Why click-through rates don?t make sense in mobile
A few months ago, I was sent a post-campaign analysis from a major mobile network. The deck reviewed the success of a film studio?s recent mobile display advertising initiative promoting the Blue-ray release of an Oscar-nominated film.
As I went through the slides, I had a sense of déjà vu: it was remarkably similar to presentations I would put together as an account manager working for an online marketing services provider ? five years ago.
If the first slide seemed reasonable reviewing the targeting and giving some basic impression and click numbers, the next made my brow furrow, with its featured click-through rates per placement and call-outs for the highest and lowest performers.
The third slide frustrated further with click-through rates by size, and the fourth had me full-on disenchanted.
The network had listed two recommendations for future advertising: drop placements that generated low click-through rates, and run for a longer time period to increase ticket sales.
I was flabbergasted at what I had read.
This network had spent an entire presentation talking about clicks, and then decided to bring up lead generation?
Perhaps most importantly, where was this network?s ability to communicate value beyond the click?
Licks, not clicks
In display advertising, the word ?performance? is a funny one, just like ?optimize.? What does it mean, really, when someone says ?I optimized my campaign by focusing on the ad that performed the best??
Those two charged words remind me of an English class in which the teacher, frustrated at our inability to use descriptive adjectives, asked us if we would be interested in taking home a nice big bag full of good stuff.
The teacher made her point.
How could you truly know if you wanted the bag unless you knew what was actually in it? Likewise, how can you ever optimize advertising initiatives if you do not really understand how they are performing?
Imagine that you have bought 1million impressions which have resulted in a healthy 0.5 percent click-through rate.
Of those 5,000 clicks, a strong 1 percent convert within the session, generating 50 click-to-conversions. Wow, you think, seems solid. Is it really, though?
By scrutinizing the data using a simple cost-per-acquisition model, you find that those 50 conversions actually ran you a $200 CPA.
Unless you are Bergdorf Goodman and selling Gucci loafers ? and I can guarantee the program will not generate a 1 percent conversion rate from click ? the campaign simply does not pay out fiscally.
Now let us add in another data point to round out the argument.
Envision that these same 1million impressions were served against 100,000 uniques.
Again, why judge media efficacy using 50 click-to-conversions, or even 5,000 clicks, when there is an audience of 100,000 consumers whose behavioral intentions are begging to be discovered?
The data do not improve when looking at the supply side of the business.
Recall that our network reference above advocated dropping sites to increase future click-through rates.
If you follow that to its logical conclusion, fewer sites generally means a smaller audience which means fewer consumers are exposed to messaging.
For publishers, this means a reduced fill rate and unhappiness in their network relationships.
For advertisers, fewer uniques messaged directly correlates to a reduction in the probability of influencing future consumer behavior. Everyone loses.
Using the same data as in our click-through rate example, we find that had we followed the network?s recommendation to increase click-through rate, not only would we have cut our population of messaged consumers in half, but our revenue would have also declined by 40 percent.
As I survey the incredible talent in the mobile space, I am dismayed that the industry continues to use clicks as a proxy for any indication of performance.
We have CEOs of large mobile networks who were heavy hitters online.
Many advertisers buying mobile display also purchase online, and are well-versed in non-click metrics.
Major players in this industry have and continue to use insights outside of clicks in online display advertising ? why do not they insist upon it in mobile?
It seems akin to accepting a dial-up Internet connection after having experienced the speed of a T3 line: a gargantuan step backward.
Given the simple mathematics above, there is no reasonable basis to use clicks when measuring display advertising or financially justifying continued mobile spend.
Yes, clicks are convenient. They are simple and fit easily into spreadsheets.
Unfortunately, clicks also provide non-significant sample size, rather uninteresting data, do not fiscally justify spend and hinder expansive advertising initiatives, as in our network example. It is pretty clear: clicks are for kids.
Elizabeth Joy Zalman is cofounder and chief marketing officer of Media Armor, a Boston-based cross-platform verification and measurement company. Reach her at .