to 39,900 less than a year later. In Germany the launch price was 699 -DM before being launch to 499 by April 1996. In France, the console sold initially for 3000F before being reduced to 1,490F after E3 1996. In Europe Sega also mimicked the aggressive bundling found elsewhere: In early 1997 Sega were selling a bundle that contained an extra controller, World Wide Soccer and Sega Rally for 1590F/449,95DM.

There were other markets outside of Japan, the US and Europe that were a mixed bag of course – South Korean sales were poor enough to force local distributer Samsung to retreat from the console business altogether while Brazil managed to shift all of its (err, 700) launch consoles despite the console costing the same as 15 games – but without strong performances in the US and Europe I don’t think these markets would have been large enough to really carry the console anyway.

In an age before the internet was ubiquitous, the Brazilian press was unaware of the surprise US launch at the time of print. They reported that Brazil was the first territory to receive the console outside of Japan

Could the Saturn Have Been Sega’s Future?

So then, all of this has building towards the big question: could the Saturn have still been Sega’s future? A casual look at Sega’s finances suggests this to be the case. If we look at the figures from their 1998 annual report, Sega were still able to generate a small profit in the sessions ending in March 1995, 1996 and 1997. It was only when the ditched the Saturn and spent a huge amount on developing the Dreamcast that they generated their first (dramatic) loss.

The six year summary contained in Sega’s 1998 annual report.

However, if we look beneath the surface all wasn’t well at Sega HQ. As we can see from a fantastic new article from MD Shock, the clouds were brewing for Sega as early as 1993. As a Japanese company that made most of its money from exporting to the US and Europe, Sega was suddenly hit by the rapidly rising value of the yen. If you were an American holidaying in Japan in 1992, one dollar would buy you 127 yen. If you returned in 1994 your same dollar would have bought just 102 yen. Any transaction Sega needed to do in USD (like, say, arranging shipping of tens of thousands of games cartridges to the US) suddenly got a lot more expensive, while every dollar earned in the us was worth less when converted back into Japanese yen. On top of that, the Megadrive market was maturing: We can see from Sega’s finances that money from game royalties rose from 1,195,000 yen in 1993 to 9,233,000 in 1995. This looks on the surface to be a dramatic increase in revenue but in reality it was actually a double edged sword: The rise in these royalties meant that games were buying more games from third parties and less directly from Sega. If Sega made twice as much money on their own titles as they did on third parties, this 8 million yen growth could actually be interpreted as an 8 million yen loss. When it came to developing for the Saturn Sega were famously less supportive of third parties than Sony were with the PlayStation. This may go some way to explaining why.

Still, leaving aside royalties for a moment, our main focus here should be on financial performance in the years from 1995 to 1998, and here we see things definitely aren’t all that rosy. From their high point in 1994, Sega’s revenues from consumer products were in a constant state of decline while their costs seem to have increased at a similar rate. The high costs Sega endured as they geared up to release the Saturn are understandable (an eye-watering 270,305,000 yen), but once the console was released they didn’t really decrease. Though 1995 costs dropped to a slightly less eye-watering 257,371,000, in 1996 Sega’s aggressive attempts to buy their way into Western markets saw costs rise beyond 1994’s water mark to 270,968,000. Sega might have been able to post a small profit in 1996 and 1997, but realistically it was growth in their sales of arcade machines that allowed this to happen. If Sega’s income from arcade machines had stayed flat between 1995-7, Sega would have made an 18,272,000 loss in 1996 and a 31,269,000 loss in 1997. Sega’s eventual 43,300,000 loss in 1998 is often attributed to the cost of developing the Dreamcast and dumping their unwanted Saturn stock, but I think it’s probably more representative of the condition of Sega’s post-1994 consumer product finances than it would appear on the surface.

Could Sega have put off the Dreamcast for an extra year and pushed harder at the Saturn? Given that the trend for Sega’s consumer sales was to decrease while the trend for their costs was to rise, the outlook doesn’t look good. Though there’s an outside chance Sega could have bucked the trend between and made n unlikely come back, attempting to do so would be risky. Sega would have needed to sell a large number of consoles and a much larger number of games and they would also have needed to do so without making further dramatic cuts to the prices of software or hardware. Given the pressure being exerted by Sony and Nintendo, this condition feels particularly unlikely.

