Who Will Pay for Mobile Data?

(Part 1: The End is Nearer Than You Think)

The most important question in mobile computing is who's going to pay for all that mobile data we're supposed to use in the next few years.  The question doesn't get much discussion online, but it's at the heart of the most intense debates in mobile, including net neutrality and the wireless bandwidth "crisis."

How many users will pay for their own mobile data service?  Will web companies also pay?  Will the government step in?  And most important, how much money are any of them willing to pay?  The answers will shape the future of every company involved in mobile, and will have a profound effect on everyone who uses a mobile phone.

Here's a quick summary of my answers:
     --Because of economics and user psychology, I think we're headed for a slowdown in the growth of mobile data.  The unlimited, exponential growth forecasts are wrong.  I believe the ultimate mobile data market will be smaller, and much more segmented, than most people expect.
     --Even if I'm wrong about user demand, we're still headed for a slowdown in growth because the cellular networks can't grow fast enough to handle all the traffic being forecasted.  This is due to physics and can't be changed; you could just as easily change the phases of the moon. 
     --Many of the proposals to "fix" the problem would probably make it far, far worse.  We could end up with a cellular data network that resembles American cable TV: slow to innovate, dominated by a few players, and subject to intense politics, with users caught in the middle.
     --In the future, cellular data for the majority of users will likely be metered, and the majority of people will need to be enticed into using it.  That creates some excellent unaddressed opportunities for everyone from handset companies to app developers. 

This is a very complicated issue, so I'm going to cover it in three posts this week:

Today's post talks about the forecasted growth of wireless data, and why I think growth won't continue the way most people are expecting.  That creates some big challenges for mobile data companies, but also some fantastic opportunities.

Tomorrow I'll talk about the alternate scenario, in which mobile data growth continues at the same rate, eventually colliding with natural limits on the amount of data that can travel over a cellular network.  I'll discuss how that collision drives the rhetoric about a bandwidth "crisis" and the debate about net neutrality.

In the third part, I'll discuss what it all means for the industry, and give my take on what we should do about it.

To start today's post, let's look at the predicted growth of mobile data...

The forecasts for mobile data growth are so sunny they could burn your skin

Every mobile phone will be a smartphone.  Smartphones are already used by about a third of the US mobile population (link, link), and ownership rates are similar in parts of Europe (link).  Horace Dediu says half of US phone users will be on smartphones by the end of 2011 (link), and non-smarthones will be virtually extinct a year later (link). 

Mobile data traffic will explode.  Cisco says global mobile data traffic will increase 26X from 2010 to 2015 (link).  The growth will be driven by increased use of smartphones, and also a rise in the number of notebook PCs connected to the cellular networks.  Notebook computers generate an order of magnitude more data traffic than smartphones, so even a small number of cellular notebooks drives a huge increase in traffic.

Mobile app shipments are off the chart.  Apple says iOS users download 62 apps per device on average, for a total download rate of 206 apps every second (link). 

The big tech companies are focused on mobile.  Many of the hottest tech companies, most recently Facebook, say their biggest focus for this year and beyond is nailing the mobile opportunity (link). 

PCs will be replaced by smartphones.  A Google vice president says smartphones will render PCs irrelevant by 2013 (link).

The growing consensus is that our current cabled, PC-centric computing world will soon be replaced by an untethered world in which everyone uses smartphones and tablets to do their computing on the go.  The mobile network we all envision will be just as flexible and carefree to use as today's wired Internet, but with the added benefit that you can use it anywhere, anytime, with a wide variety of different devices. 

It sounds cool, but unfortunately no one has ever asked users if we're all willing to pay for that mobile data service.  I think most of us aren't.

We won't pay for all the mobile data we want

We'll all carry smartphones, but...  The forecast for mobile data growth is based in part on the assumption that in the near future most or all mobile phones will be smartphones.  You can make a good case for that assumption.  The price of smartphone components is continually decreasing, so at some point the parts cost a smartphone will be the same as a feature phone is today.  Even if prices aren't completely equal, once they get close, it's more cost efficient for a mobile phone company to base its phones on a smartphone OS because it requires less rewriting of apps and support software for each new phone.  The handset companies have an incentive to switch to smartphone hardware.

So I have no doubt that in the next couple of years most phones sold in the developed world will technically be smartphones.  However, I think it's not reasonable to assume that they'll all be used as smartphones, because many users won't be willing to pay the data charges.

As I've written before (link), when I was at Palm we did a lot of research on mobile phone users in the US and Europe, and we found that about a third of them were willing to pay extra for new mobile data features in addition to voice and texting.  Some of them were more interested in entertainment, some in business communication, and some in information management.  These are the people who have been buying iPhones and BlackBerries. 

