Categories

Measuring iPhone demand

When Apple announced five million iPhone 5 were sold to end users over its launch weekend, I was surprised. Not because my guess had been around 6 million, but because the company had set expectations by announcing a doubling of pre-orders from the year-ago 4S launch.

Instead of doubling its performance for the launch weekend the company only sold 25% more units. How can there be this discrepancy? Is this a sign that demand is not growing at the rate we’ve become accustomed to? Is it a sign that there are shortages of components or labor or other production problems?

No, probably none of the above.

What we saw in the 5 million figure is what the company was able to deliver in the hands of buyers. It’s possible that there were people who did not get a phone when they wanted it, and at the same time it’s possible that some phones were available for sale and did not get bought.

This is because Apple offers the product through multiple channels. Some channels like Apple Stores may have gotten too many units while other channels like their on-line store, operator stores or retail partners did not get enough.

In other words, we have a situation of over- and under-supply (or over- and under-demand) simultaneously because the product is misallocated.

Can this really happen?

The fact that pre-orders were twice as fast indicates that online purchasing was much more popular. The company’s press release suggested that only “a majority” of online orders shipped so far and we know that day one delivery of units was only offered to online buyers for 40 minutes.[1] At the same time some stores seemed to have plenty of stock.

This is easily explained. It’s possible that this year a larger group of consumers decided that braving long lines is not their cup of tea. Instead, this larger proportion could have chosen to bravely endure waiting for a package at home rather than the companionship of their fellow man. Maybe the population of buyers is broader this year as one would expect from such a large base, and there are fewer who have the time or inclination to queue for a device.

If the shift to online from in-store was unforeseen, the company could easily have allocated too many units to retail and not enough to online resulting in weeks-long waits for orders placed online and relatively short waits in queues.

It’s easy to fault the company for making this error. However, we can only calibrate our expectations on what we’ve seen before. If the company measured demand in its stores and on-line for three years and saw a pattern, it would make sense to allocate the limited supply according to the same pattern of demand[2].

But overall, the performance of the product is relatively solid. The iPhone 4S launched into seven countries in 2011 with a total combined population of 704 million and sold 4 million units in three days. The iPhone 5 launched into nine countries in 2012 with a total combined population of 721 million and sold 5 million in the same amount of time.

That means that 0.57% of the addressable market bought an iPhone 4S in three days and 0.69% bought an iPhone 5.  A handier measure might be that 7 out of 1000 inhabitants in the launch countries bought an iPhone 5 in the first three days.

The pattern for the last three phones is shown below:

The growth from 4S to 5 is subdued, but it fails to account for the in-transit inventory.  Assuming that due to the misallocation of stock, unfilled online orders number one million then the growth would look as follows:

For the markets iPhone launched in, this would be a fairly healthy take-up rate. Anecdotal evidence I’ve seen suggests that 1% penetration on day one is not uncommon.

Notes:

  1. The fact that more than those initial promises were met means that the company quickly re-allocated stock and production to online after they sensed there was a problem.
  2. Demand estimates are also driving the allocation of production capacity and capital expenditures for process equipment and pre-purchase agreements for components. All indications are that these allocations are running at the usual growth rate.
  • disqus_ynCwxUzUCg

    “The growth from 4S to 5 is subdued, but it fails to account for the in-transit inventory. “. Is the assumption that we knew the 4S units, or 4 for that matter, in transit? I appreciate your thoughtful analysis of a wildly interesting company. I also appreciate your dry humor in “…rather than their companionship with their fellow man”.

    • http://www.asymco.com Horace Dediu

      The quantity in question is the number of units which were pre-ordered that did not get delivered during the weekend. The 4S took about a day to reach “1-2 weeks” status. Therefore the units that were declared “pre-orders” (1 million) probably were delivered on day one and were included in the “opening weekend” sales. The 5 took one hour to reach “1-2 weeks” so the first 24 hours of preorders (2 million total) did not all get delivered. In fact, we can only guarantee that the first 40 minutes of orders got in before the “opening weekend” deadline. So my point is that if we assume all the preorders had been fulfilled by opening weekend, the situation would not look so bad. Technically Apple cannot count pre-orders that did not get delivered (and signed for) during opening weekend. You have a situation where people “bought” and “paid for” product a week in advance but it does not get counted as opening weekend sales.

      • cellojoe

        It seems intuitively true That when the market for technology product is expanding you saturate Tech zealots long before you stop creating customers. Also, as you pointed out, Frenzied lines discourage more placid souls. We might be approaching an upper bound for iPhone frenzy but not an upper bound for iPhone sales at all.

      • cellojoe

        And the frenzy leveling off or even dying down might be good for the company’s image.

      • http://www.isophist.com/ Emilio Orione

        I don’t think so, statistically the frenzies, innovators, are almost 2.5% of total users. The more the buyers the more the frenzies.
        As for movies the first week end income is strongly correlated with total income, the first three day of sales should be strongly correlated with the amount of desire for the product and so with the total number of sales.
        But if the product if supply constrained you can’t really estimate the total demand, never mind what analysts say.

      • cellojoe

        Thanks great information regarding the percentage of innovators. Where does the data come from? I’m just curious. I was wondering if we could divide innovators into subcategories. Innovators were willing to stand in line etc. and innovators would rather stay comfortable at home and simply order via a few clicks.

      • http://www.isophist.com/ Emilio Orione

        I posted a link in a preceding comment of this thread.
        It is taken from the statistic of technology adoption. Here is another link http://en.wikipedia.org/wiki/Early_adopter

      • https://twitter.com/#!/azulum azulum

        This is simply a bell curve (which if plotted as cumulative adoption becomes an S-curve). Be very skeptical of the precision that the bell curve implies, particularly for individual products, less so for aggregate technologies.

        Consider that though half of the United States has smartphones, the population of the US is about 5% of the world population — so should we say every American with a smartphone is an innovator, or at the very least an early adopter? I’d say no. There are so many other factors to take into account: income, opportunity, utility, affordability, time scale and so on. The adoption curve is a guideline to help make educated guesses, not a rule to which all things apply.

