Yesterday the news came out that ad network AdMob had acquired the assets of "mediation layer"/aggregator AdWhirl. This likely sets the stage for an exchange play (in earnest) by AdMob, analogous to what Yahoo has with RightMedia on the PC.
AdWhirl enabled iPhone developers to sign up for ad units/inventory from a wide range of ad networks, to improve fill rates -- it was promising 100% -- and optimize CPM/CPC revenue in addition. At one point AdMob had said it was no longer going to work with AdWhirl but later decided not to go through with that move. Now the former has bought AdWhirl.
Deal terms where not disclosed. AdWhirl had reported $1 million in funding and so I'm guessing that the purchase price was south of $10 million (perhaps several million shy). Reportedly AdWhirl will now expand to encompass other apps platforms, including Android, BlackBerry and so on.
In expressing its objections to AdWhirl's methodology at the time, AdMob VP Ali Diab said in a blog post:
What we have also found over the last few months is that mediation layers significantly impair AdMob’s ability to optimize the selection of ads for the apps that choose to use them, by obstructing our view of these applications’ traffic profiles. These traffic profiles are a key input parameter that we use at AdMob to select the right ad for the right app at the right time. By working directly with the publisher, AdMob is able to generate a much more accurate profile of the traffic, in terms of both volume and timing, generated by a specific iPhone app, which will enable us to optimize the revenue potential that we can deliver for the app.
Now AdMob has embraced the "mediation" concept, perhaps shrewdly so. AdWhirl can now be the basis of a broader marketplace or exchange that will reach multiple smartphone platforms. AdMob thus gets a kind of "hedge" if the mobile ad market should start to heavily favor exchanges vs. networks. Other networks, such as Quattro, have been experimenting with delivering third party ads along with their own to publishers.
AdMob CEO Omar Hamoui provided the following statement to TechCrunch:
Our open source solution will be both for the client and the server. We anticipate there will be many independent servers run by developers, and possibly our competitors as well. Once we release the code into the community, we think it will be adopted widely.
Most developers use mediation layers for percentage based inventory allocation. This is not something that you can game or manipulate. The mediation component either fulfills the percentage allocation or not. We are committed to making this solution as open as it needs to be to make everybody comfortable. We also expect the market will hold us accountable to this course of action.
The AdWhirl team has visited our offices over the past several days as we have worked through this deal. However any claims that we have historically been sharing data with AdWhirl or manipulating how it works for our benefit is completely false.
Since the news broke we’ve been talking to developers and they agree that the key is an open and transparent solution. We expect to be held to that.
The article in which that quote appeared asked the question whether AdMob would maintain AdWhirl as a "neutral" exchange or whether it was going to favor its own advertising. Commenting on the deal, along those lines, Greystripe CEO Michael Chang said the following about a potential "conflict of interest" for AdMob:
“We believe that it is a conflict of interest for an ad network to own a mediation company. A mediation company needs to be a separate entity to do its job of unbiased optimizing of a number of ad networks.”
If AdWhirl does remain neutral and open as AdMob CEO Omar Hamoui promises it should continue to thrive. If it becomes biased toward AdMob's own inventory or in some other way, new "meditors" will rise up to replace it.
That number, $2.4 billion, represents AdMob's estimate of how much users are roughly spending on an annualized basis on paid apps from the iTunes store. These figures are part of AdMob's July, 2009 mobile metrics report. Much of the report is in turn based on a mobile survey of 1,000 users (see additional data). Some of the top-level findings are as follows (emphasis added):
Based on these figures AdMob extrapolates and estimates the value of the iPhone apps market and that of the Android Market for paid apps:
Here are some of the charts accompanying the report:
And among smartphones on its network, AdMob reports the following mobile Web activity (US):
Compare with last month's data:
Compare July global share:
Related: Earlier this month we reported on earlier data from AdMob on app downloads and user behavior.
Update: Developers question the figures/math here.
Millennial Media has released its July "Scorecard for Mobile Advertising Reach and Targeting." The company cites Nielsen estimates that the US mobile Internet now has 61.1 million users (this is probably still a bit low). On the basis of that figure, Millennial claims 75% reach on the mobile Internet.
Here are selected bits of data from the report:
Here are several charts from the report:
Compare the top pie chart above with the market share of the top four US carriers and MetroPCS, according to self-reported (Q2) subscriber figures:
One must look at all the reports (AdMob, Opera, Millennial and so on) because each presents a slightly different picture of the mobile Internet, device penetration and user behavior.
