Here are how the three main Japanese carriers fared during the month of September:

NTT DOCOMO lost 66,800 subscribers, au/KDDI gained 232,700 and SoftBank 270,700.

DOCOMO loses market share again, SoftBank is the top carrier in terms of subscribers gain for 21 months in a row1 and au/KDDI continues the nice streak started a year ago, fending off SoftBank’s attempt to becoming #2.

Let’s stress the fact that DOCOMO is still by far and large the biggest carrier in the country: it has 22 million more subscribers than its closest competitor.

Right after the release of those numbers last week, dozens of articles—often parroting each other—went on about how the iPhone was a failure at DOCOMO. I refrained from writing immediately, waiting for the noise to dwindle down. Because it’s more complicated than that.

What’s in a full month

First and foremost, the numbers do cover the entire month of September. You know, a full month, like 30 days. The iPhone 5s/5c were released on September 20. Like a third of that full month.

To make it clear: the iPhone launch cannot be entirely responsible for the fate of any Japanese carrier this September .

This is barely the fault of analysts getting too excited, as DOCOMO itself has blamed limited iPhone stocks for the drop in its subscriber base.

A simple look at the numbers tells you that this is not the case. DOCOMO had 27.5% share of iPhone 5s market in its first week of sales—the last week of the month that we’re interested in. au/KDDI had 30%2.

So, with similar sales share, au/KDDI won more than 200k customers thanks to the iPhone launch and DOCOMO lost more than 50k because of the same exact launch? And stocks would have been much more limited on DOCOMO than its competitors?

Let’s be serious.

In a nutshell, why the subscriber loss? The iPhone 5 and DOCOMO itself .

The iPhone 5

If you look at the sales numbers for the launch week-end and the week immediately following it, it’s not difficult to witness that one of the factors in the iPhone domination is the very strong sale of the previous model, the iPhone 5.

This isn’t new and has been clear at every iPhone launch: there’s always a massive uplift by the sale of the previous iPhone model in Japan .

au/KDDI and SoftBank are heavily discounting the iPhone 5, making it basically “free”. It was not only expected by customers, but also advertised before the release of the new model. This is a weapon that DOCOMO doesn’t have: no discounted iPhone 5, no fall-back stock and offers.

DOCOMO lost more than 130,000 customers who used the mobile number portability scheme3, which tends to support that they got lured by very attractive offers for that smartphone4

This is why I had suggested that DOCOMO could have been better at marketing the 5c pushing it more heavily than its counterparts. Truth to be told, it could have meant a cutthroat price war—the three operators react to each other’s pricing like symphonies—that would have not created any real winners in the end (beside the customer), though I’m not certain that SoftBank or au/KDDI would have wanted a situation where there was basically no differentiation between the 5 and the 5c.

The other important factor, an unknown one, is the terms signed with Apple. There might be some limitations in the freedom the carriers can have, restricting the ability to actually even think about a full-blown price war.

DOCOMO’s other problem

It remains that reading the loss of subscribers only with the iPhone in mind is both biased and naive. I’ve said it repeatedly, here and elsewhere, one cannot read what the iPhone will do to DOCOMO at this stage. You have frontloaded sales for early adopters, likely to already have switched to au/KDDI and especially SoftBank in the past years and just trading their phone (with contract). You have price-sensitive customers appealed by the heavily previous model that DOCOMO doesn’t carry. You have limited stock for everyone on the flagship models.

DOCOMO’s problems have started way before this. The carrier somewhat suffers from an image problem and from a complex price structure problem. iPhone or no iPhone, it has repeatedly shown an inability to counter the two challengers, and this for more than 5 years now5.

If you look at the market share of the the carrier in 2006 and compared it with the graphic at the top of this post, you’ll immediately understand:

The loss of subscribers during the month of September has been heavy. Almost a record, yes. But, again, let’s not pretend this is new, the carrier has had a bad year, losing customers in four of the last twelve months.

It’s DOCOMO’s strategy that it is at fault here. While it keeps innovating in infrastructure technologies, it still behaves like it were in a world where it’s the market-definer, the untouchable leader that sets the standards for the others6. It’s a mindset conundrum.

