Categories

When will smartphones reach saturation?

50% penetration for smartphones in the US turned out to be fairly predictable. I tracked my predictions based on comScore data and the results are shown in the following graph.

Screen Shot 2013-01-04 at 1-4-6.50.52 PM

I began making predictions in May 2010 and updated the estimate every month since. As the chart shows, the prediction converged to early September 2012 by early 2011. The actual date was sometime in late August 2012.

The reason this prediction was relatively easy was because the rate of penetration growth was mostly linear throughout the period of 2009 to 2012 or from 17% to 50%.

In fact, the consistency continues even above 50%. The following chart shows the weekly new smartphone users add rate. No slowing seems to be evident for the few months since 50% penetration was achieved.

Screen Shot 2013-01-04 at 1-4-6.51.02 PM

So if we assume that this rate will be preserved or increase slightly, how soon can we expect smartphones will reach 80% penetration in the US?

Screen Shot 2013-01-04 at 1-4-6.51.09 PM

My guess is now August 2014.

  • deemery

    Do you see similar growth (1st/2nd derivatives :-) in Europe? What about other parts of the world? Somewhere (but I don’t remember where) I heard a prediction that smartphone adoption in 3rd World may well outpace expectations, because smartphones (rather than PCs or tablets) will become the primary on-ramp for the Internet.

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

      There is some sparse UK data which shows it to be very similar. I don’t have data for other countries.

  • http://alexandersmith.co/ Benjamin Alexander

    I tried your Walmart test, and my gut says August ’14 is a tad conservative. Based on your comments in the most recent CP I wonder if suppliers are now defining the timeline?

  • Micromeme

    Somewhere between 50% and saturation the sales process for phone vendors switches from dominated by selling to customers with feature phones to dominated by selling to customers who are already in their ecosystem or churning from another ecosystem. I think apple is already (in US) mostly selling to their own customers and snatching some BB/android customers. But where is samsung in that process? When is everyone just fighting with churn?

  • Chuck

    Do you have an estimate for the percentage of the mobile phone market that simply doesn’t want to pay for a data plan? At current prices, I’d guess (and it’s only a guess) that it’s somewhat greater than 20%.

    • El Aura

      Depends on the market. In Germany, you can get data plans from €5/month onwards (though they have really low data caps, €10/months gets you decent amounts) add another €10/month for voice and you are set.

      I would say, €10/month is not a big barrier to entry, add a €100 for entry-level or better a used smartphone and a lot of people can afford it.

      • http://www.facebook.com/profile.php?id=741717344 Dick Applebaum

        “€10/months gets you decent amounts) add another €10/month for voice and you are set”

        It is interesting: the order in which you presented the monthly service offerings… Data first, then add voice…

        Actually, it’s really all data…

        But, one wonders if you could get by with data only and messaging… With no voice/texting plan.

      • Peter Millard

        This. Here in the UK you can get a basic smartphone (Samsung Galaxy Ace or Galaxy Y) for £7.50 ($12) a month (175 mins, 5000 texts, 250Mb) on a 24-month contract – that’s it, no upfront costs at all.

        I’ve just passed an old iPhone 4 along to my father and picked up a SIM-only deal for £5 ($8) a month on a 30-day contract with 100 mins, 100 texts &100Mb – more than my Dad’s ever likely to use. But if he does, I can change the plan anytime – plenty of SIM-only plans without data.

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

      Network operators know how to manage this very well. They segment very carefully by what buyers are willing to pay and offer packages that meet their budgets. This is, in fact, the primary objective of mobile service marketing.

  • http://twitter.com/fivetonsflax fivetonsflax

    What is the source of the bumps in the forecast graph? Seasonality?

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

      Christmas. You can tell it happens every Dec.

  • Walt French

    Let me guess that a “typical” feature phone is $60 off-contract, while a smartphone is $400. (Others are welcome to construct US ASPs less casually.) So the real question about the split between the two is not much to do with the phone capability; it’s about whether a person wants internet communications (plus various game, GPS and other goodies) in a handheld, enough to pay $300+ for them.

