iPhone Launch Patterns

When the iPhone 4S launched, one million units were pre-ordered and 4 million units were sold during its opening weekend. That made the daily rate during the 4S weekend 1.3 million units/day or one third faster than the pre-order rate of 1 million units/day.

When the iPhone 5 launched, 2 million were pre-ordered and “over” 5 million were sold during during the opening weekend. That made the daily rate during the  launch weekend about 1.7 million which was about 15% slower than the pre-order rate. However, a few months later the 5 launched in China setting an opening rate of 2 million in three days or about 666k/day. Adding China’s rate to the Rest of World rate yields about 2.4 million/day or about 20% faster than the pre-order rate.

When the iPhone 6/6Plus launched, 4 million were pre-ordered and 10 million were sold during the opening weekend. That made a daily rate during the launch weekend about 3.3 million, again lower than the 4 million/day in pre-orders. However, just like the 5, the 6 launch excluded China. If we assume that a China launch would have run 30% faster than the 5 launch[1] then my estimate of launch performance for the iPhone range is shown in the graph below:

Screen Shot 2014-09-22 at 12.35.01 PM

I included in the graph the various other launch volume data we have available.

I also included lines showing how pre-order volumes relate to weekend values for the products where we know both.

It therefore does not seem improbable that had China been available (and at the time when it will be) the iPhone launch weekend rate for the 6/6Plus combo would have been about 4 million/day. A rate consistent with the history for the product.

  1. Considering that this year China distribution includes China Mobile a 30% increase from two years ago is, in my opinion, conservative []
  • guy

    I have to wonder if the “over 10 million” referenced in the press release is lowering the bar somewhat for future years (e.g. maybe the actual number was 11m but this wasn’t disclosed to ease the comparison for next year)

    Compare the wording:

    “Apple® today announced it has sold a record-breaking nine million new iPhone® 5s and iPhone 5c models…”

    “Apple® today announced it has sold over 10 million new iPhone® 6 and iPhone 6
    Plus models, a new record…”

    • Jonshf

      Or, as Sacto_Joe has quite wordedly stated, the number is supply constrained. All they have to do is produce more before the opening weekend and then they have the option to add more countries to the mix. I’m sure the number will increase in a conveniently steady manner year after year.

      The opening weekend is exciting yet all but meaningless. We’ll have to see how quickly Apple can catch up to demand and where the sales rate levels off to on a longer term basis.

      • Sacto_Joe

        Is “wordedly” a word? #:^)

      • Jonshf

        If you understood the meaning then by definition it succeeded in being a word, although perhaps not in the exclusive club of any official English language, at least not yet.

  • chaodoze

    If Apple were supply constrained, why would launching in China increase the launch rate so significantly?

    • Jeff g

      Presumably because they would have preplanned in such away to increase supply in preparation for known China demand. Apple is phenomenally adept at budgeting their “supply constraints” to expertly balance a multitude of manufacturing, marketing, sales and distribution concerns.

      • Sacto_Joe

        And yet…every year for the last several years they have multi-week shipping backlogs for months. If the hens don’t lay, you can’t sell the eggs.

      • EVula

        Those shipping backlogs are for phones they sold, so your “you can’t sell the eggs” comment isn’t exactly accurate in this instance.

      • Sacto_Joe

        Apple doesn’t count them as sold until they ship, and neither do I.

      • Mark

        As question for my edification: does Apple count an order when it is shipped or when the credit card is charged or payment made?

      • Sacto_Joe

        My understanding is that the credit card is charged when the article is shipped.

    • Emilio Orione

      Removing constrains cost a lot of money and usually initially constrained supply meets demand in a few months. Does the initial constrain is a problem? It creates hypes, it saves money for buying additional production lines that a few months later would be unuseful.
      Production capacity is created based on the estimates for the entire production volume of the product. Initial ramp to meet initial demand is done with preproduction (before the shipping date). Preproduction quantity is time constrained and so the initial supply constrain is inevitable.
      It is good thing for Apple anyway since initial customers will probably wait for their iphone a little bit more instead of choosing another phone.

      • Sacto_Joe

        “Removing [constraints] [costs] a lot of money…” Unless you know something I don’t know, making money costs money.