Accepting the iPhone on its network was the first step7 and early signs in October show that DOCOMO is faring very good, taking the top spots in mobile phone sales in Japan with the 5s, confirming what I thought: the iPhone was needed.

It’s the first time we have a standardized device on all three carriers in Japan. SoftBank and au/KDDI cannot differentiate between “we have it and you don’t” anymore. The iPhone 5c will take off8 in a second cycle of—more conservative—customers. The 5s is a flagship phone that flies off the shelves.

If DOCOMO ends up not making any in-roads with a device that is so adored in Japan, it will have only itself to blame.




  1. 41 months out of 43! 

  2. The 5c breakdown was similar—while only counting for a fifth of the sales. 

  3. the customer gets to keep its number, carriers often offer special deals for those 

  4. The differential, or about 65,000, are new customers. It’s impossible to know how many of those got on-board for the iPhone 5s/c, the full month of October will show us a better overall picture of the dynamic between the three carriers 

  5. and contrary to some reports, DOCOMO has seen a worse loss of subscribers: in December 2006 

  6. the “benevolent dictator” theory, that brought such innovation as NFC or mobile internet earlier in Japan than anywhere else 

  7. though I’m sure it also got a better deal by signing now that Apple needs growth rather than 5 years ago. 

  8. I don’t read too much in the recent reports of supply chain production cuts by Apple 

IT Media:

On October 1, Evernote and NTT DOCOMO launched a campaign offering a one year premium subscription to the cloud service

This is not a new deal per se (it actually dates back to 2010, the same year the company opened its Tokyo bureau), as the two companies have had a similar agreement for smartphones and tablets1. The DOCOMO x Evernote deal is now expanded to the iPhone .

Evernote is extremely successful in Japan. I have been lucky to visit their offices several times and the numbers of books written about them in Japan is just staggering2.

The newly-introduced Evernote Market heavily features Japan, just look at the video (and for those who know her, let’s see how many times you can spot Mika, the marketing director of Evernote Japan).

  1. it is believed that DOCOMO carries the iPad, but no official news is expected before the middle of the month 

  2. more than 40 when I last visited 

Following up on the massive win for the iPhone during the launch weekend, let’s look at how the iPhone behaved now that we have a full week of sales.

Weekly phone sales ranking

Here are BCN‘s rankings, from Sept 23 to 29:

  1. iPhone 5 16GB (au/KDDI) — was #3 (+2)
  2. iPhone 5s 64GB (SoftBank) — was #2 (0)
  3. iPhone 5s 32GB (au/KDDI) — was #4 (+1)
  4. iPhone 5 16GB (SoftBank) — was #5 (+1)
  5. iPhone 5s 32GB (SoftBank) — was #1 (-4)
  6. iPhone 5s 64GB (au/KDDI) — was #9 (+3)
  7. iPhone 5s 64GB (DOCOMO) — was #6 (-1)
  8. iPhone 5s 32GB (DOCOMO) — was #7 (-1)
  9. Samsung Galaxy S4 — was #10 (+1)
  10. iPhone 5s 16GB (SoftBank) — was #8 (-2)
  11. Kyocera Gratina — was #11 (0)
  12. Sony Xperia A — was #12 (0)
  13. iPhone 5c 16GB (SoftBank) — was #15 (+2)
  14. Panasonic P-01 E — was #18 (+4)
  15. iPhone 5c 16GB (au/KDDI) — was #19 (+4)
  16. iPhone 5 32GB (SoftBank) — was #16 (0)
  17. iPhone 5c 32GB (DOCOMO) — was #17 (0)
  18. iPhone 5c 32GB (SoftBank) — was #25 (+7)
  19. iPhone 5 32GB (au/KDDI) — was #20 (+1)
  20. Sharp Aquos Phone Serie SHL22 — was #24 (+4)

Down the list, the iPhone 5c 32GB by au/KDDI is up 4 places to #24. The iPhone 5s 16GB by DOCOMO is down 13 spots to #26, the exact same fall for the same phone on au/KDDI at #26. The last appearance is made by the iPhone 5c 16GB on DOCOMO at #33 down three spots.