The other two thirds of phone users were not willing to pay extra for any new sort of mobile data.  Some of them didn't have enough income, some of them just weren't interested, but they all flat-out refused to consider spending extra.  The Palm surveys were conducted several years ago, but since then I have seen no evidence to suggest the basic situation has changed.  On the contrary, the most recent research I've seen was done by Forrester in 2009, and it suggested the unwilling-to-pay share of the population may have dropped from 66% to about 60%.  So there is movement toward more willingness to pay, but it's very gradual.

It's hard for me to believe that most of those 60% will be willing to add about $400 a year to their mobile phone bills just for the privilege of checking their e-mail on the bus or streaming songs from Pandora onto their phones.  And remember, that research data is based on people in some of the richest countries in the world.  It may map fairly well to other rich countries like Japan and South Korea.  But in the developing world, average personal incomes can't possibly support big mobile data bills.  Most people there will need to sip data through a straw rather than gulping it from a mug.

So I have a fundamental disagreement with many industry analysts about how the mobile data market will develop.  A graphic would help explain...

This has a huge impact on what will happen next.  The consensus view says that with only a third of the population in the US and Europe owning smartphones today, the prospects for growth are fantastic -- we can still sell to the other two thirds!  And after that we'll move on to the rest of the world.  The segmented view says that with a third of the population using smartphones in the US and Europe, we've already sold to most of the world's population willing to pay for big data plans.  In this view, data plan growth will start to slow in the US and Europe by sometime in 2012.

The best way to check which scenario is right would be to conduct some market research on user willingness to pay for data plans.  If you work in a mobile tech company, you should be doing that, urgently.  For those of us without six-figure market research budgets, there are some warning signs to look for.  If the segmented view is correct, we should start seeing more price sensitivity as we use up the late adopters of data plans.  One sign would be price promotions on smartphones...

AT&T's most recent iPhone advertising (link).

Another sign would be a shift in the mix toward lower-cost data plans...

Growth in data plans, 2010 vs. 2009.  In all five countries, growth is higher in mid to low-tier plans (under 50 euros / 35 pounds a month).  Source: Comscore (link)

This isn't conclusive evidence, but you don't get conclusive evidence until something has already happened.  There's enough evidence that we should be talking very seriously about possible saturation of the user segment willing to pay for mobile data.

What it means

So to recap, in a few years I think the majority of phone users will have smartphones but won't necessarily pay for today's data plans.  Some of the phones will connect to the web by WiFi only, while others will be on pay-as-you-go plans and won't be used for much data at all.  The situation is analogous to what happened with cameraphones.  Almost all of us have cameras in our phones, but most of us don't send picture messages because of the cost. 

This stratification of data use will have some pretty profound impacts on the mobile market:

A change in the crisis.  The first effect will be that we'd hear a lot less about the wireless bandwidth "crisis."  Operators will all of a sudden feel a lot less pressure to expand their networks and get more spectrum.  However, they will not be happy.  Slowing data growth will probably make them miss their revenue forecasts, hurting their stock prices.  Some operators may end up with excess capacity, resulting in renewed price competition in data plans, and putting more pressure on earnings.  So instead of a bandwidth crisis we'll suddenly have an overcapacity crisis.

Pressure on mobile startups.  Right now mobile apps are seen as a hot investment area because there's so much growth.  There's a lot of venture capital available.  If growth of mobile data slows, the rate of investment will slow also, as investors look for the next hot thing.  This won't be a disaster for today's mobile app companies, but it would make life harder for new entrants.  Also, companies that are investing on the assumption of endless growth might find themselves overextended.

Data will go a la carte.  But the biggest change is that to make mobile data grow further, we'll need to entice people into using it.  The challenge will be getting them to pay for little bits of data service, one app or one occasion at a time.  This requires a different sort of data plan, different apps, and a different user experience on the phone...

Enticement becomes job one

A mobile data slowdown will create an enormous opportunity for smartphone companies and app developers to create a different sort of relationship with phone users.  Most users will be perfectly willing to use data; they just won't want to pay for the plans.  The single most important task for driving mobile data growth will be to gradually entice these people into using data a bit at a time.  This creates several big business opportunities:

"Toll free" applications.  Just as we enable toll-free phone numbers in which the recipient of the call pays, we should enable toll-free apps and websites in which the app or site vendor pays for the cellular data charge.  I can picture several uses for this:
     --Some sites or apps might be willing to pay the data charge because they earn enough from ads to cover the cost.  For example, I am willing to bet that Google and Bing would both pay the data charges for a mobile search on their sites. 
     --Some sites or apps might be mobile supplements to paid PC web apps whose monthly service fee is large enough to cover the mobile data cost.  This might apply to a music streaming service or a file storage service.
     --Some third parties might be willing to cover the service fees for an app or website.  For example, the movie Rio sponsored a version of Angry Birds.  Picture them doing the same thing with a web app that transfers data.  They don't want the users hesitating to use their app, so they will pay the data charges.