        If you need an example of an area where the bell-curve-based S curve breaks down, look no further than traffic on the internet. Sure there is a physical upper bound (currently bandwidth), but that upper bound is so large and getting larger that it acts as though there is no limit.

      • Luis Alejandro Masanti

        I’m with you, Horace. I think the problem emerges from Apple being honest/obliged by the law to “only account for signed sales.”
        Samsung and the other cellphone makers only spoke of “delivered” phones to the channel.
        In that last sense, “first weekend in sales” for Apple means almost the same that “devices delivered,” because demand is greater that devices delivered.
        In next or second shareholders meeting we will know the truth on sales… only from Apple. Samsung, Amazon et als. will still be speaking of “deliver to the channel.”

      • franksspam

        And actually we didn’t pay for it when we preordered the phones. I preordered on the 14th and AT&T did not charge my card until the phones shipped this week.

      • http://www.asymco.com Horace Dediu

        I know, which is why I used quotes.

      • franksspam

        Sorry, I was in a meeting last week where someone was using air quotes incessantly even though there was no need based on what he was saying. I guess I’m still shell shocked ;-)

      • Guest

        My experience with Apple products is my credit card is not charged until the item is shipped. Isn’t this also true for the iPhone?

      • http://www.asymco.com Horace Dediu

        That’s correct.

  • http://www.isophist.com/ Emilio Orione

    I know that manipulating data too much yields to unrealistic results, but all the news about iphones 5 missing demand are so light that I feel I can try a little speculation.
    I dare say that the first buyers of the iphone 5 were “innovators”, the name that is assigned to the first 2.5% of a technology product consumers. The next 13.5% are early adopters, then there is early majority, last majority and skeptics http://lrrpublic.cli.det.nsw.edu.au/lrrSecure/Sites/Web/13289/ezine/year_2008/may/article_innovation2.htm.
    So if 5 million are the 2.5% of total users, the projected sell of iPhones 5 could be as much as 200 millions in one year.
    If we use a lager time span for innovators the total amount increase obviously so 200 is the lower limit.
    Quite an accomplishment.
    Is this more or less analysts’ prediction?

    • Ted_T

      I don’t think you can talk about “inovators” when you are talking about version 5 of the world’s most popular product. The first X% buying the original 2007 iPhone were no doubt “inovators”. I don’t think anyone buying the iPhone 5 qualifies, and thus you simply can’t make a valid calculation of overal adoption based on the first day.

      What will make the real difference is if the iPhone goes on sale via China Mobile, if Apple makes a genuine push in current Blackberry strongholds like India and Indonesia. Will Apple keep or resurrect the 3GS for developing world prepaid markets? None of that can be deduced by what happens on day 1.

      • http://www.isophist.com/ Emilio Orione

        I am sorry for the spelling. Innovators is just a name given to the first buyers of a product, they are the people who cherish the product, who really appreciate it and were waiting for it, whatever iteration the product is.
        It does not matter if it is not the first iphone, is a new iphone and “innovators” are people really caring for its new features.

        For movies the first week end income determines the total income, you can think of the first days iphone sales as a proxy for the desire of the new product. The more the desire the more the total sales. There is correlation, maybe not so strong but there is.

      • Ted_T

        Look at the iPad 2 — it had very poor initial sales — did that in any way predict future sales? It works with wide release movies, because there is no supply constraint and because movies often earn a huge percentage of their total gross in the first weekend, so there is nothing to predict — you are looking at real results.

        On the other hand even with movies, the true blockbusters, the one with legs, can’t be deduced from the opening weekend — take Star Wars or Avatar as examples — movies that went on to become the biggest grossing movie of all time during their respective first runs. There is no way to predict that, even with 20/20 hindsight, by looking at their initial grosses.

        I posit the same for the iPhone 5 — again, what happened last weekend is simply not a predictor for what happens in China, etc.

      • http://policydiary.com/ John S. Wilson

        iPad 2 had poor initial sales? News to me. What are you basing that on?

      • obarthelemy

        1- Innovators may be appropriate for early buyers of a new product. iP5 is just a light update to a very mature design. Pre-orderers show at best brand loyalty, not a search of innovation. Plus that 2.5% ratio varies a lot by product segment.

        2- The logistics of films and phones are very different: films don’t have to be manufactured for each viewer, shipped, stored…. So is the sociology: going to the movies is a social, group activity. Not buying a phone. Draxing parallels between films and phone seems like a big overreach.

    • franksspam

      Except that we already have the information for the 4S. If we apply the 2.5% to the 4 million sold the opening weekend we would get 160 million, which is not what they sold.

      • http://www.facebook.com/joseph.carducci1 Joseph Carducci

        The 4S is still on sale is it not?

      • franksspam

        Note the portion I’m quoting from the post I responded to, “could be as much as 200 millions in one year”

        He clearly said in one year. The quarter we are in right now is the end of that year and unless they double their largest quarter of the year they will not come close to the 160 million.

  • Scott

    Interesting article about the impact of the new screen on supply constraints.

    http://www.bloomberg.com/news/2012-09-25/apple-use-of-thin-display-seen-driving-iphone-5-supply-shortfall.html

  • RobDK

    Good analysis, Horace.

    I wonder whether anyone knows whether there is a logistical limit to the number of devices a company can distribute to 9 different markets within a short period of time, like the 2-3 days Apple had here? In other words is there a limit to how many products UPS, FedEx, etc. can deliver, both to private individuals and to retail outlets, in a short period of time?

    Maybe it is logistically impossible to deliver 10 million iPhones around the world within 2-3 days?!

    • Scott

      @RobDK, that’s what I’ve been thinking too. Once you hit that limit it just won’t be possible to increase sales. I think the amount would probably vary per country according to size, distribution network infrastructure etc. I wonder if we are reaching this in some countries now.

      It’s interesting the shift towards online purchase as opposed to queuing – it’s certainly something that I’ve changed to. Partially because I don’t have the luxury of waiting 4 to 6 hours for a new device.