        “…usually initially constrained supply meets demand in a few months…” You left out that those “few months” are during the holiday season. How many sales is Apple leaving on the table over those few months? (More on this lower down.)

        “It creates hype…” Do you honestly believe Apple is holding back production to “create hype”? If so, that’s an opinion, and one I very much don’t share. Does it “create hype” whether or not it is on purpose? That’s the wrong question. The right question is: Would it create even more “hype” to actually have sufficient product on hand to match demand?

        “[It saves money to not buy additional production lines that would be useless a few months later]” Maybe and maybe not. What’s it worth to Apple to sell, say, 30-50 million iPhones that it wouldn’t otherwise sell (that’s my guess on how many iPhones Apple will leave on the table this winter)?

        “Production capacity is created based on the estimates for the entire production volume of the product.” That becomes a self-fulfilling prophecy when production capacity is far below yearly demand. Here’s how ridiculous this idea can be: If Apple had a production capacity of 1 million phones a year, it could guarantee selling every one of them over the space of a year. Would it be a good idea? Nope.

        “Initial ramp to meet initial demand is done with preproduction (before the shipping date).” That’s true. But there are limiters. For example, until the design is “locked”, you can’t begin production.

        “Preproduction quantity is time constrained and so the initial supply constrain is inevitable.” Also true. But that’s the whole point: The initial supply depends completely on the production capacity and the date production starts. If you curtail either of those too much, you can’t come close to matching demand. But even that presumes that you created a highly desirable product in the first place – which you can’t really know until the product gets out there.

        “It is [a] good thing for Apple anyway since initial customers will probably wait for their iphone a little bit more instead of choosing another phone.” First, “good” is relative. Is it as “good” as selling another 30-50 million iPhones between now and a year from now? IMHO, no. Second, note the word “probably” in there. You should have used the word “might”, because we don’t really know if people will wait. We don’t know, for example, what other smartphones are about to be unveiled. We don’t know how effective the negative campaign of Apple’s competition may turn out to be. We don’t know if the economy will drop into the toilet a month from now and everyone will begin once more to husband their cash and credit.

        And thus, chaodoze’s original comment is still relevant.

        Here’s the thing: I totally agree that it’s chancy for Apple to utterly max out production far in advance of sales. But it has been overly cautious and left sales on the table for years now. I know, because I’ve been pointing it out for years now. It would be one thing if Apple was living hand to mouth rather than being flush almost beyond measure. I could understand that. But with the kind of cash Apple has available, it makes little sense to me for Apple not to be a LOT more courageous on this issue, at least every once in a while. This year, if Apple had had 20 million iPhones to sell, I have no doubt that they would have sold all 20 million in the first three days.

        Now wouldn’t THAT have been something!

      • tmay

        Perhaps the constraint is simply an all too common production glitch according to the Wall Street Journal referenced below:

        It is stated that the current iPhone 6+ screen yield is 50-60% and that the 6 is 85%.

        It is also noted that Displaymate rated the iPhone 6+ screen the best LCD screen ever tested and only one of two to get all “greens”.

        Apple has a history of pushing some technologies into production that aren’t fully baked, yet seem to get it ironed out pretty quickly. I’m not sure more production lines would have made the difference in this case, but I am happy that the display technology is such a high standard.

        Funny how 365 x 540 looks almost like 200 M units yearly capacity, surely a bit higher than what will actually be needed, and even then, with the 6 and 6+ be next year’s mid range products?

      • tmay

        let me restate the last paragraph more accurately:

        “Funny how 365 days x 540k looks almost like 200m units yearly capacity, surely a bit higher than what will actually be needed, and even then, will the 6 and 6+ be next year’s mid range products?

      • MarkS2002

        I absolutely agree. Maybe Apple gets a permanent exemption; but as long as there are Galaxy’s across the street and the new Note coming out before Apple gets into China with that inventory, having a 4-6 week delay for customers who have money in their hands for the 6+ does not reflect well on Mr Cook. It would have killed me to make excuses for my wholesalers when I was in retail.

      • Tatil_S

        Would you rather wait another month for the release, so that you can stock up on more iP6+ units (and sit on that inventory for weeks) while all of those customers will still be there with “money in their hands”, none of them able to buy an iP6+ (instead of at least some who could) and Galaxies will still be across the street? Sounds like a wonderful plan, please leave your resume, so that we can call you as soon as we fire Tim.