In June JumpTap was issued what appears to be a potentially sweeping mobile advertising patent. Another patent was just issued (no. 7,577,665) to JumpTap and is discussed in the press release out this morning. The one-line description of the thrust of what is being patented is "user characteristic influenced search results."
Here's the important language regarding when the patent would be implicated:
(ii) receiving from the carrier information relating to a plurality of user characteristics derived by the carrier, wherein the plurality of user characteristics are at least:
(a) two or more demographics associated with the user, wherein the demographics are obtained from the billing system of the carrier; and
(b) shopping habits of the user as recorded through use of the mobile communication facility, wherein the shopping habits include views of or purchases of goods or services; and
(c) duration of on-line interactions by the user from using the mobile communication facility; and
(d) usage patterns of the mobile communication facility including past mobile communication facility transactions comprising click-throughs; and
(e) previous search queries entered by the user via the mobile communication facility . . .
This language appears to require data being passed from the carrier to the engine. I'll speculate then that it wouldn't apply to "off deck" search or circumstances where the engine isn't getting information from the carrier. So, for example, I could argue it wouldn't apply to Google search on smartphones but it might well apply in the following hypothetical scenario: Bing search on Verizon handsets where Verizon user data is used to influence what ads to show in search results.
As with all patents, it's typically not until a court gets to rule on the meaning of the language that we get a sense of what this truly means. But JumpTap now has two broad mobile search advertising-related patents.
Based on a recent survey of iPhone users within the AdMob network, the company found the following attitudes and behavior regarding apps discovery and downloads (I don't know the sample size). I'm going to assume that these results are fairly representative of the larger iPhone user population.
Most people are downloading between 1-6 apps per month. They're using 4-6 apps on a weekly basis, but a substantial number are using between 7-10. The inference I draw from this is users have limited attention. While there may be 30 or 40 (or more) apps on users' iPhones, they're only going to use a number that can fit on the home screen.
What the slide above shows is that the top methods for discovering apps are:
The final two slides show that most respondents have downloaded branded apps and that users (60%) are favorably disposed toward them. Indeed, apps are proving to be a very effective tool for brands. We can expect more and more growth among branded apps.
AdAge features a piece that advises mobile marketers to start looking beyond the iPhone. Yes, that's prudent advice as a general matter; the iPhone represents a tiny slice of the handset market. However, if you're talking apps then it's all about the iPhone -- at least right now. The app experience is less significant on Android; it will grow over time and as more Android devices enter the market it will become a potentially important apps platform.
Rhomobile and Appcelerator (and others) have "write once" native apps development platforms. This scenario will become more common, allowing publishers and developers to more efficiently build apps across the major smartphone platforms. Right now, however, among smartphones it's really mostly about the iPhone:
According to Nielsen (pre 3Gs), iPhone users:
As the AdAge article points out, if you want reach beyond the iPhone, you can and should think about the mobile Internet. Google itself is increasingly banking on the browser as its "cross-platform" strategy. However if you need an app to create the optimal user experience then it's: iPhone, BlackBerry, Android, Windows Mobile and Symbian (outside the US) in order of potential reach.
Related: After iPhone exclusivity ends, AT&T is looking to a "plethora" of devices to constitute its "second act." AT&T reportedly envisions a broad range of connected devices with related consumer subscriptions to the network.
PaidContent has used Nielsen data (from an article in Mobile Marketer) on unique visitors to compile a list of the largest mobile ad networks. Here's the list according to the site (based on monthly unique reach in the US):
I invite the companies on the list (or comScore) to dispute it and provide their numbers.
AdMob says this morning that it served its "100 billionth ad." The company has been around since 2006 served its "billionth" ad in 2007. (McDonald's comparisons seem inevitable here.)
Here are some data cited in the blog post announcing the 100 billionth ad:
Meanwhile, AdMob has decided, at least temporarily, to not discontinue support for "ad mediators" AdWhirl and TapJoy:
In order to ensure that we are focused on doing what is best for our publishers, we have decided to temporarily delay the discontinuation of support for mediation layers until we can introduce a comprehensive plan that meets our publishers’ needs while also continuing to provide advertisers and end users with the best possible experience on mobile.