      I would think that this is an area where the limits would be approaching soon (unless they move towards more delivery companies). You get the feel that the current delivery organisations are stretched to the limit.

      • Michael W. Wellman

        From

        “On a day expected to be the busiest in FedEx’s 40-year history, […snip…] the Memphis, Tenn.-based shipper expected to move more than 17 million packages worldwide, double the company’s average daily volume.”

        “UPS expects its peak shipping day to fall on Dec. 22, when the company projects a worldwide delivery volume of 26 million packages, 60 percent over normal daily volume”

        So, between FedEx and UPS, their previous record peak capacity (at least back in December, 2011) was 43 million packages (worldwide); however, their respective average daily volumes of ~8.5 million and ~16 million packages only leaves respective “spare peak capacity” of ~8.5 million and 10 million. So, in theory, between FedEx and UPS, they could deliver an additional 28.5 million packages if they expected to do so.

        However, this likely overstates their surge capacity since they ramp temporary workers for Christmas season and are unlikely to do so as aggressively for a single day surge. Also, it’s likely that not all of this capacity would be available in the 9 countries to which Apple was initially shipping.

        Regardless, it’s something of an interesting upper limit.

        reinharden

      • Tatil_S

        The bottleneck might actually be the trans-Pacific flights. I believe landing rights of cargo planes in Chinese airports are negotiated years in advance, so there is probably no spare capacity other than displacement of some other shipments temporarily. When the goods are already in the US warehouses for Christmas, some arriving through ships rather than planes, it is probably easier to temporarily increase distribution capacity.

        From Bloomberg:
        http://www.businessweek.com/magazine/apples-supplychain-secret-hoard-lasers-11032011.html

        “To ensure that the company’s new, translucent blue iMacs would be widely available at Christmas the following year, Jobs paid $50 million to buy up all the available holiday air freight space, says John Martin, a logistics executive who worked with Jobs to arrange the flights. The move handicapped rivals such as Compaq that later wanted to book air transport.”

    • Tatil_S

      It cannot be that difficult. Santa has been doing that and more every Christmas for centuries.

      • Steve W

        If you got Santa to bring you an iPhone, then I am ordering through you this Christmas. You must be “Xtra Nice”!

      • JohnDoey

        If Santa is the only example, that suggests no.

    • franksspam

      That is a very good point. I think that once we get close to that we could see a shift in preorder policy requiring even those that preorder to pick up from the store.

    • orthorim

      My guess is the task of physically making the devices is a much harder problem than distribution once they’re built. FedEx, UPS, DHL, USPS – these companies kick ass when it comes to delivering lots of stuff all over the world quickly. If you don’t have enough, get more planes there. If one of these is at capacity, add another one.

      Making more devices if FoxConn is at capacity? That’s a lot harder. Getting a new production line up and running is something that has a lead time of months, that’s very expensive, and that’s even before taking into account that all the suppliers need to make more of everything too, and ship it different places.

    • Martin

      That’s my thinking too. Unlike holiday sales where the delivery comanies hire extra workers, they can’t really justify that for a weekend. Further, unlike most of their residential deliveries, every iPhone required a signature which is MUCH slower than just leaving on the stoop. My wife received hers at 9:30am on Friday and the UPS guy said he truck was mostly iPhones and that he was going to have a very long day.

      With store shipments, those are bundled up in bulk and are MUCH easier to deliver. Also much more efficient as the individual shipping boxes are 2-3x larger than the retail box.

  • LTMP

    I was surprised that my carrier (Rogers in Toronto) offered me a $50 credit if it took longer than 2 weeks to provide me with my pre ordered iPhone 5. I was very hopeful that Apple had solved it’s launch supply issues because of this.

    I don’t know if this offer was made to all Rogers clients who were at the end of their contracts, but it suggests that Rogers was very confident in Apple’s ability to supply phones as fast as they could sell them.

    • http://www.facebook.com/bob.arker.731 Bob Arker

      Not quite sure if this piece of anecdotal evidence says anything. You could have been first on the line for pre-orders and Rogers could have been very sure that they would receive a next shipment before 2 weeks.

      • LTMP

        That is entirely possible, and that is why I mentioned that I didn’t know if the offer was made to all customers who were up for renewal.

        I think that the evidence is somewhat more than anecdotal, however, since this is a case of Canada’s largest carrier making an offer that I haven’t heard of before regarding the iPhone.

      • Tatil_S

        “You” have not heard before.. “You” received an offer… It still sounds like an anecdote to me. :)

      • LTMP

        Just to be a nitpick, Anecdotal refers to a data set of one. For instance, my iPhone broke, therefore all iPhones are crap. This is an argument based on (possibly) non representative anecdotal evidence.

        If a bird watcher gets a photograph of a bird and says “I photographed this bird, and it has never, to the best of my knowledge, been reported or photographed before” this is not anecdotal.

        From the first statement, all that can be determined is that one iPhone broke, it may be a symptom of a larger issue, or it may just be one faulty phone, or it might even be one user who doesn’t know how to use the phone properly.

        The second statement shows evidence of a brand new phenomenon, and is empirical in nature, despite the fact that, upon further review by peers, it might turn out to be a previously sighted bird.

        Despite the fact that I believe that Rogers should treat me as an individual, and make offers to me that are better than the offers it makes to anyone else, I think that it is safe to presume that the offer was made to many other people.

        I put forth the the argument that this offer might be indicative of Rogers confidence in Apple’s supply capabilities. Other hypotheses are certainly possible and are equally valid.

      • LTMP

        Just to be a nitpick, Anecdotal refers to a data set of one. For instance, my iPhone broke, therefore all iPhones are crap. This is an argument based on (possibly) non representative anecdotal evidence.

        If a bird watcher gets a photograph of a bird and says “I photographed this bird, and it has never, to the best of my knowledge, been reported or photographed before” this is not anecdotal.