    Some lack the means; others lack the attraction. But while the smartphone revolution is interesting in its having gone exponential by being positioned as a better phone, I think the “job to be done” approach will eventually translate the revolution into not a disruption of cellphones, but the fact that suddenly in the early 21st century, the overwhelming majority of ordinary people in wealthy countries, started carrying a personal computer with them 24X7. That seems a little astonishing.

    And not particularly captured by whether a person carries a dumb or smart phone. The interesting question is what fraction of the economically capable populace chose to have a PC in their pocket.

    “Today we’re introducing three revolutionary products. The first is a wide-screen iPod with touch controls. The second is a revolutionary phone and the third is an Internet communications device. So three things: a wide-screen iPod…”

    • http://twitter.com/FlopTech FlopTech Engineering

      Re: “That seems a little astonishing. It has played out that the pocket PC suddenly became not a dorky geegaw, but a mass-market, must-have object. …”

      Almost. Just replace “pocket PC” and “PC in their pocket” with “post-PC device” and you’ve nailed it.

    • KirkBurgess

      I think the average cost of a smartphone at $400 is only typical in the US market. Internationally the bulk of smartphones sold are of the sub $200 variety running android 2.x.

    • http://www.facebook.com/profile.php?id=741717344 Dick Applebaum

      “it’s about whether a person wants internet communications (plus various game, GPS and other goodies) in a handheld, enough to pay $300+ and hundreds or thousands more in data charges for them.”

      I think you are correct that the desirability of a PC in your pocket was launched by the iPhone… Before the iPhone, a cell phone made me feel tethered to the leash of an expensive pager. After the iPhone, the situation was reversed — I was enabled and the world was at my service.

      Your point about data charges/contracts (quoted above) is the key to 100% smartphone penetration, IMO! The smart phone manufacturer who solves that issue will remove any vestigial advantage of a dumb phone… The issue will not be resolved by the carriers or OS platform suppliers as it is antithetical to their interests.

    • indi_research

      Horace

      Have you cross checked ComScore US data against actual US carrier OS market share.?

      It seems that when you compare with carrier activations data filed with SEC (surely a more reliable source than Comscore?) Comscore greatly inflates Android’s market share compared to iPhones. This would have the effect of inflating US smartphone penetration

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

        I have looked at activation data from operators and compared it to comScore. The problem with activation data is that it does not tell us about replacement rates. comScore data may be flawed but it’s frequent (monthly) and consistent. The thing that is valuable is the trend rather than any implied precision.

    • rational2

      Let’s take a step back and look at auto ownership. Autos are lot more expensive to buy (even an old car, relative to the price of a smartphone) and to maintain (insurance, gas, etc). Yet, there is high penetration of auto ownership among driving age populace in developed countries (extend that to motorcycle/scooter ownership in developing countries). If people accepted some inconvenience auto ownership burden can be reduced, thereby freeing people from a significant financial burden. Yet, that ain’t happening?

      Why?

      Our daily living is structured to be dependent on autos, so people feel compelled to pay to own one lest they become marginally less competitive in the society due to reduced mobility.

      The same thing started to happen with the PC in the 90s. People started to feel they are missing out and becoming less competitive if they didn’t have a personal computer. Smartphones are the next step, enabling and empowering people with more capabilities (GPS, camera, wifi etc). Very soon, if it hasn’t happened already, people in developed countries will find a smartphone to be a necessity.

      All of this to drive home the conclusion that smartphones will be owned and maintained because they will be considered indispensable. It helps these are significantly cheaper than other must haves mentioned above. In many cases you can replace, or prolong life of, a full fledged PC.

  • http://beautyandthesoftware.blogspot.com/ Adrian Constantin

    Horace, your observations from the previous post on “The Last Feature Phone” could suggest that saturation might be reached closer to 100%.