It took quite awhile for "ad exchanges" to arise online. Years after the proliferation of ad networks ad exchanges came along to aggregate inventory and bring more "liquidity" to the market. They also aim to overcome market fragmentation. Indeed, there are more than 400 online ad networks and, depending on how one defines exchange, 5-10 online ad exchanges.
A parallel trend is taking off in mobile, even before mobile ad networks have reached anything approaching maturity. Today mobile ad exchange Mobclix announced that it had developed the largest mobile ad exchange:
Through partnerships with over 20 leading ad networks, Mobclix provides the largest community of developers, advertisers, agencies and ad networks on one common platform for consolidated relationship management. As marketers leverage the Mobclix platform for precise audience targeting based on geographic, demographic and behavioral insights, developers use its analytics and advertising expertise to gain access to multiple networks in order to secure 100 percent fill rates and the highest eCPMs.
Mobile ad networks (partial list):
Mobile "exchanges," "mediators" and/or marketplaces (partial list):
Also some of the "networks" in the first list are starting to incorporate quasi-exchange features and third party inventory.
Pandora, the top Internet radio brand -- really the YouTube of Internet radio -- and the top free app for the iPhone in 2008 has struck an ad sales deal with Clear Channel, according to AdAge:
Clear Channel's radio ad sales rep firm, Katz Media Group, will start selling display and audio ads for Pandora through its Katz 360 digital sales group
Pandora comes to Katz 360 with a fledgling display-ad business that just introduced audio inventory in late 2008, yielding a total $18 million in 2008, or 95% of the company's revenue. Doug Stern, Pandora's director-audio sales, expected that total to double by the end of 2009 before the Katz 360 deal, and is now even more optimistic that Pandora will fit into advertisers' mind-set as often as it does listeners.
To date, Pandora has been using Google on its iPhone app to monetize inventory with AdSense banners. Apparently there are occasional audio ads that appear on Pandora, but I've never encountered one.
I'm a daily (and heavy) user of Pandora but, as I said to founder Tim Westergren at the EconSM show earlier this year: the minute I start hearing audio ads interrupting the music I'm gone.
Pandora has spoken in the past about a subscription offering (for its heaviest users). Audio ads could be part of a dual strategy to monetize inventory but also to drive those (like me) who listen to Pandora, in part because there are no ads, to a subscription offering.
We'll see how it plays out but there's danger in injecting conventional audio advertising into Pandora and alienating its audience -- or at least me.
Pandora, which owes its success in large measure to its iPhone app, has now become a clear (channel) acquisition target.
Mobile video ad platform and network Transpera put out a release today showing dramatic growth of video consumption on partner MSNBC's mobile properties:
Since first introducing video into the msnbc.com mobile Web product in March, monthly video views across the full suite of msnbc.com mobile product offerings has increased 207% - that’s an impressive 3x more video views in just over 4 months.
Further, according to msnbc.com, 60% of its mobile users are the most sophisticated class of smartphone user: they are savvy, active, and shop online.
InsightExpress performed the research on mobile video users for Transpera. Among "mobile users who watch mobile video at least once a week," the company found the following:
Source: Charts and data InsightExpress
Mobile ad network and search vendor JumpTap now says that its "premium ad network" reaches 42% of the US mobile internet audience, or 22 million people*, citing Nielsen figures. The premium network includes:
go2 Media, Hearst Magazines digital media, LatCel, Mobclix, MoFuse, and Weather Underground to an already impressive list of top tier properties. The network currently serves ads for hundreds of mobile sites, carriers and application developers including AT&T, Alltel, Ask.com, E! Online, Kargo (includes MotorTrend, Tiger Beat, Shape, US Magazine, Star Magazine, & BlackBook), LimeLife, MSNBC, and Shazam.
The company put out a release this morning that highlights its "momentum" in the market. I spoke briefly Friday to Paran Johar, JumpTap's CMO. He told me, among other things, that he's seeing "increasing mobile budgets, increasing seriousness" among advertisers.
We discussed a range of industry metrics and he told me that the JumpTap's advertisers break out 60-40 in favor of CPM vs CPC. This is similar to what AdMob has indicated.
Johar asserts that JumpTap is distinguished from other mobile networks by its carrier data targeting. In the US, for on-deck advertising with AT&T, US Cellular, Virgin Mobile and Boost (both Sprint properties), a range of carrier data are passed so that they can be leveraged for more relevant ad serving.
