The Fire Phone hasn't exactly caught fire. Source: Amazon.com.

We already knew Amazon.com's (AMZN -1.65%) entry into the smartphone market, the Fire Phone, wasn't doing well. Right out of the gate, things didn't look good for Amazon, and after 20 days it sold just an estimated 35,000 units. After two months, Amazon cut the price to $0.99. It doesn't take a genius to tell you that's not a good sign, and it turns out sales were well below Amazon's expectations for the device.

On the company's third-quarter conference call, CFO Tom Szkutak told analysts the company had to write down $170 million in Fire Phone inventory and supplier commitment costs related to the device. After the writedown, Amazon is still left with a whopping $83 million of Fire Phone inventory.

It seems Amazon was just a little over-optimistic about its smartphone, but just how big of a flop is the Fire Phone?

It looks bad on the Surface
It was only about a year ago when Microsoft (MSFT -2.45%) had to write down inventory of its new tablet computer entry -- the Surface RT. After eight months of lackluster sales, Microsoft decided to take the hit and write down $900 million of the initial Surface RT inventory.

By that measure, Amazon's $170 million faux pas doesn't look so bad. Even on a per-month basis -- which seems like the most fair way of evaluating these writedowns -- Amazon's writedown isn't nearly as big as Microsoft's.

Some back-of-the-envelope math pinned Microsoft's additional Surface RT inventory around 6 million units, although it was probably much lower. The same type of calculation implies that Amazon has an extra 875,000 Fire Phones it can't sell. Its $83 million worth of inventory with a bill of materials sitting around $205 implies that Amazon thinks it can liquidate about 400,000 units. Even that might be a stretch, though.

The point, though, is that although Amazon had a colossal flop with the Fire Phone that caused it to miss earnings guidance, Microsoft's flop with the RT was significantly worse.

How did this happen?
This isn't Amazon's first time selling hardware (nor was it Microsoft's first time when it introduced the Surface), so how did it over-estimate demand so badly?

There were a few problems with the Fire Phone right from the launch. First, its exclusivity with AT&T limited its addressable market to about one-third of potential smartphone purchasers. The Fire Phone wasn't compelling enough to make people switch carriers just to get it.

Second, the Fire Phone was priced at the high end of the market, which was a huge detour from Amazon's sales model. The Kindle Fire tablets have seen successful because they're pretty good tablets sold at a relatively low cost.

The high price is in part due to Amazon's decision to work with AT&T. The legacy subsidy model meant that it made no sense for Amazon to sell the phone for less than $450 (off contract). In addition, Amazon didn't want to position its phone at the low end with the free Android phones, because the main goal of the Fire Phone is to attract people to buy things on Amazon. Amazon is looking for people that will use the phone to spend more money, not people looking to save as much as possible.

In a market that's increasingly moving away from the subsidy model, Amazon might have had better success without carrier exclusivity and by offering the phone at a price competitive with Motorola's offerings of $350 to $500 with no contract.

There's still hope
Even after Microsoft's tremendous flop with the Surface RT, the company remained committed to the Surface device, and it has slowly gained momentum. Last quarter, Microsoft sold $908 million worth of Surface tablets, more than double from the same period a year ago. The numbers may be skewed by the early launch of the Surface 3 in June, compared with the Surface 2 launch in October of last year, but still, the growth is promising.

If Amazon can learn from its mistakes and manage its expectations better, it could turn things around for the Fire Phone. It won't be easy after a terrible first release, which averaged just two stars in reviews on Amazon.com, but Microsoft shows that it's possible.