I’m hearing refreshing outbursts of sanity recently on mobile-data pricing (and puzzlingly, grumbling from people I normally agree with). Usage-based data pricing is inevitable. Just because the rumblings are coming from phone companies doesn’t mean they’re wrong.

The Problem · Here’s the simplest way I can put it: Fixed-price unlimited-volume data pricing is a totally, unfixably broken idea. Because once the network operator has your monthly payment, they’re powerfully incented to keep you from using the network; there’s only downside in people enriching their lives via the Internet. It’s painfully obvious that the world really needs a pedal-to-the-metal damn-the-expense buildout of wireless data capacity. But how on earth are we going to get that if the incentives are all backward?

Why? · First, by way of background: Verizon warns demand will outstrip LTE capacity ‘as early as 2013’. Now, one ought to take the regulatory filings of any large business with a grain of salt; but this one rings true to me. Everyone I know with a smartphone keeps finding cool new things to do with it, and every one of them makes good use of an Internet connection. Plus there are in excess of a million new smartphone users every day.

Who? · Time-Warner wants variable pricing. T-Mobile does too, and you can add Vodaphone, Ericsson, Telkomsel (from Indonesia), Telenor (Norway), and IBM to the list.

How? · I can sense Net-heads all over the world rolling their eyes and shaking their heads: “Nope, I gotta have flat-rate data”, they’re thinking, “because otherwise the telephone company will rape me.” Let’s not go into whose fault it is that people feel this way; they do and it has to be dealt with. Here’s how:

  • Ruthless transparency from the telco as to how much you’ll pay for every MB.

  • A price-per-MB that, off the top, feels reasonable.

  • A price-per-MB that falls during the course of the month as you use more and more.

  • A lightweight API so that you can obsessively check how much you spent in the last week or hour or ten minutes. So that after a while you won’t need to.

Do those things, and people will sign up for bandwidth by the bucketload, because it delivers intense real-life value.

For Extra Credit · Pay commissions! If you’ve got the economics pointing in the right direction and you want people to use more bandwidth not less, then what you do is encourage people to write apps that use lots, and pay the developers a little kick-back.

This is why I just about puked when, back in February, some telecom exec said that app developers should kick money into the pot to cover the bandwidth cost so their users wouldn’t worry about it. This in spite of the fact that the app developers are getting crumbs while the telcos harvest gold.

I tweeted: “This notion that apps should pay for bandwidth is insane. Telcos should pay developers a commission for helping them sell bandwidth.”

I got some puzzled responses to that, and at least one major news organization reached out and asked if I wanted to expand on that, on Google’s behalf. Uh, please look at that disclaimer in the right margin: I have no idea if Google has a position on how bandwidth should be priced, and if we did, I’m 100% sure I wouldn’t be the one we pick to broadcast it.

Speaking of Distortions · Based on the telco commentary, I’m optimistic that we’ll get to a more rational data pricing regime. That will still leave a horribly distorted market, particularly in North America, because of the phone subsidies.

If this problem isn’t obvious to you, read Carriers Whine: We Wuz Robbed!, an exquisitely-cynical rant from Jean-Louis Gassée.

Just imagine how much clearer things would be if people bought phones from phone builders not networks, and people paid networks just for network services not phones. Sigh.


Comment feed for ongoing:Comments feed

From: Will S (Mar 26 2012, at 08:48)

Normally I agree with you Tim, but not this time. I do not miss the per minute charged dial up days. Usage based billing is far too confusing for most people, and leads to bill shock. Unlimited with caps works because the vast majority of people don't use anywhere near their limits, and they make up for those that "abuse" it and go over their limits.


From: John Cowan (Mar 26 2012, at 08:50)

With a handful of changes this post would work fine as a rant against fixed-price wired data (and wired voice, for that matter). And yet these are the order of the day, with no sign of that changing.


From: Zach (Mar 26 2012, at 08:59)

There's one other missing piece- time-of-day based charges. The reason for these charges is that you have to build your network to handle peak load, even though most of the time that capacity will be mostly idle. By charging more for data used in non-peak periods you incentivize customers to timeshift their data usage. I'd certainly think twice about downloading that movie if it was going to eat into my "premium" data cap.

I think the most likely model will look very similar to how voice calls are charged, with two buckets- daytime data and nighttime/weekend data.


From: Bud Gibson (Mar 26 2012, at 09:17)

In the US, we have almost exactly the situation you describe with the iPad. Unfortunately, your ability to switch networks with the device is slightly hindered here in the US (can switch on 3g but not LTE). I have to say the GB pricing is a bit steep, but it is month-to-month at least.

I'd like more competition in this space.

Unfortunately, I think I'm going to have to wait for Apple to do it. They're the only ones with a big enough stick to force the telcos hands. Even that may not be enough.