JumpTap started as a white label search provider, seeking to counter Google and Yahoo! with carriers. As Google and Yahoo! (and Microsoft) started to do more direct carrier deals, the company began to develop its ad network and that part of the business now represents the future.
JumpTap and Medio Systems used to be almost indistinguishable in terms of their claims and presentation of their capabilities to the market. However I haven't heard anything from Medio in quite some time. JumpTap appears to be going strong.
*Nielsen assumes a total US mobile Internet population of approximately 53 million. We believe the numbers are larger than that, although they decline if frequency of usage becomes a critical variable.
Skyhook Wireless has released a white paper entitled "Developer’s Guide to In-Application Advertising: How developers today can make money off apps" (.pdf). As the title indicates, it's aimed at mobile app developers or would-be mobile developers. The document offers a range of "how to" information and advice, including best practices.
It's a kind of crash course on mobile advertising and the mobile ecosystem for those unfamilar with the wonderful world of apps or how to make money with them. In addition, there are also interesting bits of data sprinkled throughout, from Skyhook's recent survey of mobile app developers. For example, location and demographic targeting appear to be the most desired capabilities or qualities among developers:
At the end, the report also features a list of vendors: ad networks, analytics providers and "ad enablers."
This morning's conference call with Yahoo! CEO Carol Bartz and Microsoft CEO Steve Ballmer covered a broad range of topics related to the Microsoft-Yahoo! search deal. Here are some excerpts from the call transcript related to mobile:
We want to invest in what is really important to our future success, including winning audience properties, display advertising capabilities, and mobile experiences . . .
I'm excited that this agreement provides us with the focus and the resources to invest in our audience properties, display advertising leadership, and mobile experiences . . .
We have the option of using the Microsoft technology for the mobile Web experience. It's not exclusive as it is on the PC. But we certainly have the benefit of it, and we will start, in fact, exercising that right to do it.
So, the only difference is it's not exclusive. If somewhere down the road we wanted to switch, we could, but there's no intent in that arena.
I think all of us would say we don't know what we don't know about all of the scenarios that we're going to try to invest in, in the mobile case. If -- and Carol can speak for herself, but it won't make sense to do a whole separate crawl of the Internet for Web sites to do mobile search, and yet what that really looks like, even whether the ad model for mobile search looks like the ad model frankly for PC search, I think that's an interesting question, and this gives Yahoo! flexibility to consider that broadly.
think the thing is for us is what we're very interested in, just like by the way with PC-based, we're very interested on doubling down on the mobile experience to integrate search as part of that, to integrate our content such as our normal finance, news, sports, homepage, that sort of thing. So, being able to have an integrated search is important, and it also frees us up to, as I said, really invest in the other areas of the mobile experience.
So, again it's a partnership that is very supportive, and allows Yahoo! to do what it does well, and that's really be the center of information, entertainment, friends, family, activity, and that sort of thing, both the desktop and the mobile experience we're looking for.
What this means is that Yahoo!, although not required, will be using Bing algorithmic search index for mobile. It makes sense from an economic perspective. Part of the reason Yahoo! is doing this deal is to outsource technology development and reduce costs. It would thus make no sense to invest in search for mobile and not on the PC.
Mobile search monetization and how that will go is a little more unclear. One version of the story is: direct overlap between PC and mobile. But that's been more Google's approach. Yahoo! has treated PC and mobile search separately. So I'm unsure how this will play out. In the broader relationship Yahoo! will sell "premium search" (complex campagins to large advertisers) while Microsoft will handle the self-service aspect of the business. Yahoo! will thus manage the PC + mobile search and display ad campaigns for large advertisers.
Separately, there are open questions about how differentiated Yahoo! Search and Bing will be on the PC side, but they will certainly be very differentiated on the mobile side (compare m.bing.com with new.m.yahoo.com) despite the fact that they will use the same index.
Microsoft and Yahoo! announced their long-awaited search deal this morning. Here's the press release which includes videos from both CEOs. Here's a site set up to explain the deal and here are the terms laid out by the release:
According to remarks made on the conference call, Yahoo! doesn't have to use Bing/MSFT's technology in mobile. There's more flexibility there for Yahoo! to maintain a separate search technology or work with other partners if it chose to do so. However it sounded like Yahoo! would be using the MSFT/Bing technology there as well. Part of the "flexibility" on the mobile side comes from uncertainty regarding how mobile advertising will play out.