      • MarkS2002

        Well. Tried for a response but it suddenly disappeared on my iPad and, laying in bed, I am not going to try to recreate it. Suffice to say that this is a problem that has continued since at least the days of the iPhone. Those were heady times as Jobs’ little company that could took on Microsoft and RIMM and later Android; but geeze, Louise, they are now the richest publicly held company, certainly they could finance another line to ameliorate this annual drama. It is not that they need to make all 85 MM before they open their doors; but if the analysts were predicting 10MM and assuming they could afford to run some focus groups to figure out likely demands, I expect that that stock would be in boxes and ready to go. God knows what this would have been like had they had the permits ready for a China opening as well. No, I certainly don’t want Tim replaced; but if Maps was an embarrassment, this is a problem that can and ought to be be addressed. Further, when the Watch is properly introduced, I am going to expect that here will be more than pictures of the Gold Watch at

      • MarkS2002

        You will have to fill in the last sentence, since the editing function is not giving me a keyboard. But I figure you have gotten my expectation for how a leading company operates.


      • Emilio Orione

        The adoption curve of new technology is well established. Initial constrain limits early adopters sales, you plan production capacity for mainstream and try to meet early adopters demand with preproduction starting as soon as you can.
        Early adopters are technology enthusiast for your product and this enthusiasm means fidelity, they will buy your product anyway.
        I know that it is better to serve demand at the moment, but demand has steep curves while production is an orizzontal line, planning production for early demand is, even if feasible there are constrains in what you can achieve with a production chain startup, a huge waste of money after the first months.
        You do it only if unmet demand can be diverted to competitors, but apple customer satisfaction and customer loyalty are high and even higher between early adopters.

    • Sumocat

      Supply is not just volume but also distribution. Not only does it take longer to ship iPhones across the ocean, but Apple has significantly fewer outlets in China than in the U.S. Apple Stores alone are 12 vs. 257. Same volume, speedier distribution. Wouldn’t matter long-term, but it does for the time-limited launch period.

      • Sacto_Joe

        Apple ships iPhones by air, not by sea. Still, you’re right. But even in that case there are solutions. Apple could ship a lot more to their stores and outlets rather than individuals. Customers who really wanted their iPhone right away would be only too happy to travel a bit to pick them up.

      • Sumocat

        I meant “ship” in the general sense, but yes, I suppose that is the correct inference because I used “across” instead of “over” the ocean.

    • 程肯

      Horace is doing a hypothetical to compare oranges to oranges. The iPhone 5 did not launch contemporaneously in China and the US, so there was little to no supply constraint then. Using the same extrapolation for the 6/6+, if you add a predicted China sales launch, some time after the US launch, then you get an unconstrained total.

  • Sacto_Joe

    I agree, chaodoze. I don’t think we’re going to need to worry about demand all the way through Chinese New Year. The real question is, what’s the production rate at, and what effect will it have on product availability for fy ’14 Q4, fy ’15 Q1, and fy ’15 Q2? And that’s a question that applies across the product lineup.

  • normm

    Several of my friends didn’t preorder because they weren’t sure which model they wanted. That phenomenon may have impacted the ratio of preorders to initial sales.

  • Viking

    I am trying to understand the proper way to look at the 5C numbers from last year. Yes, last year the 5C was ‘new’. However, it was replacing the 5. To understand how well the new phones (6 and 6+) are selling it is more accurate to look at last years sales of 5S only; it looks to me like sales may be up 50%.

    Looking at 5C sales last year it appears there was significant channel fill in the first weekend numbers (some estimate 3 million units). This likely shifted sales from the Dec quarter to the Sept quarter.

    The telling comment from Tim Cook was “We would like to thank all of our customers for making this our best launch ever, shattering all previous sell-through records by a large margin.” My read of the tea leaves is because 5C sales are included in last years numbers (and were very large) many people are underestimating how strong sales of 6 and 6+ currently are. Looking at average selling price per unit (and comparing opening weekend sales last year and is year) I would expect a 10-15% increase.

    I wonder why Apple did not increase guidance for the current quarter (something they did last year). Perhaps sales earlier in the month were very slow as more people than usual held of purchasing a new phone. Weaving it all together it looks like Apple has hit a home run with the 6 and 6+.