Although it's easy to talk in the abstract about toll-free data apps and websites, it will be hard to implement them.  We'll need a payment clearinghouse that standardizes and manages the transfer payments between developers and mobile operators.  That was done for toll-free numbers, so I assume it should be straightforward, but there's still a lot of work to do. 

We'll also need a way to let the users know about toll-free apps and websites.  I think this is a task for the operating system -- it should identify the toll-free apps and sites automatically and enable them on phones that don't have data plans.  The operators also have work to do, because they'll need to track the data used by the toll-free apps and make sure it's not charged to the user.

It might also be good to have a top-level web domain for toll-free mobile sites.  I think .up (for "unpaid") is available.

There is also an important role for government here: Don't screw this up.  We need to be sure that any net neutrality regulations don't accidentally ban toll-free sites and apps.  It's possible that toll-free apps and sites will end up being the main way most people access mobile data, and it's critical not to cut off that possibility.  (I'll discuss net neutrality in a lot more detail in the second and third parts of this post.)

After-sales billing is critical.  Mobile and web developers have already figured this out: In many cases, your best chance of making money is to give away your base product and charge for upgrades and add-ons.  That business model becomes even more important in a world where most users don't have a mobile data plan.  How do you gradually get people hooked on your product when they're not willing to even pay for the cost of connecting to your website?

For some developers the answer will be that you just ignore those customers (and in that case you'd better base your forecast on selling to only a third of the population).  But for other developers, there will be an art in figuring out how to write a very data-efficient app or website that delivers enough value to hook a user with a data charge so low that you can pay it, at least during a trial period.  That sort of art is a great opportunity for differentiation.

Micropayment is critical as well.  Because developers need to experiment in incremental billing, it's critically important that they be able to easily bill customers in very small amounts.  The best system for doing that looks to be Google's recently-announced In-App Payments system, which is supposed to launch this summer.  Google will charge a flat 5% of your revenue no matter how small the transaction.  This is a huge improvement over PayPal and Amazon FPS, both of which charge 5% plus 5 cents per transaction (in other words, they take 40% of a 25-cent transaction). 

If the operators want to facilitate this sort of billing through their own infrastructure, they'll need to match Google's terms.  Operators that are wise enough to enable this may be able to build tight alliances with the most innovative websites and apps, but my guess is that most operators won't be able to get comfortable with a cut as small as 5%.  In that case, they should just get out of the way and let Google (and its competitors) operate.

Smartphones must entice.  This is a huge opportunity for companies that make handsets and mobile operating systems.  Smartphones today are designed for unlimited data plans -- here's the browser, click away; here's the app store, download something.  Those apps will be ignored by a user who has a limited data plan.  Instead, the phone itself will need to show the user individual functions and apps they can use for small bits of money.  Want directions?  That'll cost you 25 cents.  Want to download an ebook?  That's a buck.  Folks in Europe already understand this sort of world well, because so many users there are on pay-as-you-go plans.  But to most Americans it's a new concept.  Get used to it.  Think of mobile data like an a la carte menu in a restaurant, except that for data the options are almost infinite -- so the phone will need to learn about the user and customize the offers to his or her particular interests.

This model of infinite customization and a la carte ordering requires a fundamental redesign of the user experience of the smartphone.  That means it is a huge opportunity for differentiation, maybe the biggest single opportunity in mobile computing.  Apple is the leading vendor in smartphones for people with large data plans.  Although Android is catching up on many countries, often it seems to be selling to the more price-sensitive end of the market (note the lower sales of paid apps on Android compared to iPhone).   So it makes sense that the "enticement phone" would be built on Android.  I'd like to think Google would do it, but intuitive and well-integrated user experience is not its strong suit.  So maybe it'll be an Android vendor.  Or maybe Nokia will do it.  Or even Microsoft.  Whoever gets it right first has a very good chance to be the other dominant smartphone vendor.

Or maybe Apple will do it first, and end up the leader in all smartphone price bands.  It wouldn't surprise me.

What if I'm wrong?

So that's what I think is going to happen: there will be a natural slowdown in the growth of mobile data as we use up the customers willing to pay for it, and the most critical task for mobile data companies will be enticing people to use their services a bit at a time.  But what if I'm wrong?  What if the whole population is so excited about smartphones that everyone is willing to pay for big mobile data plans?  How does the world look then?

I'll cover that tomorrow, in part 2 (link).  In the meantime, please post comments and questions.  This is a huge, complex issue, and I don't pretend to have it all figured out.


Rob said...

Hi Mike,

I'm sure you are right about most people not wanting to pay for a fat data contract, but I think you are wrong in your conclusion.

The idea of toll-free data sites, and managing connections on a micropayment basis is just way too complicated. By necessity, you'd need to have permissions, confirmations, pricing info - and that would just make the whole process too cumbersome and offputting for these folks who don't care that much about data.

