BTW, NARIP asked me to link to this – I did a simple write-up on how to distribute music via mobile.

Where should I focus my mobile ad spend?
As expected, the panel answered “…test, test, test…”. With all the different combinations of ad networks and ad units etc, the easiest approach to see what is working is to test – $5K still continues to be the average “good” test budget.

Has the mobile web advertising market gotten more efficient?
In some ways, this was sort of a trick question. In an efficient market, you have less variance (ie mediation engines tend to focus more on ad unit types and placement on the page as opposed to the benefits of culling from multiple ad networks). Efficiency also means that CPMs have normalized meaning that the 50 cent eCPMs we see now in mobile may not ever get back to $1+ (or at least until the network is large enough that it can take advantage of a new parameter such as granular location). Obviously during recessionary times, a greater piece of the ad spend is in direct-response (ie performance-based) as opposed to brand advertising, so this doesn’t help either. So yes, the panelists did agree that the market was getting more efficient.

Is ad mediation necessary?
Not surprisingly, all of the panelists agreed that we need mediation but it ultimately comes down to testing. Mediation is a funny thing and in many ways reminds me of the SMS aggregator business. Yes, there are services like AdMarvel and TapJoy that are direct-to-consumer facing mobile ad mediation services but it’s not like every other mobile ad network isn’t pulling ads from other partners as well – they are just not exposing it as a D2C service. It’s like in the old days of SMS before everyone had “direct-connects” – you send an SMS to Verizon through MX Telecom and they then send it through their competitor, Verisign – funny how that works.

Does the mobile ad spend have a ceiling as a percentage of the total ad budget?
It was not in the panelists best interest to say yes but everyone agreed that there is a ceiling for digital and then a ceiling for mobile within digital. More interesting was the panelists agreement that mobile is now just a piece of selling digital as opposed to necessarily being called-out in the media buy. This is a remarkable state since this implies that the industry has matured and mobile has moved out of the traditional “innovation ad spend”!

Is there an iPhone bias when selling to the brands?
Mixed answers here – some agreed that when selling mobile, it implied iPhone. Others said, it wasn’t called-out by device. Regardless, a chunk of the mobile advertising industry has morphed into in-application advertising, essentially PPD (pay per download). The beauty of this, is that for the first time, the advertiser can calculate a real ROI (ie I buy an ad and does it result in a purchase of my app in the app store – this is much akin to the biggest ecommerce advertisers on the web measuring purchase conversion).

IAB Mobile and MMA convergence?
We hope so :) We definitely need fewer ad units to target – some brands are definitely already beginning to complain about how much they have to spend on creative.

Are we ready for behavioral advertising in mobile?
Unanimous consensus was no and not for a long time – the network is barely able to target on granular location, it’s going to take a long time before we have ad servers that can ingest behavioral data as well as networks that can consistently fill against this data.

Is PPC (pay per call) picking-up?
ATT Interactive (with Yellow Pages and Ingenio ownership) obviously said yes, but agreed that it’s still a long ways out. Solving the hyper-local problem is difficult, there is a reason, I get a Pizza Hut ad as opposed to the mom-and-pop shop. On a separate note, probably another blog topic, I’ve heard that PPC performance is varying significantly by carrier – not certain yet, but it seems bad coverage means worse performing PPC!

Is lead-gen on the horizon?
Everyone was in absolute agreement here – mobile Offerpal was on it’s way! Onthe contrary, video ad units was definitely not in the horizon. There were some tests in the past but there is just not enough interest for mobile video ads yet, mostly because the experience isn’t there yet.

Will the top 10 advertisers on the PC become the top 10 advertisers on mobile?
Not much response from the panel here, after some thought, some of them said yes. This answer validates a few assumptions. One, that mobile and PC ad networks are converging (ad units will of course remain different). Two, that the mobile will become as effective of a transaction platform as is the PC (ie credit card payments for Amazon products on mobile would have to be seamless). Three, the browsing behavior on mobile would ultimately converge with the PC – with better browsers, we are starting to see that kind of convergence where for example, Skyfire’s top 100 visited sites closely mimics the Alexa 100.

Will platform providers, specifically Android, introduce their own advertising APIs for in-application? How do you compete?
The panel absolutely agreed that this was coming and as expected, they said it’s about testing with as many partners as you can and seeing what performs for you the best. Interestingly, there is definitely a trend of carriers and/or OEMs introducing or partnering with ad networks and analytics companies (ie T-Mobile US has partnered with Flurry for analytics and 3 ad networks for mobile; Motorola has partnered with Amobee) – interesting to see how this all plays out.

Those were the interesting tidbits, feel free to ping me offline if you want more. BTW, the panel was composed of Ali Diab at Admob, Michael Rubin at ATT Interactive, Ragner Kruse at Smaato, Paul Cheng at Velti and David Katz at Yahoo.

FYI, if you are not familiar with Carnival of the Mobilists – it’s a network of bloggers who write about mobile and take turns each week hosting the best posts. It’s a great way to drive traffic but also a great to get a quick summary of the week.

Meanwhile, I’ve been working on power-metering my life w/ one of my favorite apps, Nokia Energy Profiler – more on that soon.

