Note: This post originally appeared on Joe’s blog, and is re-posted with his permission.
Ask nearly anyone if they want ads in their mobile experience and most of them will respond with a stentorian “
no“. The late Steve Jobs even stated in 2010 that mobile ads “ suck“. Yet the dollar spend in mobile “ads” is going up exponentially and is predicted to soar over the next few years, from $3.3 billion in 2011 to a whopping $20.6 billion by 2015.
Yet therein lies the rub. If consumers are so turned off by mobile ads then how does mobile ad-based revenue increase over 600% in the next few years?
A simple response can be the sheer number of smartphones and tablets that will be in use by 2015:
1 billion and 250 million, respectively according to Transparency Market Research, will in lockstep increase the number of endpoints for ads. Yet given that we look at our phones on average 150 times a day, that revenue number seems rather low. But why? Because people don’t like ads!
However, people not liking ads is not the only problem. Attempting to map the dot-com web advertising
pricing models onto mobile ads poses technological and user experience-based problems as well.
Mobile ad pricing – CPC, CPA or CPM?
Typically there are three common pricing models for desktop web-based advertisements. Unfortunately, each one has its own set of issues in the mobile space.
Cost-per-click (CPC) in the mobile context poses big issues because, well, there is no clicking on a smartphone or tablet and even if you changed the metric to “cost-per-tap”, false positives would be quite high given the small screen real estate and how easy it would be to accidentally tap an ad (arguably a poor user experience). Morever, the King of click-based ad revenue, Google, recently stated on their conference call on April 12th, 2012 that “ Aggregate cost-per-click growth was down 12% and down 6% quarter-over-quarter.” These decreases are on the dot-com sites as well! I can’t imaging this boding well for a mobile CPC model.
CPA, or cost-per-acquisition (or conversion), is minimally viable at best. Think about it. You click on an ad for a new drill from a hardware retailer inside an iPhone app. Now, you are redirected outside of the app to the product detail page of the drill on Home Depot’s website (hopefully it is mobile-formatted!) where you then need to then enter your shipping and billing information alongside your credit card information. As you can imagine, this experience on a 4.3 inch screen and without a physical qwerty keyboard will lead to huge drop-offs in conversion.
CPM model, cost-per-mille or cost per thousands of times an ad is displayed (presents an impression), might be the most relevant, but posts its own set of problems. First, how is one to measure an “impression” in a mobile app? If I’m rapidly scrolling a list view of information (think a few hundred tweets for example) and an advertisement flies across the screen as one of those hundreds of list items, does that count as an impression? Are API calls to ad networks like Millennial Media or Greystripe considered the metric for measuring impressions?
Banner ads are somewhat effective for display yet they are a nuisance according to users. Is the risk of displaying the ad and disrupting the user experience worth the revenue gained? If you have a 4 or 5 star-rated app in the Appstore are you willing to risk lowering your rating to integrate ads?
web browsing context. A native application using native user interface components doesn’t have cookies whatsoever! Therefore, sophisticated tooling needs to be created so developers can make tracking and reporting of impression data inside the native app context a reality.
The solution? Ad-like objects and media buys
So how are publishers and app developers alike supposed to handle the aforementioned issues? My solution is to focus on the content of your app and mesh in ads, but not in the traditional sense, but in an ad-like object.
Think of ads as natural piece of content in the flow of an application, something that doesn’t take away from the native experience and actually
feels like part of it. This is much easier said than done, but is possible. In this blog post alone there is an ad and if you’ve read this far, you’ve already encountered it.
App developers and content publishers will have to work closely initially to get past the archaic way of serving digital ads (banner displays, interrupting videos, etc.) and focus more on tight and frictionless app integration. Imagine applications where ad networks plug in rich media and interactive content into areas of the application where the
developer decided ad-like objects’ content could live. Think of it almost as product placement inside of an app.
These locations inside of the app have to be determined
a priori by the developer while he or she is actually developing the application – not afterwards, which is currently the model for most of the mobile-focused ad networks. “Drop our SDK into your currently built app and watch the money flow in,” says many an ad network. This is certainly not the case for the greater share of applications.
Digital product placements or “ad-like objects” may or may not follow the traditional CPC, CPA or CPM models; instead, they should be treated initially as
– a flat, fixed fee for the placement and preference of content inside the application. This pricing model is not nearly as optimized as the CP-* models of the desktop web realm, but may allow for flexibility in not only the piloting of such unique programs, but actually pulling from a media buys media buy budget as opposed to a company’s digital advertising budget. For developers, this is certainly more attractive as media buying budgets are typically much larger than digital ad budgets.
In the future…
There is no doubt in my mind that someone, somewhere will figure out a transaction-based payment model for ad-like objects as media buys may be incredibly difficult to price. Why so difficult? Because an ad’s endpoint (the myriad of mobile and casual computing devices) continues to change and with that, the types of applications (and games) that are developed continue to evolve as well. As newer sensors are added to our smartphones, expect new types of applications and subsequently new types of ads.
Imagine chemical sensors that can detect the trace of alcohol on your breath. Sounds crazy now, but so did the idea streaming HD video on your phone. If you’ve had too much to drink, maybe an app you use frequently pipes in an ad-like object for a local taxi to pick you up at your current location because it recognizes your blood-alcohol level is above the legal limit of the State you are currently in and the ad-like object recognizes you are more than walking distance to your home. All of these data points can currently be determined by the GPS sensors on your phone and some web service calls to APIs.
Another concept that will enhance the ad-like experience will be heavily centered on ad-like objects’ ability to take advantage of the contextually relevant data associated with a smartphone and the user of said smartphone. Location is an obvious one, but what about meta-data that extends simply
from location? The weather, sporting events, social/music events, traffic, population density, colloquial idioms, political influences, etc. Now include past “click” (tap) history on ad-like objects and even a user’s social graph. The type of ad-like objects should be and could be highly relevant to the end user and more importantly not take away from the natural feeling of the app.