        From the first statement, all that can be determined is that one iPhone broke, it may be a symptom of a larger issue, or it may just be one faulty phone, or it might even be one user who doesn’t know how to use the phone properly.

        The second statement shows evidence of a brand new phenomenon, and is empirical in nature, despite the fact that, upon further review by peers, it might turn out to be a previously sighted bird.

        Despite the fact that I believe that Rogers should treat me as an individual, and make offers to me that are better than the offers it makes to anyone else, I think that it is safe to presume that the offer was made to many other people.

        I put forth the the argument that this offer might be indicative of Rogers confidence in Apple’s supply capabilities. Other hypotheses are certainly possible and are equally valid.

      • Tatil_S

        I was just pulling your leg, but you still got only one anecdote and a claim that this offer was never seen before. We still don’t know whether Rogers made these offers unbeknownst to you (I assume you are not a Rogers analyst unusually familiar with its marketing schemes). We also don’t know whether this offer is made to every customer within the first 24 minutes or 24 hours, to customers who spend over $200 per month or who were with the company for more than 10 years or maybe who only joined two years ago (if Rogers saw that people tend to jump 2 years after joining and it wanted to retain more of these customers). That is about as anecdotal for trying to draw conclusions about Apple’s or Rogers’ confidence in having enough supplies as one person’s failing phone for judging the reliability of a handset.

      • LTMP

        Fair points, all.

        I have to concede that my hypothesis is premature given the lack of additional data.

        Ironically, this is the very thing I was discussing with Horace in another thread of comments on this posting :)

      • LTMP

        That is entirely possible, and that is why I mentioned that I didn’t know if the offer was made to all customers who were up for renewal.

        I think that the evidence is somewhat more than anecdotal, however, since this is a case of Canada’s largest carrier making an offer that I haven’t heard of before regarding the iPhone.

    • KirkBurgess

      An opposing view is that Rogers knew that it would not be able to keep up with preorder deliveries, so was offering a credit to stop people defecting to other carriers that could provide an iPhone earlier.

      • LTMP

        I had considered this, and it is a valid point. We have six carriers in Canada that sell the iPhone now, so competition to keep customers is always a big factor.

        However, since each carrier is likely to have the same supply constraints, I think that this is, at least, not the only factor involved.

      • LTMP

        I had considered this, and it is a valid point. We have six carriers in Canada that sell the iPhone now, so competition to keep customers is always a big factor.

        However, since each carrier is likely to have the same supply constraints, I think that this is, at least, not the only factor involved.

  • LTMP

    Horace: As a counterpoint to this post, I think it might be useful to look at the first weekend results from the viewpoint that sales growth is actually slowing substantially.

    There are certainly a lot of factors that could contribute to this, such as maturation of the US market, the economic situation in the EU and the availability of Android devices that are ‘good enough’ at lower price points.

    These factors undoubtably have had an impact on iPhones sales. I think it would be wise to try to determine just how big this impact has been, and will be.

    • AdamChew

      I don’t know whether you realize that the some online stores in the nine countries which offer the iPhones have no stock of the phone. I tried to order one but none is available.

      Believe me the demand is much much better than what many said.

    • Accent_Sweden

      We’ll have to wait until the fourth quarter report in about a month to know the sales until end of September. Only then will we know how fast Apple was able to deliver and demand in general. That will allow us to determine whether Apple has saturated the market and peaked on iPhone growth. I doubt they have, but I don’t see any other information that will tell us this before we get next month’s numbers.

    • franksspam

      Your analysis is extremely premature when you consider the fact that if you try to order an iPhone 5 from Apple right now it will give you a 3-4 week delivery time. There is no way that you can square what we know, 5 million confirmed sold, double the preorders, longer lines than the 4S, 3-4 week wait, with your analysis.

    • franksspam

      Your analysis is extremely premature when you consider the fact that if you try to order an iPhone 5 from Apple right now it will give you a 3-4 week delivery time. There is no way that you can square what we know, 5 million confirmed sold, double the preorders, longer lines than the 4S, 3-4 week wait, with your analysis.

      • http://www.facebook.com/zsolt.vasvari Zsolt Vasvari

        On the opposite side, from the 5 million you need to subtract the product that’s in stores and not in customers’ hands. Are there more people with backordered online sales than unsold product in stores? No idea.

    • http://www.asymco.com Horace Dediu

      There are three pieces of data: (1) 24 hour pre-orders were 100% higher than the year before, (2) Deliveries on launch day were “sold out” in one hour and (3) opening weekend sales were 5 million or 25% higher than the year before over roughly the same distribution. I put forward a hypothesis which explains all three pieces of data. Which alternate hypothesis do you propose will fit these same data?

      • JohnDoey

        But but but … Android.

      • LTMP

        I think it is fair to say that the data points don’t tell us all that much about demand. Since supply has been far outstripped by demand at every launch, only longer term data can offer insight into demand growth.

        These three data points only tell us that Apple has managed to increase launch weekend supply by 25%, and that this increase has still failed to meet launch weekend demand.

        These are good indicators that demand is probably at least as high as it was at the launch of the 4S, but it doesn’t (on its own) paint a reliable picture for demand over the next quarter.

        In order to do that, I think we need to try to look at larger changes to the spending habits of mobile phone buyers.

        I realize that gathering the data to forge a decent analysis might be impossible, and I certainly don’t think that I’m up to the task of gathering the data or interpreting it.

        It would be hard to argue that there aren’t significant factors pushing downward on iPhone sales. So it seems to me that it is wise to try to quantify that downward pressure.

        I’m long on Apple, and have a very large portion of my retirement savings with them, so my knee jerk reaction when reading your post was to cheer. My second reaction was that, while your hypothesis holds up, it might not be relevant.

      • LTMP

        I think it is fair to say that the data points don’t tell us all that much about demand. Since supply has been far outstripped by demand at every launch, only longer term data can offer insight into demand growth.

        These three data points only tell us that Apple has managed to increase launch weekend supply by 25%, and that this increase has still failed to meet launch weekend demand.

