Most of you have already read the details of AT&T's pricing announcement of last week. There are some clear advantages and disadvantages of the new approach. On the plus side, AT&T has:

• Lowered the entry price for wireless data, thus opening up important new market segments and making data more affordable at the low end;
• Developed a more favorable price point ($14.95) to add additional devices/users for data services, which I think will be especially important for households and families; and
• Set the tone for greater awareness among users of how much data they are consuming, the undercurrent being that even with all these fantastic devices, wireless economics are different from fixed broadband economics. Those who are concerned about overage will think more proactively about using WiFi where available.
  

This is a good start, and is the opening salvo in what is sure to be a new wave of pricing activity in the industry over the coming months. But now that we're headed down this road, I do have some thoughts on what could be done differently. Here are some thoughts:

Plan Optimization. AT&T is trying to keep it simple, with its approach of adding 200 MB in $15 increments for those subscribing to the $14.95 plan. However, 200 MB plan users who go over their allotment are then paying $30 for 400 MB total, compared to $25 for 2 GB for those on that plan. The $15 for an extra 200 MB makes even less sense (and looks like gouging) when 2 GB plan users only have to pay $10 for an additional 1 GB.

AT&T would argue that those who consistently exceed the 200 MB limit should upgrade to the 2 GB plan. I'd like to see the option for what I call "plan optimization": go over the 200 MB limit, and you can choose to upgrade to the 2 GB plan for that month - which can then either continue permanently or revert back to the plan. This would give AT&T greater opportunity to upgrade users on a more proactive basis, rather than upset them over constant overage of the relatively paltry 200 MB increments.

Very heavy users will pay a lot more than they do today. 5GB of iPhone usage (an unlikely number) will cost $60 -- the same, ironically, as its DataConnect 5 GB broadband plan.

Buckets of Shared Megabytes. Wireless data pricing, whether flat rate or usage-based, is still principally based on a per-device, per-subscriber approach. But the average individual/household owns multiple devices, with varying data requirements and usage scenarios. Wireless operators in the U.S. have been hugely successful with family/group plans for voice. Why not take the same approach for data? As the master account holder, I would buy a "bucket" of MB (or likely GB), which could then be shared across multiple members and devices in my household. This would be great for additional household members, as well as for devices with different usage profiles. A smartphone is going to have a more persistent need to connect to the cellular network; but devices like the iPad, portable gaming device, or even a laptop might have more infrequent, occasional, cellular connectivity needs.

Allow Multiple Devices to Connect. As a corollary to the "bucket" plan approach, I don't believe users should have to make individual data plan decisions for each and every wireless-enabled device they own. A broadband subscription can be shared across devices, and an iTunes service can be shared across multiple PCs. Operators have gone "half-way" with cellular data, allowing other devices to draft off a primary device (or MiFi) using WiFi. I'd like to users be given the ability to authorize up to a certain number of devices that can share in a data plan. A side benefit of this is greater stickiness to a carrier's data service, since there would be less incentive to shop around for a lower, or prepaid plan from another operator for a device with less need for a persistent cellular connection.

What About Broadband? The plans AT&T announced last week were aimed at smartphones. Lens readers know I have long been advocating for lower, more flexible price points for mobile broadband from laptops. T-Mobile and Metro PCS have led the way with lower prices, and there are some good prepaid options (per day/week/month) from Virgin Mobile and Verizon Wireless. But with the delta shrinking between computing device and phone device, and products such as the iPad that are in-between, I believe having a separate, significantly premium-priced mobile broadband offering makes less sense.

AT&T's broadband plans now look downright wacky. 200 MB for iPhone users costs $15, but it costs $35 for laptop users. Huh?

Start Thinking Anew About Voice and Text. Messaging has been a cash cow. Text plans remain incremental to wireless data subscriptions in most cases. Lines are blurring, however, between messaging of various forms - voice, e-mail, text, IM, and social networks. Today's user accesses Facebook, Twitter, or LinkedIn via their data account, but must pay extra for SMS and MMS. There will be tension around this as users add multiple devices to the network, as smartphone adoption flows down to the younger, text-centric segment of the market, and as users/households start really examining their total monthly spend.

It is time to start thinking about how legacy messaging and voice plans are going to be "sunsetted", in a 4G, "everything is IP", multi messaging platform world. The subscriber relationship will be less about discrete services (voice-messaging-data) and more about a "connectivity" plan that is shared across multiple devices and users. This will require tough decisions around legacy - and highly profitable - plans.

Make Some Moves in the Enterprise. Enterprise pricing has not changed markedly over the past three years. While this segment remains higher margin, I do believe two moves are needed. First, mobile broadband pricing needs to be adjusted, and to be more in line with smartphone pricing, especially since tethering options have expanded. There should also be more prevalent plans that provide some pricing flexibility for occasional/infrequent users who represent the next phase of growth for the market.

Customer Education. I'll have more to say about this in a future Lens. But with the dawn of usage pricing, operators have an interesting challenge with regards to customer education. AT&T has developed the right "early warning" tools to protect consumers (and themselves) from bill shock. However, truly understanding data consumption is tricky. Sometimes it is obvious, such as downloading TV program from iTunes. But there's murky ground with respect to:

• Embedded content. It's one thing to go to YouTube...and another to play embedded content within an app, such as a video clip on CNN.
• Apps requiring persistent connectivity. A good example here is applications using location services, such as navigation. Also, it's not always clear when one is actually "connected" to the network. Does my RunKeeper app stay constantly connected, for example? And what about integrated messaging services that are constantly sending information to and from Facebook, Twitter, and the like?
• Over the air synchronization. More and more services are using over the air synchronization, for example managing e-mail, contacts, and calendar using the "cloud" such as Gmail and MobileMe. 
• Advertising. Ads to count against the data buckets.  It's not an issue right now given current consumption patterns - but with more sophisticated, multimedia ads being developed for mobile, we'll have to keep our eye on this.
  

The operators have developed some reasonably good tools: AT&T has usage scenarios and a pretty nifty "Usage calculator", which shows how moderate usage of anything multimedia gets you to 2 GB pretty fast. The more troublesome area to predict consumption is with apps requiring more persistent connectivity, as stated above, and the issue that the customer is not always in control of their data consumption, such as when they are sent large files, attachments, video ads, and so on.

Operators need to make these tools prominent, especially via collateral, at the "My Account" site, and at point of sale. I would also recommend adding more functionality to the "early warning system" For example, inform users of high-usage apps; warn them if a file they are about to download is above a certain size; or recommend for certain downloads that they use WiFi. At point of sale, or when signing up for data service, there can be an option for users to opt in to this early warning system, similar to what they do now for privacy settings.

It will be interesting to see what other operators do. I believe Sprint and T-Mobile will stick with aggressive "unlimited" plans, to maintain competitive advantage in the marketplace. Plus, their data networks are less taxed at the present time. For Verizon, which has intimated it will migrate to usage-based pricing, the question is whether it will take the "airline" approach of mimicking AT&T's plan or whether it might do something a little different. I'd bet on the latter.