Mobile advertising. Remixed.

Posts Tagged ‘measurement’

We must liberate mobile metrics to make advertising measurable

Monday, December 28th, 2009

GoMo News today posted an article I wrote calling for the industry to be more open with mobile metrics.  Here it is again:

Mobile advertisers face a big problem: Fragmentation.  Not of technology, but of information.

Say you’re one of the majority of mobile advertisers wishing to drive traffic to their mobile internet site or mobile app by advertising on other mobile sites and apps.  How do you measure the effectiveness of your campaigns?  The agencies and ad networks alone can’t tell you. A high exposure rate is meaningless unless you get clicks, and yet a high click through rate is not always a good thing.  Your site analytics alone can’t tell you because you don’t know if those high conversion rates came at a prohibitive price.

So at a minimum you need to track spend on the ads and conversions on your site / app for each and every campaign you run.  And this is where fragmentation laughs in your face and turns your marketing team into copy-pasting slaves.  The problem is threefold:

-        duplication of reporting from ad networks and agencies

-        duplication of conversion tracking on your site/app

-        stitching it together

Firstly, you’ve got to aggregate ad performance data in a meaningful way from multiple ad networks and agencies.  Some of the more forward-thinking ones are realising this and planning an API to give their clients direct access to their data (I wrote about this recently: Ad networks: emancipate your clients data!).  But precious few of them are doing this and it will probably be years before they all do, by which time many will have been acquired or sidelined.  Meantime, you have to navigate your way through many different, sometimes complex, sometimes downright awful reporting interfaces to manually extract your data.

Secondly, for those ad networks that offer ‘conversion tracking’ features, you need to integrate their tracking code into your site or app in order to see full CPA performance data for each ad running on their network.  This sounds like a great idea, until you realise how many separate tracking codes you’re going to need to integrate to cover all the ad networks.  Once you’ve got more than a couple of these ‘trackers’ running, response times are going to suffer.  Don’t worry though, hardly any offer this yet anyway.  You could use an analytics provider to track conversions instead, but you’ll need separate tracker codes for your mobile sites (Admob, Bango and Omniture are the most well-known with Google only recently releasing a mobile analytics service) and for your apps (the recent merger of Flurry and Pinch Media is looking a good bet here but still does not cover all the device base).

These two issues compound the problem of how to stitch it all together.  Obviously there’s the sheer effort of just having to collect and capture common-denomination, comparable data from all these sources into one place and then match it all up.  Then there are time zone and currency differences.  Nearly all ad networks and some analytics providers think you live just up the road from them.  How are you supposed to compare ad spend in the US delivered by an ad network headquartered in India with conversion data from your analytics provider in the UK?

For me, there is one clear and relatively simple solution to this problem:  Ad networks and analytics providers give your clients their data in a time zone independent format and liberate it with an API. That way, they can easily combine and compare ad performance end to end for each and every one of their campaigns.  The advertisers and ad agencies will be the final deciders on this, and they will choose those companies in the ecosystem that are able to give them the data they need whenever and however they want it.

Clarity and control will drive growth of mobile advertising

Thursday, November 19th, 2009

There is currently an imbalance in supply and demand in the mobile advertising market. There is no shortage of mobile websites, portals and apps out there to advertise on, but there are not enough big spending advertisers to fill the inventory with high-value ads. The majority of publishers are the long tail, using blind ad networks (such as AdMob) to sell their inventory on a CPC basis. The kind of advertisers using these ad networks have to be comfortable with a loss of control over where and how their ads are displayed. Mobile content companies and small-time advertisers abound, with the ringtone and adult entertainment sector being notably well represented. Much of this advertising is taking limited or no advantage of the benefits the mobile channel offers such as location awareness or demographic / behavioural targeting.

There are big brands advertising on mobile, but in a different way and in small doses; they are often working with agencies and ad networks that can tell them which sites their ads will be displayed on and the user demographics of those sites because it’s important to control their brand image. They are also advertising on the iPhone where they can find more well-off consumers, despite in some cases losing control over which apps their ads will appear in.

For mobile advertising to really take off, the big brands and non-mobile companies need to spend more money on it, which will be driven by:

  • More control over the quality of their ad campaigns, driven largely by greater smartphone adoption and data speeds
  • Greater confidence in where and how their ads are displayed so they can control their brand image
  • More accurate measurement of ads, which depends on agencies and ad networks providing improved ad performance data in a way that can be aggregated and compared, and also on better analytics of what happens on the advertiser’s properties in order to give an accurate end to end picture of short and long term ad performance and ROI
  • Greater cross-channel integration of the conception, delivery and measurement of ad campaigns across multiple media in addition to mobile.

I think that the technical challenge of ad delivery to a fragmented mobile device base is largely on the way to being solved. The measurement of unique users and understanding who they in greater depth are has not … but this is an envisioned advantage of mobile over traditional media, and I do not see it as a brake for growth right now. And since most of the privacy issues are again associated with the new envisioned capabilities of mobile, they are in my opinion simply (though not straightforwardly) something that needs to be managed as the industry continues to innovate.

Image attributed to: http://www.flickr.com/photos/celinet/ / CC BY-SA 2.0

Should we kill the CPM on mobile?

Friday, September 25th, 2009

An interesting post today on Techcrunch suggests that we should kill the CPM to save the online advertising industry.  As the argument goes, by using CPM as a prime measurement of campaign ROI advertisers and publishers are missing the real goals of relevance, engagement and conversions, and consumers suffer as a result.

I agree that CPM is not a good primary measurement of advertising success, even for brand awareness only campaigns.  However do we really have to ‘kill it’ to make people stop using it that way?  I argue that we shouldn’t, and certainly not on mobile.  Here’s why:

  • CPM is still a viable and useful measure, if not the primary one.  As a publisher selling ad space on your site, you want to maximise revenue for that space.  Each page view needs to deliver great content and a great monetisation opportunity if that’s your business model.  So you want to track if those extra pages of content you add are bringing in the money or not.  As a ‘brand awareness’ advertiser where your goal is to create long term loyalty to your brand, CPM is a good measure to compare the cost of advertising across potential publishers you may choose for a given campaign …. assuming you also know how to factor in the relative quality of each publishers’ sites and the relative value of each publishers’ audiences.
  • Market dynamics will rectify the situation anyway.  If you sell only on CPM, or you buy only on CPM, without looking at other measures of engagement and action, then you will lose out.  Poor content sites with lots of ads will command lower and lower CPMs because advertisers value your disenchanted audience less.  Poorly targeted, cheap CPM campaigns will cost you more in the long run as an advertiser because they won’t have the desired effect.  Users will vote with their mouse/thumb and avoid sites with poor content and ad overload, and go to better quality sites where there are fewer, more relevant ads.
  • On mobile, the above applies, except even more.  You cannot fill a mobile webpage with more than 2 or maybe 3 ads maximum (for a long page) without really annoying the visitor. Their perception of your site degrades in proportion to the volume of advertising you fill it up with.  The quality of your mobile pages needs to be even more engaging because of longer download times and the small screen size.

It’s true that CPC and CPA are becoming more useful measures for ‘performance advertisers’ who want their ads to generate immediate action and response.  But CPM still has its place as a useful measure, and should not be so easily dismissed.