        These are good indicators that demand is probably at least as high as it was at the launch of the 4S, but it doesn’t (on its own) paint a reliable picture for demand over the next quarter.

        In order to do that, I think we need to try to look at larger changes to the spending habits of mobile phone buyers.

        I realize that gathering the data to forge a decent analysis might be impossible, and I certainly don’t think that I’m up to the task of gathering the data or interpreting it.

        It would be hard to argue that there aren’t significant factors pushing downward on iPhone sales. So it seems to me that it is wise to try to quantify that downward pressure.

        I’m long on Apple, and have a very large portion of my retirement savings with them, so my knee jerk reaction when reading your post was to cheer. My second reaction was that, while your hypothesis holds up, it might not be relevant.

      • LTMP

        I think it is fair to say that the data points don’t tell us all that much about demand. Since supply has been far outstripped by demand at every launch, only longer term data can offer insight into demand growth.

        These three data points only tell us that Apple has managed to increase launch weekend supply by 25%, and that this increase has still failed to meet launch weekend demand.

        These are good indicators that demand is probably at least as high as it was at the launch of the 4S, but it doesn’t (on its own) paint a reliable picture for demand over the next quarter.

        In order to do that, I think we need to try to look at larger changes to the spending habits of mobile phone buyers.

        I realize that gathering the data to forge a decent analysis might be impossible, and I certainly don’t think that I’m up to the task of gathering the data or interpreting it.

        It would be hard to argue that there aren’t significant factors pushing downward on iPhone sales. So it seems to me that it is wise to try to quantify that downward pressure.

        I’m long on Apple, and have a very large portion of my retirement savings with them, so my knee jerk reaction when reading your post was to cheer. My second reaction was that, while your hypothesis holds up, it might not be relevant.

      • http://twitter.com/WalterMilliken Walter Milliken

        I don’t see any downward pressure showing in the data Horace cites — the total supply was larger for launch weekend than last year (5M vs. 4M), and it sold out faster. The only alternative hypothesis I can see here to Horace’s is that for some reason people moved up their buying to try to get to the front of the line, while total demand was the same or dropped off. I.e. the shape of the buyer curve has changed… but I can’t think of any evidence that suggests that has happened.

        I think we did see the shape of the overall iPhone 4S sales curve change a bit versus earlier years (by shifting sales earlier in the cycle); it looks like Apple met total demand earlier, so the Q3 and Q4 sales dropped off more quickly this year than in prior years. But the area under the curve was still much larger than for the iPhone 4.

        I do think Apple’s iPhone sales growth rate has (necessarily) slowed down in the more saturated markets. But the aytm.com survey data (Horace’s tweet points to the Fortune Apple 2.0 article about it) seems to suggest that Apple will start growing market share at Android’s expense in the saturated US market, apparently to something like sales parity over the next six months. We will have to see what the actual sales data looks like, though — a market survey isn’t very solid data.

      • http://www.asymco.com Horace Dediu

        I disagree. I think launch supply was about 50% higher than last year but it was misallocated for the first three days. Further, based on capital equipment purchases, I expect that production will increase over the next few weeks to reach a level at least 80% higher than last year’s rate. The methods I use were described first in March 2011. http://www.asymco.com/2011/03/28/predicting-iphone-sales-for-dummies/

      • Westech

        Horace, I disagree with you. Apple’s main concern was and is filling a huge supply chain and insuring that additional units are available to keep it operating. This would require an initial inventory of at least 25 million units. They pre-determined the amount they wanted to sell in the first few days to maximize getting their retail outlets up to speed. Their history is that they will feed product into the chain at a rate which will result in individual outlets running out on a daily basis while they build inventory to extend availability to the next group of countries.

        Apple would like to be able to snap their fingers and fill the chain overnight but as demand and volume increases with each new product this becomes more difficult. They are balancing the initial inventory and the time it takes to do this with getting to market ASAP and with expanding breadth of availability world wide. If they wanted to they could have sold 10 million units in the first three days but this would have compromised their ability to fill their supply chain.

        It is impossible to project consumer demand from the sales dat for the first three days.

      • Tatil_S

        Why do you think initial inventory needs to be 25 million? How many do you think Apple can make in a week (It sold about 25 million all of last quarter)? Why would it make sense to sit on this much inventory and tie up capital, rather than just launching the product a few weeks earlier?

      • westech

        They need to do it to fill the supply chain. They were making the iPhone 5 at least a month before introduction. This is the only way they can handle the business in an orderly manner. And now they are building inventory for the next round of countries.

        Production stars out slowly and ramps up. I estimate that Apple will sell 60 million phones in the December quarter, maybe 700 thousand a day. I would guess that they started out production at 250 thousand a day and have ramped up to perhaps double that amount so far. If memory serves me correctly Apple aims for a six to eight week inventory at equilibrium.

      • Tatil_S

        Yes, they need to fill the supply chain, but 12 weeks worth of supply sounds excessive to me. Luckily for both of us, Horace is delving into it in his next post. His estimate is about 16 million produced in Aug and Sept.

      • Tatil_S

        Yes, they need to fill the supply chain, but 12 weeks worth of supply sounds excessive to me. Luckily for both of us, Horace is delving into it in his next post. His estimate is about 16 million produced in Aug and Sept.

      • LTMP

        I’m learning that it is a horrible thing to disagree with you Horace.

        First, I have no doubt that you are better at analysis than I could ever be, second, I really want you to be right, and third, I get down voted a lot :)

      • http://www.asymco.com Horace Dediu

        It should not be horrible. Your comments force me to question my assumptions. I’m always in doubt of my analysis. It’s only through these challenges that I am able to gain confidence in my results. There should be no confidence in unchallenged results.

      • Westech

        Apple had a distribution plan for some number of iPhones. There was allocation to fill the pipeline, first day orders, and to hold a large number of them, maybe most of them, for replacing inventory at the various points of distribution. They determined that they wanted to sell 5 million plus on the first three days. I suspect that Apples’s beginning inventory was at least 25 million phones. Their objective was to keep a steady flow of phones to the field without having to live hand to mouth. The 5 million number is what Apple wanted to be able to report.