Look to sms in the US to get a clearer view of what will happen.

Originally, this was a premium feature on some plans and networks. Then it became more ubiquitous, and more expensive bundles for high users started to appear. Meanwhile, casual users who send occasional messages find that it is just bundled as 'free' in their basic package.

I think we'll see exactly the same with data. Some people will pay $100s of dollars for high usage plans. The 60% remaining will be like the occasional sms users. A basic data allowance will be a differentiating factor for them when they choose a plan, but they won't pay more for it (or at least, they won't knowingly pay much more).

These low-users are wary of bill shock, and will only use the data if they can trust it. They only want email, a bit of facebook, and occasional google mapping or wikipedia (of course - each user has a different set of low-usage needs).

I'm on a plan in the UK that meets this perfectly. The whole bundle costs about $25/month. It includes more minutes and sms than I can use, and 1 or 2 gig (I genuinely don't know) of data.

Critically - if I go over the data limit, then they block streaming/downloading, but continue to provide basic http/email. They may throttle the connection.

I have been on this for a few months and have no idea as to whether I have ever hit the limit. That's exactly the point.

Chuck Till said...

Timely! I'm looking forward to Parts 2 and 3.

I come from the network end of the business, not the handset end. I wonder if the carriers' pricing for data is truly and fully compensatory... in other words, do the the data prices actually provide an adequate operating margin and ROI if all costs are factored in. I suspect that their current pricing is not fully compensatory and is predicated -- in part -- on marginal cost calculations rather than fully loaded calculations, particularly with respect to acquiring additional spectrum.

If my suspicion is correct, a day of judgment is forthcoming -- unless Mike is correct that data demand will taper off. Meanwhile we see AT&T acquiring T-Mobile in an attempt (desperation?) to achieve a cost basis that's consonant with their pricing.

Jonas said...

I agree with Rob that toll free data sites are too complicated. There is such a huge variety in what sites people want to access.

Small affordable packages of data around 1GB/month is sufficient for most smartphone users and can be priced reasonably (I pay around 14USD/month for subscription including 1GB/month)

Still there will be users for which this is too expensive. They could pay per day and chose to only use data certain days (currently available at around 50 US cents per day).


Russ said...

I have to say I agree with Rob. If you had asked me in 2009 when all I had was an LG Cookie (sold at the time as a low end smartphone) whether I would pay for data, the answer would have been absolutely not. I was actually on a tariff where I had the option of all you can eat data for 30p a day on any day I chose to use it. The experience was so poor that in the 2 years I was on the contract (which was £15/mth), I doubt I used it more than 10 times. Fast forward 2 years and I have an HTC Desire HD with a £25/mth tariff which similar to Rob, gives me more minutes than I can use, more texts than I use and 1GB of data - way more than I use plus the phone was thrown in for free.

I'm normally in range of wi-fi, BUT, sometimes I'm not, so the quality of the phone plus the data access (and the fairly substantial amount of data) for what I would consider a fairly small extra outlay for me, feels absolutely worth it.

Christian said...

At the risk of sounding really original, I also agree with Rob, for the reasons cited by my fellow "commenters".

I would also add, that unless OEMs and OS makers change the architecture of their smartphone product(s), data will continue to gorw (maybe not exponentially, but still...). The fact is that many of us that do switch to smartphones are getting "hooked" on the seamless updating of statuses, data, emails, weather, whatever else you may think of and lo and behold, you need data to do that.

Chances are, the more apps or functionality your smartphone has, the more data you'll/they'll consume.

also, let us not forget that many of us were born when a phone was... a phone. I look at my friends' kids (mine are still a bit too young) and some family and they've never been in a world where cellphones did not exist. Furthermore, most of their own experience with a phone (being in their teens) has been with a smartphone of some kind (I do live in a first world country however). I doubt these future decision maker and buyers will shun data which is something that for them is as normal in a cellphone as voice.

Seeing that you also mention your own background in order to lend some credence/authority to your arguments, I'll do the same: 12 years working at carriers (3 of them so far: 1 telco, 2 MSO) in the product development/management groups with 2 of these being greendield operators (i.e. new networks and infrastructure, so state of the art, no pcs/edge legacy).

I'm looking forward to reading part 2 and 3!

Michael Mace said...

Thanks very much for the excellent comments, folks! This is exactly the sort of discussion I was hoping for.

I should explain some more about the market research we did at Palm. We didn't ask people directly if they were willing to pay for mobile data, we gave them a long list of potential data-related features and scenarios, and asked them over and over if they would possibly be willing to pay extra for any of them. I think we were pretty thorough about scraping in all of the people who were potentially open to data.

So I hear you that your own attitudes toward data have changed, but I don't know where you would have come out on that survey. You might have been in the huge group we classified as latent customers, who are included the 30-40% of willing data buyers shown in the chart. The fact that you're reading this weblog, and that you're using smartphones already, means you are far, far, far more engaged with mobile issues than the average phone user out there. No offense, but you are not a representative sample of the user base. Not even close.