Still, another massive week for the iPhone 5s in Japan .

Remember that those ranking are derived from tracking points of sales—covering 40% of the OTC electronic stores—, notably not including Apple Stores, nor carrier shops. It has thus to be read as an indicator only, one that has proven solid historically1.

5s v. 5c

The situation remains almost unchanged from the launch weekend if we look at how the two models stacks against each other. The 5S still outsells the 5c by a wide margin . Unsurprising at this stage.

iPhone 5 v. 5c

Why does the iPhone 5 rank so high? au/KDDI and SoftBank are massively subsidizing the 16GB model rendering it “free”2. In other words, they’re dumping the stock. Back when the 5 was released, there was a very similar behavior with the 4s.

It’s almost ironic that au/KDDI would push this offer so strongly, since they’re so adamant on marketing their 800MHz LTE band …a band that the iPhone 5 lacks3.

Note that the networks are offering great discounts on the 5c, they just seem to first be wanting to clear the 5 first. DOCOMO seems on a holding pattern of wait and see, whereas they should probably use the angle of marketing a “free” 5c 16GB versus the old 5 its competition pushes.

Limited 5s supplies

Limited 5s supplies are being reported in shops around the country. This is the main reason behind the big drops in the rankings compared to last weekend, although one cannot yet read the short-term success of the new models (in other words: how frontloaded they were in the first three days).

This is also where the methodology of BCN—for all solid that it is as mentioned above—finds its limitation, as Apple Stores and flagship carrier shops are getting priority for replenishment.

DOCOMO’s share

I have the feeling that DOCOMO is not making the best job out of the iPhone . Comparing with last week, the carrier actually lost iPhone market share to its competitors throughout the week. It swapped its second place with au/KDDI4.

The lag in getting the carrier email to work on the device (not mentioning it’s not through a push method) but also the perceived lack of traditional DOCOMO offerings of back-ups, battery support and other services (through the Premier Club for instance) might partially explain this. But, again, I sense its marketing being at fault, probably because of a last-minute deal with Apple leaving no time to truly build any strategy.

SoftBank, clearly, continues to dominate the iPhone market . It’s way too early to seriously analyze the reasons. But it should win the subscribers gains batttle when the numbers are released later this month.

Breakdown by capacity

BCN broke down the sales by capacity, offering some additional insights.

As expected with an early adopter crowd, the more expensive models are the most popular.

If you look at the same numbers by carrier, you can see that SoftBank and DOCOMO share a similar pattern, the latter having more emphasis on the high-end model and the former a more even distribution overall. The 32GB version looks very successful at au/KDDI.

I wouldn’t read too much into this, as it could just be borne out of inventory issues. It’s only in the mid-term that we will be able to learn if carriers strategies are really into play.

The iPhone 5c is almost at 50/50 between the two available storage capacity: 49.1% for the 32GB and 50.9% for the 16GB.
The notable outlier is DOCOMO: the aggressive pricing of the 32GB compared with the two other networks could explain this differentiation.

In the end, the first full week of iPhone sales in Japan has cemented SoftBank as the market leader for that device, with DOCOMO and au/KDDI battling for second place.

At this time, it’s impossible to know how many of the overall sales came from existing carrier subscribers (DOCOMO and SoftBank being the most competitive), carrier switchers with number portability (au/KDDI and SoftBank being cheaper) or simply new contracts (DOCOMO seemingly offering a better deal).

It appears that DOCOMO is either not doing the best job in communicating its good prices, or suffering from an image problem, at least with early adopters. The second wave of subscribers, until next March, will give some clarity this murky question.

  1. another research company, GfK, reports rankings that are almost similar to those ones 

  2. for new contracts 

  3. that iPhone 5 version is only compatible with the 2100MHz band 1 LTE that au/KDDI offers, the 5s/5c also access the building-penetrating 800MHz band 18; note that if you’d want to also use the 1500MHz band 11 LTE, you should think about a Xperia UL for instance 

  4. note that the numbers are very similar even if you break them down by 5c v. 5s