From: Nathan (Mar 26 2012, at 09:49)

I think you're discounting the psychological aspect of the pay-for-usage model. I don't give a second thought to checking sports scores or telling my email client to sync every 5 picoseconds or visiting a youtube link in an email because I'm on a capped-but-theoretically-unlimited plan.

Let's say my carrier used me as the baseline for their usage-based pricing scheme, such that I would be paying about the same each month as I currently do. Every time I launched the SportsCenter app or used an ad-supported app that downloads ads from the Internet or tapped a youtube link or grabbed a file from Dropbox, I would be mentally calculating how much it was costing me. An attitude of plenty would be replaced by an attitude of scarcity. I would grow to hate websites with large images inline because reading the photographical musings of one Mr. T. Bray would be actively costing me money.

Even if my monthly bill was about the same or even less than it was before, my attitude toward my carrier would, I believe, change from uneasy truce to open hostility as I curse their name every time I have to pull down a several-megabyte PDF when I'm out of wi-fi range.

If changing the pricing scheme is what needs to happen for our infrastructure to improve, I could, grudgingly, get behind it. I'm not convinced that telcos wouldn't just use the money for even more outrageous executive compensation packages and televised mudslinging, though.


From: stephen o'grady (Mar 26 2012, at 09:53)

I'd love to see handset and service decoupled, but the price would have to come down substantially for that to work. Setting aside the price expectation that's been created by the subsidies over the years, there's the fact that the lifespan of a phone is, on average, less than that of a laptop or tablet, making their relative cost a problem.

The failure of the direct-to-customer Nexus One wasn't one of model, in my opinion, it was pricepoint.

Separately, I hope your optimism is justified with regard to the carriers. Personally, I'm skeptical, but maybe it's just because I've been paying $1500/MB for text messages for too long.


From: Tim Swan (Mar 26 2012, at 10:09)

If we lived in a society where there was free competition to build and reuse the infrastructure, I'd agree. However, we don't. Many people are restricted to the carriers they can use, which lets the carriers restrict provision and capacity pretty much at a whim.

In the UK, I am currently paying 10 GBP per month for unlimited data, unlimited texts and 100 minutes of UK calls, with calls to people on the same carrier being totally free. This company is selling on service on top of the O2 network. I'm assuming that both this company and O2 are making a profit out of the deal, which begs the question why this price point seems totally unachievable by the major carriers themselves. equivalent price points from most major carriers here are upwards of 3 times more expensive. The price point I currently pay is only available because there's provider competition there. It's also very unusual for people to use my provider (they don't advertise heavily, for example).

we should be looking at less than a penny a meg (or a cent, or a euro, or whatever). Once the main carriers agree with that, then you're on to something.


From: Norman Walsh (Mar 26 2012, at 10:14)

In some imaginary world where the telcos could be persuaded to charge a per/mb price that was reasonable, this might work. In this world, there's no evidence to suggest that they'd attempt anything less than rapacious pricing. Look at international roaming rates, for example.

I got lost in Prague. In the few short minutes that I had data enabled in order to find myself on a map, I burned through $50. Better than sleeping on a park bench in the snow, but nevertheless outrageous.


From: Nick Johnson (Mar 26 2012, at 10:18)

As always, it's instructive to look at what the world does outside the US and europe, where internet bandwidth is often more scarce.

Pay-as-you-go is common in NZ and AU, but more common is for users to buy a data plan that includes a fixed amount of bandwidth, with an overage rate for usage over the cap. It works very well, and contrary to some commenters, we don't find ourselves constantly worrying about how much checking the weather is costing us, because even in Australia, the charges are low enough that uses like email and news don't consume substantial amounts of bandwidth, and it's easy to get a cap you're not likely to go over.

It's not all roses, of course: the rates are too high, especially in NZ, and the overage charges are even higher. It's a model that can work, though.

Also, note that the same goes for wired broadband in NZ and AU - those of you living in the US and Europe only get unlimited broadband connections because the upstream pipes are fat enough you simply can't use enough bandwidth to cost the ISP appreciable amounts of money.


From: Koranteng Ofosu-Amaah (Mar 26 2012, at 10:29)

Now, now Tim... when it comes to communication networks, we have seen repeatedly that price discrimination by network operators is fool's gold despite its appeal, that network users prefer fixed pricing because of the reduced mental accounting on transaction costs, that operators can continue to make a bundle and simplify their capacity planning without recourse to fine-grained usage billing.

Read almost anything that Odlyzko has written on the subject and this one for example

Too expensive to meter: The influence of

transaction costs in transportation and



It is only with airline pricing (and that is transportation not communication networks), with the now draconian restrictions on identity and ticket transfers, that control continues to rest with operators.


From: Mike Kozlowski (Mar 26 2012, at 11:00)

What Nathan said. The psychological impact to a consumer of a metered model is brutal, and discourages use. It would do more to kill mobile innovation than just about anything. (I know that I, for one, would never install any app that contacted the internet, because do I REALLY need to be able to see the weather at a glance? Enough to pay real money for it? Can't I wait until I'm home to read Facebook? Etc.)