However, that does not mean you are wrong! Since I don't have fresh research data, I can't back up my prediction; we'll just have to wait and see what happens in the market. And I'll cover the scenario of continuing growth in the second post.

Rob wrote:

>>The idea of toll-free data sites, and managing connections on a micropayment basis is just way too complicated.

I think maybe I made it sound too complicated in my explanation. I can't believe it would be more complicated than managing a network of toll-free numbers and 900 numbers. If the mobile industry can make that work, I respectfully think it can make toll-free data work as well. And the operators have a big incentive to make it work, since it would potentially be a very lucrative revenue stream (maybe the only new revenue stream for them in mobile data).

>>By necessity, you'd need to have permissions, confirmations, pricing info - and that would just make the whole process too cumbersome and offputting for these folks who don't care that much about data.

No permissions needed for the person using the free website or data app. But complicated for the company offering the site, yet.

>>Look to sms in the US to get a clearer view of what will happen...I think we'll see exactly the same with data.

I think that is a reasonable scenario.

>>Some people will pay $100s of dollars for high usage plans. The 60% remaining will be like the occasional sms users. A basic data allowance will be a differentiating factor for them when they choose a plan, but they won't pay more for it (or at least, they won't knowingly pay much more). These low-users are wary of bill shock, and will only use the data if they can trust it.

Meaning they will use much less data, which was actually the point I was trying (awkwardly) to make. The high-growth forecasts are based on all users gulping data the way an iPhone enthusiast does.

Michael Mace said...

Chuck Till wrote:

>>I come from the network end of the business, not the handset end. I wonder if the carriers' pricing for data is truly and fully compensatory... I suspect that their current pricing is not fully compensatory and is predicated -- in part -- on marginal cost calculations rather than fully loaded calculations, particularly with respect to acquiring additional spectrum.

Bingo! I think you hit the nail on the head (and you explained it a lot more succinctly than I would have).

>>If my suspicion is correct, a day of judgment is forthcoming

I'll cover judgment day in the second post.

Jonas wrote:

>>Small affordable packages of data around 1GB/month is sufficient for most smartphone users and can be priced reasonably (I pay around 14USD/month for subscription including 1GB/month). Still there will be users for which this is too expensive. They could pay per day and chose to only use data certain days (currently available at around 50 US cents per day).

Makes sense to me.

Russ wrote:

>>I'm normally in range of wi-fi, BUT, sometimes I'm not, so the quality of the phone plus the data access (and the fairly substantial amount of data) for what I would consider a fairly small extra outlay for me, feels absolutely worth it.

Thanks. Good points, and I am glad we're getting feedback from folks outside the US. The mobile markets vary a lot from country to country, and it's very had to assemble a complete picture. I wish this weblog had more commenters from Japan and South Korea.

Christian wrote:

>>I doubt these future decision maker and buyers will shun data which is something that for them is as normal in a cellphone as voice.

Fair enough.

>>12 years working at carriers (3 of them so far: 1 telco, 2 MSO) in the product development/management groups with 2 of these being greendield operators (i.e. new networks and infrastructure, so state of the art, no pcs/edge legacy).

Excellent! It's very valuable to get comments from people who have different backgrounds than me.

Steven Hoober said...

> However, I think it's not reasonable to assume that they'll all be used as smartphones, because many users won't be willing to pay the data charges.

At least in the US, a lot of operators require a smartphone have a smartphone (data) plan. I expect this to be good eventually, and limit growth of the sector, so that forces them to change their mind about this practice. I wonder how long this will take.

Anonymous said...

i have worked at and owned cell phone stores for a few years. this is what i have noticed lately:

unlike like in other cases the smartphone early adopters actually use much less data than the people just coming along now. the people who waited but are finally getting smartphones are getting them mostly for streaming media entertainment. they use youtube, pandora, etc. extensively and plan to do video chatting. for toll free to work it must include these type of apps. anyways for apps that use small amount of data people will just pay the metered cost. but radio and video that is ad supported(and free the listener/viewer) could be very attractive to a lot of consumers.

also these smartphone newcomers definitely see wifi different. they do not accept the idea that they may have to supplement the 3g/4g they are paying the cell providers for. wifi means one of two things for these people. either they can use there smartphone features for free without a data plan or that they can tether there laptop.

i just hope the carriers have not projected that the early adopters would be the heavy users and the latecomers would use less. the exact opposite is the case.

Euroclie said...

Very interesting post, as usual, Mike!

I think that part of the micro-payment issue might be handled by the carriers instead of third party companies like google (most probably a mix of both, actually).