Had a Saturn-backing Sega not bucked the trend of the previous years – with revenue falling, costs rising and royalties payments all keeping to the trends set by the two previous years – I would argue that Sega would still have experienced a loss making year in 1998. It would probably have been a much, much smaller one (in the region of, say, 6,344,000 as opposed to Sega’s actual March 1998 loss of 43,300,000), but Sega would still have been in the red and they wouldn’t have had a new console launch to show for it. Had Sega put the Dreamcast off for another year and pushed harder at the Saturn, I don’t think it’s unreasonable to suggest that we might have never seen the Dreamcast at all.

Perhaps a better strategy, then, would have been to keep schtum. Rather than openly admit that the Saturn wasn’t their future, could Sega have openly touted the virtues of the console in public while working on the Dreamcast and slowly winding things down in the background? This would have the benefit of giving the likes of Burning Rangers, Panzer Dragoon Saga and Shining Force 3 the display of support and the marketing push they so desperately needed, but would have been incredibly difficult to pull off. Aside from the legal issues surrounding its duty to report to its shareholders, Sega would have needed to either keep manufacturing Saturn consoles it knew it would never sell, or somehow close down/drastically reduce its entire manufacturing process without word leaking out to the press (which, given that in May 1997 EGM reported on Sega slashing their orders for Saturn consoles, seems unlikely.) It would also have needed to keep the details of the Dreamcast a secret until the last possible minute – something that could have potentially alienated retailers, third parties and consumers alike.

This article from EGM issue 94 shows how difficult it would have been for Sega to try and quietly wrap up the Saturn’s production run

As messy as the death of the Saturn was (especially coming only a year and a bit after the official death of the 32x…) it might actually represent the best way out of the situation for Sega. The common narrative suggests that Sega jumped rather than were pushed – new Sega of America boss Bernie Stolar supposedly hated the Saturn and wanted to replace it as quickly as possible – but given the history of Sega’s decision making it seems unlikely that a move as bold and potentially damaging as leaving the market for two years could have been made without Sega of Japan’s buy-in.

I think it’s much fairer to conclude that Sega were pushed to dropping the Saturn by financial circumstance rather than choosing to do so as part of a wider strategy. We’ve seen that the financial situation wasn’t great – with market share in the West being bought at extreme cost and Sega’s immense Japanese momentum beginning to wane – but there might have been other factors at work as well. Despite Sega of Europe’s head of marketing remaining bullish heading into Christmas 1997, anyone walking into a shop in the UK would have seen a different story, as explained in Edge’s Saturn post mortem:

The arrival of the N64 in the marketplace has further complicated matters for Sega; retailers recognising this gap, have cut Saturn shelf space. Thus begins a self-fulfilling prophesy; with walls of PlayStation software and a growing N64 library, Sega’s machine is removed from the public eye.

Many large software chains have reduced store space for the Saturn, and independents also reported little demand. Virgin Our Price dropped the format towards the end of 1996, and in a number of Electronics Boutique stores the Saturn rivals only the Amiga in terms of shelf space.

Edge 51 (November 1997)

One obvious thing Sega needed in order to be successful was to have the console available on shop shelves. In 1997 Tesco sold groceries and Amazon sold books. To shift large numbers of hardware and software Sega needed to be taken seriously by video game retailers. If Sega couldn’t make it worthwhile for retailers to stock the console, the console would, to all meaningful intents and purposes, be dead. For the sake of prestige at least, it would be better if the decision to kill the Saturn came from Sega itself rather than the buyers at Toys’R’Us, GAME or Electronics Boutique. Given the widespread rumours’ of unsportsmanlike behaviour from Sony when it came to retail (it’s difficult to find a Western Saturn owner who wasn’t nudged towards the PlayStation by retail staff…) the battle for 32-bit retail space is probably a subject that’s best left to its own article. Still, for our purposes, I think it’s important for us to acknowledge that there were more players on the field than just the console manufacturers and consumers.