As you can imagine, an opt-in subscription is quite painful for a content provider since they can’t bank on things like breakage – the fancy term for “I forgot that I’m subscribed (and I keep paying them)”. Most mobile content companies bank on breakage – it’s a fundamental piece of their biz model which is fine as long as the customer (subscriber) knows what he’s subscribing for. Where this starts breaking-down is when you start seeing applications that use nefarious or mis-leading marketing techniques to get new customers such as the Facebook quiz I recently took, which required subscribing to a $9.99/month subscription to get your quiz results. If you didn’t notice the fine-print at the bottom, you would have been subscribed when you clicked “See my results”!

As a result, most mobile content companies are facing numerous class-action law suits and most US carriers are facing huge premium return rates – see PhonePayPlus for some of the adjudications in the UK. This is how it breaks-down (estimates):

AT&T: 20-25%
T-Mobile: 15-20%
Sprint: 7-12%
Verizon: 5-10%

20-25% of all PSMS subscriptions on ATT are returned – insane! Just to translate this, the average Level 1 customer service call is $10-15 (Level 2, which these probably never get to are approximately $40). The interesting ratio here is the potential revenue earned from subscription breakage versus the customer service call volume and it’s affect on customer retention. Also interesting is how the refund policies or inclination to offer a refund vary amongst the different operators. In any case, this acceptable is about to get flipped upside down with Sprint recently announcing that starting Oct, 09, any premium content provider who results in a 7%+ return rate for 3 consecutive months will have their PSMS shortcode pulled! Is Sprint ready to shut down the millions in breakage revenue? And the bigger Q is whether other carriers follow-suit – this will be absolutely detrimental to the mobile content industry in the US – the “ringtone” pain, has maybe just begun…

While taking the bus into the park, one of the other passengers was joking how his cellular coverage seems to extend further and further past the entrance of the park. Well, imagine a world 25-50 years from now, where to get out of coverage, you have to go to a “Mobile Reserve” where there are no cell towers installed and/or jammers are installed to intentionally kill coverage especially since solar base stations are starting to take-off. Being totally off the grid may become a thing of the past.

Well, as part of this trend, I think we are running the risk of repeating history. When J2ME was first released, a lot of the cool functionality was not available (ie access to the camera or file-system or bluetooth etc). Given the long but necessary process of the JCP to approve additional JSRs to support this added functionality, handset manufacturers jumped-the-gun and started releasing proprietary J2ME libraries to achieve certain functionality. I remember distinctly using the proprietary Motorola file system J2ME libraries and not JSR 75 which hadn’t yet been ratified or adopted. Instantly, J2ME became quite fragmented; given that I was working on a photo-upload application, we had to port to every J2ME device using different proprietary file-access libraries.

What’s interesting is that this same trend may repeat with the mobile web. As the browser becomes the app platform on the mobile device, developers are wanting to access native functionality on the phone (ie file access, the camera etc) via Javascript. Because the W3C MWI (Mobile Web Initiative) and the Bondi OMTP are still gathering support of the ecosystem and still defining standards, OEMs have begun defining proprietary Javascript to access native features of the phone. Palm is leading the charge here with WebOS and the Mojo SDK by defining custom JS to access native features of the Pre.

Going forward, I expect that most OEM browsers will take a similar approach potentially resulting in a fragmented web. As a developer, I may now have to port once again amongst different platforms.

So I’m pulling up a slide here – I reconstructed it pretty quickly, it was originally shown in an episode of South Park:

You got to love it! Is this the slide for the music Web 2.0 industry:

Ha!

1. iPhone
2. Blackberry
3. BREW
4. J2ME
5. Android

One of the game publishers said iPhone was doing 7X Blackberry which was their #2. Note, in the BB App World, apps are often listed upwards of $20. Also interesting to see Android at the bottom of the list but I fully expect that to climb especially as more Android based devices are launched later this year assuming we don’t have massive app store fragmentation on the platform. That being said, I also hear Android has the highest return rate, near 75% for some developers given how seamless the process is to return an application (clicking a button rather than having to make a phone call).

BREW is still a big money-maker and several X more than J2ME even though it’s basically a Verizon-only proposition – this story hasn’t changed from 01. The contrarian move would be to double-down on BREW given that so many developers are focused on the high-end. Also interesting is that S60 and Windows Mobile are not on the map but again, I fully expect that to change once the Ovi store is in full-force and the Windows Marketplace is up and running.

This simple table actually represents quite a bit of learnings. The most controversial row is volume – app stores have definitely made it a lot easier to achieve millions of installs but most developers still get stuck in the 100s of K of installs.

Most interesting is with revenue, many of my friends at many of the large mobile game publishers are struggling with the iPhone and other “democratized” app stores. The problem is, with carriers, they have guaranteed revenue since they have guaranteed placement – with the app store, there is no guarantee, they could spend 3 months of dev time, money on licenses and not achieve an ROI. This high-risk distribution avenue doesn’t immediately suit well for the traditional “tight margin” mobile game publishers.

And as you would expect, working with operators takes a lot longer (and a lot more money) than most people imagine. Most of this is because of the painful requirement of having to port to all their phones – I fully expect this to change since there has been increased focus on the high-end.

One final note, sort of the VC pain, how many mobile companies do you know that make millions a year that do not require operator partnerships (very few) – that’s why I’m an industry fan of the lifestyle business.