I mean, for the end user here in France, most carriers offer "free" TV streaming on the handset, or "free" GPS navigation/map, etc. so they've got an historical record of being willing to handle the billing of such services, so that the end user doesn't have to ponder the risks involved with online payments, or doesn't have to handle yet another billing solution.

It would be quite easy for carriers to add small charges to your monthly bill - or not, depending on the plan you chose, and would also let them sign agreements with selected partners for which they wouldn't charge anything at all (toll-free sites/apps).

Otherwise, I do agree with you that the mobile data hungry users market seems quite segmented and probably already mature enough to expect a slowdown in bandwidth growth. When I look around me, I see plenty of smartphone users, and they are almost always either using capped plans or pay-per-use plans if they have the choice (cheaper).

And most people still prefer to rip the CDs they have already purchased instead of paying high monthly bills to stream music on their phones, which perfectly makes sense to me.

On the other hand, it's still unclear to me how today's youg people will handle this question. I mean, for now, as long as the parents are paying the bill, they use their smartphone as a commodity, oblivious to the costs involved, and heavily consuming bandwidth with pictures, songs, videos (Facebook, YouTube...). If we fast-forward in ten years, when those teenagers/young adults have to work to pay the bills, it will be interesting to see if they keep the same usage pattern, and if they are willing/able to afford it!

One more thing: I can remember the not too distant past where unlimited data plans were available only in the US, and here in Europe we were longing for this, it seems now that things have changed quite dramatically with such plans being now more commonly available in Europe, and not too expensive at that (I get unlimited data for less than $50 per month), while US carriers are slowly but inexorably moving away from unlimited data...

Now I'll read your second post on the subject! Thanks!

ipsi said...

I hope we don't see a move towards a 'toll-free' internet, as that's likely to create barriers to entry for startups and the like.

Why? Well, unlike toll-free numbers, where you only have to manage payments between companies in the same country (as far as I know, toll-free phone numbers do not work between countries), with websites you would have to manage transactions between companies in different countries. For example, a company in Japan might wish to make their website toll-free in Australia and Germany, and now you've got rather more headaches than you have with toll-free phone numbers. Not only do you have currency-conversion fun, but that one Japanese company will likely need to discuss things with each operator in each country (3 - 4 per country, maybe? Or more?)

In other words, if I create a website which offers something cool and exciting, and I start advertising that on the internet, getting people interested, and they start coming to my website via their phones, they won't be surprised if a toll-free number I provide doesn't work for them (or is not free), for example.

But if they got to my site from some other toll-free site (Maybe Google Search), then they're going to be a bit upset if the operator didn't tell them they'd be charged for visiting my website. However, if the operator does put in a warning message, that's going to be less people coming to my site, as that will probably be rather off-putting to them.

The obvious solution is to try and make my website toll-free. But if I'm a small startup, or even a one-man operation, that's going to be a lot of work, and could potentially be a lot of money.

So, I'd have to contact operators in every country where I think my site might be popular (If I'm lucky, I'll get enough visitors from other sources to have a good idea about that), and see if it's worth making my site toll-free for each operator in that country.

If you're a small company, you might not be able to afford that.

Conversely, it's unlikely to be a problem for the likes of Google/Apple/Microsoft/etc., as they should have enough capital to afford that, and I imagine they will already have a presence in the country, which would make any negotiations easier.

Of course, that all assumes that toll-free sites would be a manual process - you'd have to lodge a request which may or may not get granted, as opposed to having it all automated. And that as a startup, you'd actually have a prayer of being able to make your website toll-free.

So that creates problems, both for consumers (because they might miss out on a better alternative, just because that alternative doesn't have the resources to allow consumers to browse their website for free), and for companies (because they're losing potential customers).

The only people it's really going to be a win for are the mobile operators, and the companies who can actually afford to have their websites be toll-free. I suspect everyone else would end up being a loser...

Ultimately, it is a net neutrality thing, and I don't like the idea that, for a lot of people, the mobile web will become a small list of websites run by large, rich companies. And that does not sit well with me at all.

Apart from that, I do agree with your analysis, and that of the other commentators.

I certainly do not see every phone being a smartphone anytime soon. Maybe in 50 or 100 years or something, when everyone who grew up without computers are dead, but even then there will probably be people who struggle enough with technology that they will want something dead simple.

Or there will be people who don't like charging their phone up every night, and want something that'll last a week or more on a charge, and I haven't heard of any research that's likely to change that in the near future (but if anyone knows about something, I'd love to hear it!).

ipsi said...

Gah! Max of 4096 characters, so here's the little bit more I had to say:

Also, in terms of the amount I pay for data:

NZ$150 for 12GiB. It lasts for up to 180 days. If I go over the 12GiB, then a new 12GiB pack will be purchased, assuming I have enough credit (as I'm on prepay). If I don't, I will be notified, but will very likely have to start paying the plan-less rate, which is extortionate. So I could pay NZ$600/month, and get 48GiB/month, or I could pay roughly NZ$25/month, and get 2GiB/month.

