Notes on the next Apple TVs

Speculation about Apple Televisions has become something of a joke. I wrote a 15 page report for Enders Analysis a year ago saying why I didn't think one would exist (short version - very low margins, little scope for higher prices, content irrevocably locked up by platform owners), and it occurs to me that I could change the dates and put it out again this year. Meanwhile, the current Apple TV remains a 'hobby'.

However, I've also been pondering acronyms and technologies. There are a couple that seem relevant.

• HDMI power.  A device plugged into an HDMI port can draw a very small amount of power from the TV. This isn't quite enough to run , say, a dongle made using off-the-shelf parts, which is why the various Chinese Android-based HDMI dongles now on the market need a (very un-Apple) USB tail, which plugs into the TV's USB port and draws power that way (see an example here). However, if you design your own chips (as Apple now does), you might now be able to get the budget low enough, if you limited what the dongle did. 
• Wifi Direct. This is the really interesting part, and it's what drives several new parts of iOS7, including AirDrop. It means (simplifying hugely) that two wifi devices can find and talk to each without needing a base station
• Another part of iOS7's networking API set allows for peer-to-peer local wireless discovery by apps - so your game can look for another device in the same room running the same game. So get your friends around, start the game and stick the action on the TV (or squirt it onto the TV, as Steve Ballmer might say) 
• And then there are the new APIs in iOS 7 for hardware game controllers. What should we make of those?

Now, consider a new 'Apple TV' that's an HDMI dongle, powered by HDMI so you can just plug it in. It uses wifi direct to advertise itself such that as soon as you plug it into the back of your TV, your iPhone pops up and tells you it's available. No need to enter a wifi key, no need to fiddle around with an keyboard on your TV screen. Maybe it uses peer-to-peer from then on, or maybe your iPhone tells it the wifi key. And to get the cost down, all it does is Airplay (which Apple has promoted to front and centre of iOS 7 by putting it on the control screen panel). So for $50, any TV becomes a remote display for any iOS device.

There are a bunch of problems with this. HDMI power is somewhat inconsistently available. Airplay (since it encodes the video feed into h264) puts a lag into the display that may be problematic for games. Airplay by itself may not be enough of a feature (but then, maybe it's possible to implement the current Apple TV feature set in this envelope). And of course developers need to support it and content owners need to refrain from blocking it. 

But extending the iOS ecosystem onto a TV for $50 seems to me a much more compelling proposition than trying to get people to break into the 5-7 years replacement cycle to buy a new Apple TV, and $50 is a much more compelling impulse/upsell price than the current $100. Equally, the touch screened device in your hand seems like a better user interface than any remote control. Meanwhile, TVs are a 1% margin business and there's very little Apple can bring to justify a much higher price (or rather, a higher price premium than just buying the existing Apple TV). The advantage, of course, is the strengthening of the broader iOS ecosystem - driving more sales of $650 phones at a 45% gross margin.