      • Westech

        Apple had a distribution plan for some number of iPhones. There was allocation to fill the pipeline, first day orders, and to hold a large number of them, maybe most of them, for replacing inventory at the various points of distribution. They determined that they wanted to sell 5 million plus on the first three days. I suspect that Apples’s beginning inventory was at least 25 million phones. Their objective was to keep a steady flow of phones to the field without having to live hand to mouth. The 5 million number is what Apple wanted to be able to report.

      • twilightmoon

        Android fantasy that iPhones are badly behind in features and are a market flop barely held aloft by Apple’s shiny logo and slick marketing.

    • http://www.facebook.com/zsolt.vasvari Zsolt Vasvari

      Dude, the AN-word is forbidden here, so don’t even bother.

      But, the point is, Apple has missed two analyst guidances in a row, their last quarter numbers and the iPhone 5 weekend sales. I don’t think it’s a coincidence, and I blame the economy in Europe and Asia where people gravitate to equivalent, but cheaper choices.

      • KirkBurgess

        Apple couldn’t supply everyone that wanted one.

        Ordering online takes 3-4 weeks to get delivered.

        There is no problem with demand for the iPhone 5.

      • http://www.facebook.com/zsolt.vasvari Zsolt Vasvari

        By all accounts, physical locations had plenty of supply and everybody who wanted one could have gotten one if they went to a store. Of course, there are people who are not within a reasonable distance from a store selling Apple products, but that must be a very small minority.

        The 2 million pre-orders are people who were already decided to buy whatever Apple released that day. The real story here is that only 3 million decided to get the product after the announcement.

        Sure, you can come up with a convoluted argument about imbalance of supply, but the simpler explanation is that people weren’t that interested/awed by the iPhone 5, and coupled with some of the negative press (maps, defective product), they decided to stay away or to take a wait-and-see approach.

      • http://twitter.com/jonmilani Jon Milani

        “By all accounts, physical locations had plenty of supply and everybody who wanted one could have gotten one if they went to a store.”

        Not in Canada.

      • unhinged

        I don’t think it’s a coincidence either – guidance from analysts has never reflected reality, nor has meeting it ever been a goal of Apple. :)

  • Pingback: Tuesday links: a guaranteed losing strategy | Abnormal Returns()

  • Jurassic

    This is a major difference between numbers given by Apple vs. those given by Samsung.

    Samsung gives numbers for units “shipped”. As soon as a truckload of Samsung phones leaves the factory, they are counted as “sold”, even if many of those units sit on merchant’s shelves gathering dust.

    Apple ONLY counts iPhones that have been sold and delivered directly to users. They don’t count those phones that have been shipped to users but not received yet, nor do they include pre-orders that are yet to ship.

    Even though Apple’s 5 Million iPhones sold in a weekend is an amazing number, the total of units actually sold could be much greater.

    • Steve W

      “Actually Sold”? Seems like 5 million is the “actual” number, unless you count chickens before they hatch.

      • chandra

        Did you read Jurassic’s full post?

      • JohnDoey

        He obviously meant “shipped.”

      • http://policydiary.com/ John S. Wilson

        No. I purchased online and have not yet received it, to me it’s sold, but to Apple it’s still only shipped. They don’t book it as a sale until it’s in a customer’s hand.

      • Mo

        According to the 10Q, shipped = sold.

        “Product is considered delivered to the customer once it has been shipped and title and risk of loss have been transferred. For most of the Company’s product sales, these criteria are met at the time the product is shipped. “

      • ronnyjotten

        you’re missing the important bit:

        “For online sales to individuals, for some sales to education customers in the U.S., and for certain other sales, the Company defers recognition of revenue until the customer receives the product because the Company retains a portion of the risk of loss on these sales during transit. ”
        many of these sales are online sales to individuals. the part you quoted likely applies to customers like at&t, verizon, etc.

      • ronnyjotten

        you’re missing the important bit:

        “For online sales to individuals, for some sales to education customers in the U.S., and for certain other sales, the Company defers recognition of revenue until the customer receives the product because the Company retains a portion of the risk of loss on these sales during transit. ”
        many of these sales are online sales to individuals. the part you quoted likely applies to customers like at&t, verizon, etc.

      • ronnyjotten

        you’re missing the important bit:

        “For online sales to individuals, for some sales to education customers in the U.S., and for certain other sales, the Company defers recognition of revenue until the customer receives the product because the Company retains a portion of the risk of loss on these sales during transit. ”
        many of these sales are online sales to individuals. the part you quoted likely applies to customers like at&t, verizon, etc.

      • http://www.asymco.com Horace Dediu

        Be careful with interpreting the “most” clause in the quote. Shipments to consumers are not the same as shipments to distributors/operators.

      • orthorim

        This is a bit of a nonsensical debate. How can you measure demand for a supply constrained product? Answer: You can’t.

        I want to buy an iPhone 5 but I can’t, I have to wait until it becomes available for online ordering here in Asia.

        The above article is reading the tea leaves. Doesn’t make a lot of sense to me but if you were going to do that didn’t they sell out of the initial batch in like 2 hours?

      • Mo

        According to the 10Q, shipped = sold.

        “Product is considered delivered to the customer once it has been shipped and title and risk of loss have been transferred. For most of the Company’s product sales, these criteria are met at the time the product is shipped. “

      • Mo

        According to the 10Q, shipped = sold.

        “Product is considered delivered to the customer once it has been shipped and title and risk of loss have been transferred. For most of the Company’s product sales, these criteria are met at the time the product is shipped. “

    • http://twitter.com/Gowan Gowan

      How could this be corrected? It’s very dishonest. Oh well. We’re talking about Samsung.

      Also, when talking of numbers, what is the deal with Amazon not sharing numbers? Obviously, it’s because the number is super-low and they’re ashamed to say… otherwise, why not report the number?