And every time I top up, I get text messages, which is good. Admittedly, I have 0 minutes included, but since I very rarely call, that's not an issue.

gzost said...

Most of the people who are getting smartphones will eventually use some of the features that require mobile data. Whether it's checking the email or facebook, chatting, reading news, watching a youtube clip, downloading data for the AGPS, or all the other uses that exist and will exist - there will be something that most find useful for their particular way of using the phone. Once this is established, wi-fi only use is no longer an option. This conversion may take a while, and there will be hold-outs, but the utility of always-at-hand will win in the end. I'm sure if you'd asked people at the beginning of the 90's whether they were willing to pay for mobile phones, you'd also have gotten only a small amount of positive responses - yet now we all have mobiles.

As for data plans:
We shouldn't think of these as the monolithic, $ 20 upwards packages. There often are already smaller packages, day-access passes, and metered charging that is no longer priced insanely high, cost caps with throttling after a limit has been reached. There will be different packages for different needs.

Anonymous said...

i have another idea of what the future of toll free mobile could look like. it involves the platform makers taking over.

i can see a future where apple or google buy companies like sprint or even t-mobile(if the merger gets blocked) and than stop taking new paid customers and instead use the network backbones to provide free wireless internet for iphones/ipads or android/chromebooks. eventually data cost will be built into the cost of the device and platform subscriptions much like it is today with some ereaders.

Denis Altudov said...

Not only you are correct, Michael, this future is much closer thank you think. There is already a large software/hardware consumer company that pursues this exact route. The name's Amazon.com and the first version of the product is Kindle/Whispernet duo. The data connectivity is free, subsidized by the content - books. Everybody knows that, but what most people don't know about is Kindle SDK "Beta" terms and conditions: each Kindle "app" has a free 100kb/month data quota, and after that the app developer is on the hook to pay for the used bandiwidth at Amazon's rates. Amazon negotiated rates with all acarriers in all countries, so I as a mobile app developer (or a book writer) don't have to do that. Amazon also provides payment infrastructure. Now, the Kindle SDK seems to be stuck in "beta", likely because the hardware is not really accomodating of apps. But guess who's coming up with a new Android tablet this summer? Amazon! Guess who already has an app store? Amazon! Guess who already has the carriers agreements in place? Amazon! Guess who has a ton of credit cards on file? Amazon!

To take your thought all gthe way trough the end - we're approaching the world of low-margin devices (btw, guess who strives on low margins?) and free connectivity, both subsidised by the content/apps. This is a lot better world to have for all providers: carriers, platform developers, intermediaries, book writers, and developers. None of us want the user to look at the $30/month data plan and think "gosh, that's a lot of money I have to commit to paying right now for a benefit I can't fully grasn at the moment". As you said, Michael, this cuts off two thirds of the audience right there. We all want the user to look at the book and think "Mary said it's a good book, and I will spend two hours readin it, so $9.97 seems like a good price for the value", or "this app saves me to hours every month, surely I can spare two bucks per month for that". I love my Kindle because the cognitive load is very low - I just paid $180 for it, and I don't have to think of it until I want to use it again. I'm not a cheapskate either, I just like the cost to be paired with value, and I like the money to be withdrawn at the same moment that value is delivered. I'm not alone in that.

Now, about the toll-free web sites. A lot of people make the assumption that web and internet are synonymous, but that's not really true. Kindle is internet, iPhone apps are internet, email is internet, but none of them are "web". Facebook and Google Docs are straddling the boundary. All of these can be made to work over Amazone "AndroKindle" (well, we do need a term for that thing that should launch later this year) infastructure. As to the web sites? It's not hard to make them work - Web site developer can publish a manifest on their site, that they agree to pay access fees to Amazon, so AndroKindle device will know which sites are ok to visit, and which are not.

One thing I love about this arrangmeent is that Amazon has wrestled all of the billing control away form the carriers - even the mighty Apple could only take control over app purchases, but not the service. The other thing I love about it, is that it creates the right incentives for everyone - Amazon has incentives and mojo to bargain with carriers, carriers have incentive to lower the cost, app developers have incentive to reduce bandwidth use.

And all of that is likely coming by the end of the year. How exciting is that?

Denis Altudov said...

Love your work, by the way. So much so, I once driven to a conference just to shake your hand. Don't know if you remember that - it was in Seattle, and I also asked if my business's bedrock belief that "wireless network connectivity will remain spotty and therefore sync will be required" will stand against time. I have since quit my day job and now do my iPhone apps full time. :)

Ram said...