    • Sacto_Joe

      Last year Apple restated its projected earnings for Q4 about now. I wouldn’t be surprised if they did so again this year. If they do, I believe it may have something to do with restating projected earnings for the whole fiscal year, which ends this month.

  • DJTaurus

    Horace please clarify what you are trying to say:

    When the iPhone 6/6Plus launched, 4 million were pre-ordered and 10 million were sold during the opening weekend. That made a daily rate during the launch weekend about 3.3 million, again lower than the 4 million/day in pre-orders.

    I understand that if China was in the equation, daily rate for the opening weekend could be more than 3.3m and bigger than the first 24h 4m preorders. If i am not wrong you are emphasizing on even greater iPhone sales that could achieved. Sorry but i don’t see any statistical value in that comparison cause imagine if Apple could produce more stock for the 24h preorders.

    • Horace Dediu

      I will say it another way: The data given by Apple for 4S, 5 and 6 launches (i.e. excluding the 5S where we did not have preorder data) is consistent and a pattern can be read from it. I present all the data (and nothing but the data) on the graph shown and you can draw your own conclusions.
      Sent from my iPad

      • Roy Rodenstein

        Two patterns can be observed:
        1) Opening Weekend units/day >= Preorder units/day
        2) units/day roughly doubles with each generation (or two)

        This would imply that in two years the iPhone 7 would see 8M units/day Preorder and ~25M units Opening Weekend.

        Given the 12-block line in NYC and similar lines in SF and London, that will certainly be an urban logistical challenge, to add to the incredible production ramp needed!

        This also implies that last year’s speculation of 6-month iPhone product cycles can be put to rest.

      • ptmmac

        I believe you mean 20 million in 2 years? 25 million is 2.5 times what happened this year. I personally discovered that there were other retailers that had stock beyond Apple and AT&T. I had a business agent with AT&T that dropped the ball and didn’t place my pre-order after confirming he had my order at 7 am on order day 1. I had to find 2 phones during the couple of hours after launch. I went to 3 places: Sams Club – plenty of ATT stock, no more Verizon but only 16gig phones, Best Buy- plenty of stock for the line there- I bought a 64 gig slate gray and a 128 gig slate gray for full market price. I went to the ATT store to try and get them to use my subsidized price for the phones. They simply refused to do this. Then I discovered that I would save over $400 over the life of the 2 phones if I simply paid the sticker price. When I got to AT&T they did have some stock of phones left, but not the exact models I was looking for. So by 10 am there were very few phones left in Athens Ga. I don’t know if they have gotten any more phones in at the AT&T store yet. I expect them to get more supplies as they become available.

    • 程肯

      I don’t think there’s an inventory limit for the first 24h pre-orders. There may be a delivery issue, but no ordering issue.

  • juanm105

    You may have a number missing in “When the iPhone 5 launched, 2 million were pre-ordered and “over” million were sold during during the opening weekend” Shouldn’t there be a “five” inserted so it reads “… and “over” FIVE million were sold during the opening weekend”?

    • 程肯

      It’s a test .Those who can do the math for the missing number pass, those who can’t, don’t.

  • obarthelemy

    Is the relationship between first-weekend sales and first-Q, first-Y sales also evolving ? ie, are first-week sales a good predictor of full-quarter/year sales ?

    • Steven Noyes

      That was my thought as well. Does the first weekend put an indication on production capacity and hence points to a prediction of how Apple thinks it will sell long term?

  • Gabriel

    This should mean about 8 billions dollars of sales for the first weekend. Anybody knows how many businesses in the world produce more then 8 billions of sales per year?

  • demodave

    From where I sit, it’s interesting to see “The Rest of the World” (ROTW) refer to “everything but China”. A short time ago, all the company internal presentations that I saw at Honeywell referred to ROTW meaning everything outside the USA. Perspectives shift, of course, and now we refer to HGR (High Growth Regions), which naturally includes China. Still, it’s an interesting turn of play.

  • Walt French

    Curious to hear what this data TELL us.

    It appears that launch week sales are constrained by production. At the rumored rate of a half million new iPhones per day, Apple (only) had three weeks worth of production when they went on sale. Undoubtedly, more could have been sold if popular sizes hadn’t run out, so this is a measure of supply-line considerations, not demand.