      Apple simply does it honestly and fairly. If it’s on order or in transit, it doesn’t count. Any other way is bogus. Not announcing numbers is, likewise, bogus.

      • http://vonnau.com Chad von Nau

        I don’t think it’s dishonest, it just reflects the on the structure of the company. Apple is tuned towards interacting with individual people whereas Samsung works with distributors. They are both honestly reporting their perspective of when the device is sold.

      • orthorim

        +1 exactly. I see Samsung phones all over the mall here in SEA, sold in literally thousands of little one-man shops. How is Samsung supposed to know when one of them sells a phone? They can’t.

        Apple exacts much more control over its distribution; each iPhone sold gets registered via iTunes. They know exactly how many are out there.

      • Tatil_S

        Samsung could easily collect activation data and monitor that in real time very easily if it wanted to.

    • ChKen

      I see this pointed out so many times, I went to the latest 10Q and pulled this snippet from the section on Revenue Recognition:

      “The Company recognizes revenue when persuasive evidence of an arrangement exists, delivery has occurred, the sales price is fixed or determinable, and collection is probable. Product is considered delivered to the customer once it has been shipped and title and risk of loss have been transferred. For most of the Company’s product sales, these criteria are met at the time the product is shipped. For online sales to individuals, for some sales to education customers in the U.S., and for certain other sales, the Company defers recognition of revenue until the customer receives the product because the Company retains a portion of the risk of loss on these sales during transit. ”

      Apple reports units sold, just like everyone else in the industry. The difference is that in the earnings conference calls, Apple also details current channel inventory levels. This allows analysts to calculate actual sell-thru, which should be a close approximation of customer sales, or as close as one could get.

      (edited to include a bit of the 10q missing)

      • orthorim

        This is nit-picking – it still holds true that Apple reports units sold to customers. Apple is lowering the number, even, to account for returns and other things that could go wrong from the time Apple ships to the time the customer receives, which is what the above is saying.

        Samsung and others don’t even have this data, so they can’t report it. Samsung reports phones shipped to merchants – whether the phone collects dust on store shelves or gets sold is none of their concern.

        For example, last year, Blackberry, if they were Apple, would report they sold 1,000 Playbooks. If they were Samsung, they’d report they shipped 1,000,000. The difference can be significant, particularly if people don’t buy the product as much as the company thought they would.

        This doesn’t mean Samsung’s numbers are wrong – they’re just measuring a different quantity.

  • John Rich

    My personal experience supports the hypothesis of, “of over- and under-supply.” The day after online orders began I first tried to order an iPhone 5 through Apple.com which failed due to page load/server overload. I then attempted to place the order on the AT&T website and had the same problem. In desperation I called AT&T (Oh, the horror) and after 20 minutes on hold I was finally able to give someone my money. Deliver time? 2-3 weeks. I now linger forlornly at my front door listlessly scanning for anything that might be a delivery vehicle.

  • John Rich

    My personal experience supports the hypothesis of, “of over- and under-supply.” The day after online orders began I first tried to order an iPhone 5 through Apple.com which failed due to page load/server overload. I then attempted to place the order on the AT&T website and had the same problem. In desperation I called AT&T (Oh, the horror) and after 20 minutes on hold I was finally able to give someone my money. Deliver time? 2-3 weeks. I now linger forlornly at my front door listlessly scanning for anything that might be a delivery vehicle.

    • http://twitter.com/kgbraund Kyle Braund

      thanks for the smile!

    • twilightmoon

      Your post should be rewritten in Haiku or limerick form.

      • Oak

        iPhone web stores fail

        a desperate voice order

        when will the truck come

      • twilightmoon

        A cold wind blows
        A check of my new iPhone in transit shows
        No sign
        Where it is,
        Nobody knows

      • John Rich

        Wow, this could be a whole new trend on Asymco, all discussions must be in Haiku. I’m sure Horace would approve;-)

      • http://www.asymco.com Horace Dediu

        Brevity is glorious.

    • Frank

      thanks for the lols. i’m lingering forlornly as well.

  • Diablojota

    There are people like me who would have upgraded but no longer get the early upgrade allowed as in times past for the iPhone. This has to have an effect as well.

    • Foo

      This.

    • KirkBurgess

      Online orders are showing weeks to deliver, there is no problem whatsoever with demand, it’s all about apple not being able to supply enough.

      If there was a problem with demand then the iPhone would be shipping “next day delivery” when ordering online.

      • twilightmoon

        That would indicate demand supply equilibrium not necessaritly a problem with the demand.

  • Pablo Carlier

    I would argue that manufacturing cannot be dimensioned for a 100% percent growth in day 1 shipments forever.

    There comes a point where you cam double your expected overall lifecycle production ability, but not starting on day 1, but incrementally.

    It is almost a given that Apple will sell as many iPhones 5 as all previous models combined. But it is easier (and much more efficient) to scale up production over time that to try to flash some pundits with a massive overproduction for day one.

    • Dewster35

      and one could argue that the reason of more leaks is because production started MUCH earlier than previous phones in order to tackle this increased demand. This stuff isn’t rocket science.

      • ChKen

        Plus, scaling production means adding suppliers, which leads to more channels for leaks.

    • orthorim

      I don’t think it’s even possible to overproduce on day one. Apple always hits on physical limitations – number of production lines, number of supplied components, and so on.

      To quote Steve Jobs when asked about iPad sales: “My main concern is, can we make one every 3 seconds?”. That’s the story of the iPhone.

      Would it make sense to delay the announcement for a month so you can build a bigger stock pile? Definitely not, you’d just be losing sales.

  • BroanCGillespie

    For the last three years my prediction of iPhone sales has been spot on. I have predicted Apple would sell “All of them” and continue to be correct.

    • Sharon_Sharalike

      Apple too could have predicted this. On the day they announced the preorder counts they knew how many had been built, and could have gone on to say “and oh yeah, we’ll sell five million this coming weekend.”