Here's a paraphrase of remarks made by the two CEOs regarding mobile:
Bartz: We have option of using MSFT technology for the mobile experience. It’s not exclusive as on the PC. If somewhere down the road we want to switch we could.We’re very interested in doubling down on the mobile experience. Having an integrated search is important.
Baller: We don’t know all the scenarios involving mobile search. This gives Yahoo flexibility on the mobile side.
The two companies on the PC side are creating a single paid search market. There's more gray in the mobile execution. However I suspect that it will also play out the same way in mobile: a combined market for mobile paid search. On the PC side that means the reach of a combined Microsoft-Yahoo is approaching 30% market share; it will be something comparable on the mobile side.
Mobile answers service ChaCha released the results of a case study today based on an SMS campaign run on its network for the movie Transformers. It ran in June before the movie's theatrical release. We feature this case study in a white paper publishing later today on SMS marketing.
Insight Express performed the study. Here are some datapoints and takeaways:
Yahoo! recently launched its new, customizable homepage for the PC. Though in development for a year it represents a gutsy and postive move for the site. And in the past couple of days Yahoo! updated its mobile site, better aligning the new PC homepage with the mobile experience (though it's still not fully aligned).
The Yahoo! Mobile iPhone app was a dramatic improvement over the old Yahoo! Go application when it launched in Q1 of this year. And the new mobile Web portal in some respects offers a better experience than even the iPhone app.
The new mobile Yahoo.com is somewhat more streamlined, simpler and cleaner than the earlier version of the mobile website. It has three principal sections: Today, My Favorites and All Sites. The Today section is programmed by Yahoo and is essentially a news site featuring news, display ads and video. My Favorites is a highly customizable area that includes mail, a feed reader and incorporates social network updates ("Social Pulse"). It also allows for simultaneous updating of Facebook and Twitter. The former Yahoo! mobile social networking tool oneConnect appears to be gone.
Finally All Sites is a useful collection of Yahoo! properties from Local and Messenger to Sports, Movies and Search. As an aside, the iPhone version of the site can tailor content (and ads) to user location because Safari now supports location (after the 3.0 software update).
The layout and overall user experience on m.yahoo.com keep improving and these most recent changes and upgrades make the new mobile version of Yahoo! probably the most useful single destination on the mobile Internet.
Opera has released its State of the Mobile Web report for June. The company said it crossed a milestone with 10 billion page views on the Opera Mini browser globally in the month.
Here are some the top-level data from this latest report on top sites and handsets:
Remember this reflects data from users who have downloaded the Opera browser and is not necessarily reflective of the mobile Internet as a whole. However it's likely to be directionally accurate.
The success that Opera has enjoyed on the BlackBerry is partly attributable to the fact that the BlackBerry mobile browser experience is quite poor at the moment.
Ad network AdMob issued its June 2009 Mobile Metrics Report this morning. It focuses on the geographic distribution of iPhone and iPod Touch owners (Apple said in its recent earnings report that there are 45 million globally.) Here are the top-level findings from the June report:
Here are some charts from the report:
All of this must be qualified by the fact that this data comes from AdMob's network, which is not identical to the mobile Internet as a whole. However it's a large network and directionally consistent with mobile Internet trends more broadly.
The distribution of iPhone/iPod Touch ownership and usage is interesting to be sure but the bombshell in here is the data that Android-based usage is now bigger than Windows Mobile, this despite only one device (effectively) in the market for Android. This underscores the competitive urgency for Microsoft to get WinMo 6.5 and 7 into the market.
Yahoo! today is unveiling a new more customizable homepage that allows users to integrate Yahoo! verticals with outside content from third party publishers, via a series of widgets or applications that appear on the left hand side of the page (see images below). They replace a static list of Yahoo! owned verticals and sites that currently reside in the left column.
As part of the new Yahoo! homepage, which has been in development for a year or so, Yahoo! will be revamping its mobile portal. If users are signed in they'll be able to access the same list of "favorites" they've set up on their homepage on mobile (image at bottom). This should make the mobile site, which has been customizable since relaunch early this year, even more user friendly.
Here's the new homepage:
Detail of left "favorites" column:
Yahoo! will also finally start autodecting devices and showing the mobile portal automatically to mobile users. If you go to Yahoo.com on the mobile Safari or Android browsers today you see a rendered version of the current Yahoo.com homepage rather than the mobile homepage, which requires users to go to new.m.yahoo.com.