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

      Yes, 100% will happen, but I chose 80% as an inflection point where I believe the predictability of the switching rate will still hold. In other words, I’m willing to project that far given current data and afterwards I expect things to slow down. The last 20% might take a bit longer.

  • Pingback: CHART OF THE DAY: Here’s When Smartphones Will Saturate The US Market | Digital Wealth

  • Jay

    Where is your Q1 2013 predictions Horace?? Waiting on it!!! you usually publish it early.

  • Antonio Marvuglia

    It is surprising you suggest assuming a preserved or increasing smartphone new users rate going into the future. There appears to be a clear peak in the Christmas 2011 season adoption rates (which shows up in early 2012 if you’re watching the running average). This is followed by an unmistakeable ~reduction~ in the smartphone new users rate since then (notice the lowest yearly peak in averaged new users rate over the entire range of data), and, given the shrinking number of possible new users, now less than half the total, the data strongly suggests we will never again see adoption rates at their winter 2011/2012 highs.

    It’s seems to be a sure bet that the smartphone penetration graph will appear as a classic S-curve If it was scaled further out. To the left of the displayed data, there must be some period of low early adoption rates, and to the right one would predict an inevitable slowing as the pool of potential smartphone users is reduced to stragglers holding out indefinitely. It is reasonable to believe there would be slow but increasing adoption to begin with, a maximum adoption rate in the middle of the curve, and a slowing rate as penetration asymptotically approaches 100%. (And I do believe it will be asymptotic for all intents and purposes; there are people who will never buy a smartphone until it is impossible to buy a phone that isn’t smart.) It would be surprising to me if the inflection point from increasing to decreasing rates of smartphone adoption ~didn’t~ occur at or near the 50% mark, and your data appears to show it just 6 months earlier.

  • Guest

    It is surprising you suggest assuming a preserved or increasing smartphone new users rate going into the future. There appears to be a clear peak in the Christmas 2011 season adoption rates (which shows up in early 2012 if you’re watching the running average). This is followed by an unmistakeable ~reduction~ in the smartphone new users rate since then (notice the lowest yearly peak in averaged new users rate over the entire range of data), and, given the shrinking number of possible new users, now less than half the total, the data strongly suggests we will never again see adoption rates at their winter 2011/2012 highs. 

It’s seems to be a sure bet that the smartphone penetration graph will appear as a classic S-curve If it was scaled further out. To the left of the displayed data, there must be some period of low early adoption rates, and to the right one would predict an inevitable slowing as the pool of potential smartphone users is reduced to stragglers holding out indefinitely. It is reasonable to believe there would be slow but increasing adoption to begin with, a maximum adoption rate in the middle of the curve, and a slowing rate as penetration asymptotically approaches 100%. (And I do believe it will be asymptotic for all intents and purposes; there are people who will never buy a smartphone until it is impossible to buy a phone that isn’t smart.) It would be surprising to me if the inflection point from increasing to decreasing rates of smartphone adoption ~didn’t~ occur at or near the 50% mark, and your data appears to show it just 6 months earlier.

  • Pingback: How Samsung Became The Hottest Smartphone Company On The Planet (SSNLF, AAPL) | Digital Wealth

  • Frankcapra03

    This would imply Apple growth will accelerate. You have the same penetration growth curve for next 2 years + all the upgrades + gains in market share, could really see Apples growth accelerate exponentially.

  • Pingback: When will smartphones reach saturation? | Trend Spectrum

  • Pingback: Four Steps to Embracing Email Marketing In The Mobile Era | The CUnet Blog

  • KimBesWork

    What is the population basis when you say 50% or 80%……50% or 80% of what exactly? – is the penetration rate based on the entire US population of 315mi+
    – or adults only or…..???

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

      The survey measures a population of cell phone users who are older than 13 and not using employer supplied phones.
      Sent from my iPad

      • KimBesWork

        Even more compelling! Thank you.

  • Pingback: Putting the Galaxy Gear into context | От работы кони дохнут x_Х