  • Evan

    My own experience would seem to confirm the misallocation problem. I pre-ordered early on the first day they started taking orders, by that time the estimated ship date was already 2 weeks. On the 21st my wife decided she wanted to upgrade as well, walked into a local apple store at noon and walked out with a new phone 20 minutes later. I just received a shipping notice for mine today…

  • http://twitter.com/bmike mike bradshaw

    Anecdotally, my experience in the midwest of the US is that the carrier partners got a huge increase in the number of phones and the distribution is moving away from Apple online selling most of the initial run to a more balanced approach of getting them in the broader retail channel. I could see reporting delays in accounting for things that would artificially suppress the true picture of demand. Oh to have access to the spreadsheets that show activation statistics (when the device is powered on), sales channels, predictions and adaptations made to react to this year’s consumer behavior surrounding launch. Also, the number of international markets seems more aggressive this year and units must be held by Apple to supply those launches if an on-sale date is pre-announced.

  • Pingback: The Apple Roundup: Google Maps for iOS 6 not ready to go after all | iMicroid()

  • juanm105

    On Thursday, 9/20 late in the day I stopped at Walmart and Sam’s Club in Seekonk, MA. Walmart told me they had not received any shipments of iPhones yet and didn’t think they would have any on Friday. I then checked with Sam’s Club and was told that they were making appointments on Friday for people who wanted an iPhone 5. They asked me when I wanted to come in. I checked the schedule and noticed that 10:00AM was open and told them I would take that time. They asked what I wanted and I told them Black, Verizon, 32Gb. They didn’t know what they would get in and said see you in the morning. Friday morning, at 7:10AM i drove through BestBuy’s parking lot and no one was there. I went to Walmart and was told they didn’t get any and don’t expect to get any. I couldn’t get into Sam’s Club since it opened at 10:00AM. I came back at 10:10AM and walked to the cellphone kiosk – two people were there both getting iPhones. They told me I could have a black Verison 16Gb but no 32Gb. I said no and they said they would call me if they got any 32GB’s in. I went back to BestBuy which was now open. I walked up to the cellphone desk where there were three people buying iPhones. They asked what I wanted, I told them. They said we have one for you. 20 minutes later i was an iPhone 5 owner (first iPhone, first smartphone). They told me earlier (but after I was there earlier) they had a lot of people buying. BUT I was able to walk in and buy one of the last six iPhones that BestBuy had. It was easy to get an iPhone … or at least it was for me. Later that day the nearby Verizon store’s parking lot and the two adjoining lots were packed!

  • http://twitter.com/Gowan Gowan

    For me, waiting on day 1 in a long queue to get a phone that may or may not be available is losing its luster. It use to be that having the new iPhone on Day 1 was a big deal. After a few of these iPhones, while all great, the importance of having it very soon after it launches is just not as important as years past. I feel, if I have an iPhone for 2 contract years, what does it matter if I have to wait an extra week or two in the long run? That’s only because I have matured some and realize that new cell phones, while truly terrific, aren’t going to make my life any more joyous. They’re nice (really nice) but not life-changing.

    Also, with 3 carriers, color and multiple SSDs, the phone I want might not even be in the store. I know that there are people willing to take less than what they had their heart set on going in, but I’m not one of them. If I’m going to be under contract, I want the exact device. Perhaps, people ordering online this time know they won’t get their iPhone on Day 1, but they know eventually, it’ll come to their home exactly in the configuration that they really want. Perhaps, this was a lesson they learned the hard way, two years ago when they stood in line for 4 hours and then had to compromise and not get the color they wanted or the SSD size they needed.

    Worse, perhaps for them, they had to leave the store because the carrier of choice was out of stock. I know many people who stick with a carrier and are not about to change. At that point, certainly thousands of unlucky people said “next time, I’ll just have to wait and order online. I might not get it Day 1 but at least I won’t waste 5 hours and then not even getting a phone.”

    I believe that after a month has gone by, we’re going to see some remarkable numbers.

    • Asad Quraishi

      I think you might be mistaking product maturity for personal maturity. The iPhone and 3G (we only got the 3G in Canada) were revolutionary. The 5 is evolutionary. It’s fantastic but perhaps not enough to wait hours in line for.

      Some people disagree with you and I though and are willing to wait to get one on day 1. All the power to them.

  • Mo

    You assume that there was misallocation for the 5, but no misallocation for the 4S. What evidence indicates that there was no misallocation for the 4S?

    • http://www.asymco.com Horace Dediu

      I posted this in a previous comment, but repeat it here: The quantity in question is the number of units which were pre-ordered that did not get delivered during the weekend. The 4S took about a day to reach “1-2 weeks” status. Therefore the units that were declared “pre-orders” (1 million) probably were delivered on day one and were included in the “opening weekend” sales. The 5 took one hour to reach “1-2 weeks” so the first 24 hours of preorders (2 million total) did not all get delivered. In fact, we can only guarantee that the first 40 minutes of orders got in before the “opening weekend” deadline [the company confirmed this in their press release by saying only a “majority” of pre-orders shipped by Monday]. So my point is that if we assume all the preorders had been fulfilled by opening weekend, the situation would not look so bad. Technically Apple cannot count pre-orders that did not get delivered (and signed for) during opening weekend. You have a situation where people “bought” and “paid for” product a week in advance but it does not get counted as opening weekend sales. This did not occur with the 4S.

  • Asad Quraishi

    I pre-ordered the phone, it was received by Rogers (Canada) on Friday and as of Monday I still do not have it (they were open Saturday). Rogers has been unable or unwilling (not enough margin to staff up?) to deal with the demand.

  • Pingback: 10 Mid-Week AM Reads | The Big Picture()

  • http://twitter.com/RobertZangrilli Robert Zangrilli

    Horace, I’m glad you ditched the “phone per country” metric in favor of “phone as a percentage of population.” It was quoted on a number of news sources as a sign of diminishing demand for the IPhone even though it isn’t really accurate.

  • Pingback: 10 Mid-Week AM Reads | CapitalInvestments()

  • Advill

    Great article, having a Horace approach to future stock price with same methodology could be quite interesting.
    Rgds