While you are right on most counts - the 60% of 'late-adopters' will use very little data and not contribute to the data tsunami, you seem to have missed the multi-device issue. The 40% of 'early-adopters' will have 4-5 mobile terminals - tablet, PC, and smart phone, all with 3G/4G connectivity and all of them consuming bandwidth. I don't know of a lot of carriers offering data packages that can be shared by multiple SIMs. Unless that happens, this segment of users will have multiple ways to grow mobile data

Kyith said...

This problem is not new. Telco 2.0 have talked extensively about this. In their recommendation, the model is to be able to monetize not just the end users but also the providers of the data content.

The key is that the telco will have to be seen more than a pipe but be a media company.

We might start to see alot of acqusition of media by telecoms as well.

Investment Moats.com

Stefan said...

Hi Mike,

What are your assumptions on the development of data plan prices?
The figures in your blog post seem very high to me.(I pay 4Eur/1GB and can cancel the contract any time).
Aren't they going to fall eg because of over-capacity?

Anonymous said...

Hi Mike,

similar to other commenters I think it won't work the way you describe it. I think the only way to sell something to those 2/3 of the users is to offer some resonable data quota for a small fee. Like 300Mb for say 5-10$. And do not expect to sell anything significant on top of this.

Just look at the history: something very similar has already been done in Europe on a large scale and failed miserably: the WAP.

It had the same merits: telcos worried about their bandwith and also tried to extract the money from the content. The only way to achieve this is to remain at control, i.e. build your own tightly controlled sandbox. Or the services and money will just flow past them. And thus the WAP was born.

It also had the "toll-free" content area and also offered "pay per view" billing possibility. And of course it offered micropayments, where user could buy something (content, tickets, games etc) and get it billed on his/hers phone bill.

OS support was also present all this time: just look at any 5-10 year old Nokia. It will make you crystal clear what will cost you money and what won't. And I think this was exactly the reason why almost noone ever went past the "from now on this will cost you money" message. It effectively worked as an entry barrier.

Users were simply not willing to be enticed that way. They seem to be more willing to buy some (semi-)flat service up-front and do not worry about the costs. Just look how the plans here in Europe are changing.

Granted, the WAP was not a particularly flashy piece of thechnology, but it was quite adequate for many services and not worse than menu of the phone itself.

Interestingly enough, the Microsoft seem to be after the same model. They promised telcos to reduce the data traffic. So far they allow just the http connections while carefully tagging request headers with application credentials. Which allows to differentiate billing per application. Just the same way as most JavaMIDP implementations did. I am curious what restrictions the upcoming socket implementation will bring. I predict the same sad pattern as in JavaMIDP.

Yann said...

Excellent and very enlightening post!

I guess most readers of this blog would always be ready to pay anything for their precious data plan so may not realize that not everybody would, but at 50 EUR a month for a typical iPhone plan in France it really gets a lot of people thinking...

I also think it's safe to say that the demographics argue in your favor: the early adopter who already are smartphone owners are also probably the most bandwidth eaters. Whatever remains of the available market to capture for smartphone manufacturers will not be a high data consumption population.

Finally, although I did not believe initially the idea of toll-free sites and apps, I then realized that this ALREADY EXISTS: 0.facebook.com is one such example, and it works like hell here in India - despite data plans being ridiculously cheap. I don't know of other examples, except for operator apps (e.g. Orange radio on iPhone in France).

Anonymous said...

When PFGBest wanted to create the world's first real-time FOREX quotes app, they chose McGinley Media. iphone app developers
droid application development

Paul said...

I'm a few months late, but thought I'd chime in.

The underlying problem with "toll free" is that there is likely to be very little "content" that yields sufficient revenue to cover the mobile data costs. Search advertising might be the one workable example. Video almost certainly will not work.

Michael Elling said...

Michael, mobile bandwidth is 20-150x overpriced. What happened to Moore's and Metcalfe's laws over the past 12 years? You are absolutely right that the carriers are killing the golden goose, but for the wrong reasons. Michael Elling @infostack

Anonymous said...

I agree with your comments about a coming mobile data crunch. If the exponential growth of people using mobile devices does indeed skyrocket, we are going to see severe bandwidth problems. In fact here in the UK, we are seeing them right now.

You neglect to mention pre-pay customers in your analysis, I am one of these. I currently pay

Three UK

Broadband Lite 1GB £10
Broadband Plus 3GB £15
Broadband Max 7GB £25
Pay Per Day 500MB* £2.99

As you can see non of these are unlimited, there is no unlimited package.

So all Pre-Pay customers pay for exactly the data that they use. The rates are far more than contract customers, seemingly for no good rational reason.

But I digress, the bandwidth crunch in the area in the area that I live in is self evident, especially at peak times. Streaming video is usually unwatchable, because of the bandwidth limitations. I suspect that this is because Three are unwilling to invest in expanding network capacity where I live.

I would be willing to pay a premium for a service that offered me high constant data rates, but this is not an option. Eventually as you say I suspect paying extra will not be an option because the network will simply run out of available limited data 3G spectrum.