And you're addressing a non-problem. What's going to incentivize Verizon to build more capacity is that if they don't, people will switch to AT&T, or Sprint. Competition is what drives investment, not metered pricing.


From: Chris Ryland (Mar 26 2012, at 11:39)

I agree that it's inevitable. How do we pay for electricity or water or any other limited quantity? We pay by the meter.

Why should data be any different?

As long as it's fairly priced by the byte (MB or GB).


From: Gordon Haff (Mar 26 2012, at 11:47)

I tend to agree usage-based pricing is inevitable--at least for a window of time. As higher speed mobile networks come online, they'll increasingly be an alternative to using wired access and then usage will REALLY explode. I don't see any way other than having pricing schemes that encourage sensible behaviors--such as using WiFi when available.

The issue is that the telcos have something of a track record of generally outrageous pricing for metered overages of standard monthly plans. It's also true, as Nathan says, that people generally prefer not to think about a meter. But the fact is that we deal with lots of metered pricing in our lives every day.


From: yachris (Mar 26 2012, at 12:10)

Checkout ting.com -- gee, pretty much exactly what you're talking about. I'm in!


From: Buck (Mar 26 2012, at 12:29)

rape ≠ paying a lot for using a phone

Such comparisons trivialise rape. Uncool!

Otherwise I was into your post, though.


From: ben (Mar 26 2012, at 13:33)

From my perspective, the salient issues are these:

— Wireless service is cartelized in the States and Canada and a bunch of other places, but given the barriers to entry and the sunk costs, I don’t see how that can be changed without imposing more disincentives on operators.

— Meanwhile, rent-seeking behavior is rampant. I rely on a landline for which I pay 50USD p/m. To get the equivalent level of service on a cellie without data, I would need to spend $100+. A 100% premium for the sake of mobility and the big heap of small-shinies that come with the phone? Really?

— Most of the data is slurped by a fraction of the customers. My ISP started metering in ’05, and one of their engineers told me over beers that they’d established that 80% of their net throughput was being consumed by 5% of their users. I have a hard time believing that things are any different vis a vis cell networks. No doubt VoD has flattened that curve a bit, but a 15:70 ratio would not surprise me a bit.

— As a customer, the one thing I absolutely DO NOT WANT is to get nickel-dimed. But that’s EXACTLY what network providers are doing, and thus why I continue to rely on a landline.

— For all the nerdgasming about mobile and responsive design &c. &c., the fact remains that in the real world of sitedev bandwidth conservation appears to be a dying art, driven and practiced by content providers solely for their own reasons. Dinging customers who don’t really grasp that issue seems unfair to me.

(Related and rather uncommon shameless self-promotion: I’m really good at lightweight site design and implementation. If that matters to you or your business, let me know.)

(Related 2x: I lost count long ago of the number of times I’ve choked back the desire to take a semi-clueless designer behind the proverbial woodshed for sticking me with a design that disregards throughput concerns altogether.)

If I were king, I would insist that network providers:

— Introduce data service tiers that cover many more sigmas;

— Reframe overage charges as a premium on the contract/pre-pay price for those same levels of data service, in lieu of the egregious per-MB b.s. customers are currently forced to deal with;

— Ship apps that fetch Content-Length values BEFORE making a proper GET request, alert customers BEFORE they visit a plan-hogging site, and perhaps even use a contracted search API to suggest alternatives;


— Submit to common-carrier status for their wireless networks.


From: Bill Seitz (Mar 26 2012, at 14:12)

Other have argued against your pricing structure well enough. I'm a fan of the "fixed-price cap that covers 80-95% of users" model, myself.

But the idea of paying commissions to app makers horrifies me. I agree it would be smart for the telcos to do, but I'd be really nervous about any app-maker who took that bribe.


From: Eric Mill (Mar 28 2012, at 07:36)

I agree that the current structure incentives networks to minimize customers' data usage. But there are two things that I think outweigh this argument.

One is that this disproportionately affects the poor, and would seriously suppress their usage of the Internet and limit how their voices are represented online. Even with reasonable, even cheap, pricing, when you're living without much of a buffer, you look for ways to save where you can, and that will have a serious impact on how you make use of your network. People can stretch and accomodate a flat rate into a minimal budget, but it's much different when every action has a cost associated.

The other is that our collective behavior online seriously changes seriously fast. It's plausible to me that each individual could be using 100x the bandwidth they do now in 10 years, and I do not trust the telco oligopoly to drop their pricing 100-fold in the same time frame.

So if I had to choose, I would rather have my ISP incentivized to minimize my data usage, than have the masses each incentivized to minimize their own data usage. The latter is a much more terrifying vision to me.


author · Dad · software · colophon · rights
picture of the day
March 25, 2012
· Business (106 fragments)
· · Internet (99 more)
· · Mobile (20 more)

By .

I am an employee
of Amazon.com, but
the opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.