Overall it seems that, going into the middle of 1997, the Saturn was in a frustratingly tragic position. Though its Japanese momentum had been stopped by Final Fantasy VII, there were still millions of Japanese Saturn owners waiting to purchase Saturn games. Performance in the West may not have been anywhere near as strong – and achieved at significant cost to the company – but history has demonstrated the position was potentially salvageable: today no one would consider the PS3 a failure, but in its second calendar year on the market the PlayStation 3 was not only being outsold by not only the Wii and the Xbox 360, but it was also selling less than its aged seven year old predecessor by a good 1.4 million units.

What Sony had in 2007, however, was money. We tend to think of Sega vs Sony as a David and Goliath affair. To be honest we should probably substitute Goliath for Godzilla or Unicron: At its 1994 high point, Sega could boast of posting 235,880,000 yen worth of sales of consumer products. In 1998 Sony reported its console sales to consumers were worth 700,000,000,000 yen. What was even worse for Sega was that games weren’t even Sony’s most profitable business: combined, sales of Sony’s audio and visual products were worth a staggering 1,999,000,000,000 yen – and that’s before we get into Sony’s cameras or their cinematic ventures. Not surprisingly, while Sega’s 43,300,000 yen loss in 1998 caused the company significant hardship, Sony was able to take the $3.3 billion losses generated in the early years of the PlayStation 3 more or less in its stride.

Consequently, though the likes of Burning Rangers, Shining Force 3, Panzer Dragoon Saga and the exciting future prospects of Shenmue and a cartridge-enhanced Virtua Fighter 3 could have ensured the Saturn went out with a bang, Sega’s financial situation prevented this from being the case. Many reasons are given for the Saturn failing – Sega dividing attention in the West with the 32x, leaving RF Modulators out of UK Saturn bundles, alienating key US retailers with the console’s surprise launch – but for me the main one is the economics of the console itself. On paper, we would regard Sega’s first console, the SG-1000, as a failure simply because it was outsold many times over by Nintendo’s Famicom. At the time, however, Sega made a good chunk of change from it:

We released our very first video game console, the SG-1000, and it sold 160,000 units. Those were huge numbers, considering Sega has only made arcade games that sold no more than several thousand units up until then.

https://www.siliconera.com/former-sega-president-talks-about-making-the-mega-drive-to-beat-nintendo/

By 1997, the Sega Saturn was definitely developing its niche. It was abundantly clear it would never achieve the mass popularity of PlayStation, but in the UK at least it was rapidly becoming the enthusiast’s choice:

Edge has noticed a surprisingly high quantity of dual PlayStation/Saturn ownership – that people buy a PlayStation for the Psygnosis/Namco/thirdparty games and a Saturn for the sega titles. Many, in fact, express the Saturn as their preferred format. How does Sega regard this? Would you say this confirms the Saturn’s position as a machine for the videogaming enthusiast?

Jo Bladen: Our internal research and database information shows this trend is increasing. There are many people now that own both machines, mainly as the low price point mean that if you want to play exclusive Sega games then £99.99 is a low barrier to entry.

Edge 51 (November 1997)

Because of the underlying economics, however, this wasn’t good enough for Sega. It’s obvious from the figures we’ve seen above that the PlayStation would have still been worth Sony’s effort even if it hadn’t sold as well. The Saturn, on the other hand, absolutely had to be a smash hit. Rather than simply hoping to repeat the Megadrive’s incredible and unexpected success, Sega put themselves in a position where the continued existence of their consumer products division more or less depended on it. By 1997 the Saturn had carved itself a reputation and a niche that would most definitely have allowed it to be a more solvent and financially viable company’s future. Unfortunately for both the console and the army of dedicated fans that continue to sing its praises to this day, that somebody couldn’t be Sega.

Enjoyed this Article?

If you enjoyed this article you can let me know on Twitter, Facebook and Instagram!

You can also find my music on Tidal, Spotify, Youtube, Soundcloud, Deezer, Tiktok,