    And because of this, management of the supply chain is on display much more than buying frenzy / demand. The reason this is significant is that in movies, some “open wide” to a hyped-up audience that gets uninterested two weeks later; other films intentionally start smaller and build their audience over weeks, on reviews and word of mouth as much as TV ads and the culture of anticipation. I have a hard time reading whether the iPhone 6/6+ is a “finally got it right” version for aficionados, or whether it’s more, “this is just the best phone today; the only reason not to buy it is the economics for an individual.”

    I think the latter, but I just get no idea how quickly this mania will sputter out. (Meanwhile, my AT&T contract permits the subsidy price in two weeks and I easily decided I would wait until AFTER a trip that would expose a new phone to special risks.)

    • handleym

      The simplest model of what is happening, in the US at least, is that the decline of the previous pricing+contract model has led to a higher bunching up of desire this time round.

      In the past everyone was more or less compelled to buy their phone 2+ years after the previous one, which meant that, because of past history as to when people bought their phone during the year, their new purchase would likewise be spread out.

      The carriers seem to be making a (somewhat half-hearted) attempt to end this regime.
      (It’s half-hearted in multiple ways. One is Horace’s usual “iPhone is a salesman for data plans”; another is that no compelling subsidy means no two year contract, and I’m a free agent able to leave ATT/VZW the moment someone else offers something better. Their attempt to square this circle has to been to throw up a wide selection of options and, I assume, hope that people will take the default option or allow the salesperson to talk them into it. In the case of ATT for example my options are
      – continue with existing plan [“essentially” $40 a month, two year contract, net value is $60 subsidy — UNADVERTISED!!!]
      – new version of my existing plan [“essentially” $45 a month, two year contract, net value is $60 subsidy — mildly advertised]
      – NEXT plan [“essentially” $45 a month, 12 or 18 month contract, net value zero subsidy — heavily advertised]
      (Net value means stripping out the cost of the phone part.)]

      OK, point is, these various moves (of which T-Mobile’s are the most aggressive) create a world where there is now much more bunching at the start gate, people who have no compelling reason not to buy the phone the day it is released.

      Longer term this seems like a problem for Apple —basically their only option is to make lots of people unhappy that they can’t get their shiny the day they want it, without “reasons” holding them back like the previous contract issues I mentioned.
      Apple can partly handle this by delaying the launch in different countries, but that’s just another variant of Apple preventing people from getting what they want.
      Other companies handle this with a more or less implicit expectation that prices will drop slowly over the next year, so that some proportion of the population self select for buying later, but Apple seems uninterested in that route.

      I don’t know Apple handles this. Maybe they can get away with doing nothing, with a situation where the rabid expect an annual or bi-annual long-line event and consider it part of the fun, while everyone else just accepts that they’ll buy when it’s convenient over the next two months? Or maybe Apple could artificially spread out demand by introducing extra differentiators (two more colors, Jade and Ruby!) but those color models are announced only going to be released in December?

  • berult

    To the court of loose tech punditry…and roaming jesters of naught, from a curmudgeon proofreader, on their annual and scripted pilgrimage to loloLand…

    A kick-ass iPhone ought be designed to espouse the curvature of badass usage. As all slated wearables eventually will.

    Aluminum, as does Apple whilst autumn riches befall, bends either from the stance till of the handicapper, or to the industrious milling of its success.

    But, bugger if I know, break, …or break for cover, you most surely will, …for they both will or ought not.

    • Sacto_Joe

      I hope you’re gathering these up, berult. Put ’em in a book, and I’d buy a copy. (Probably need to reference the issue that inspired you.)

      In any case, much appreciated!

  • Lightbow
  • Hosni

    If iPhones are identical, then a delay in the China iPhone launch would cause Apple to sell some of the inventories originally intended for China to US consumers and consumers in other nations where the launch proceded on schedule. If there are minor component differences between iPhones for China and those for other nations, then Apple would respond to the delayed China launch in September and early October by assembling more iPhones for America and fewer for China.

    For that reason, “what if” statements about Apple production and sales under the assumption of a September iPhone launch in China are not meaningful. They ignore resource and capacity constraints as well as the profit-seeking motive at Apple.