@JMendez:
I’ll take a look at the case study, thanks!

“The endless possibilities of dynamic display are actually not so great for testing. The optimization has to been done on click through rates that are really low are rarely factor in network/publisher data (where the ad was running, what position on the page, what time of day, etc) thus bringing lots of noise into the result sets.”

Fortunately, there are other options than straight CTR optimization. While CTR does factor in, most of the clients I’ve seen prefer to optimize on some hybrid of (rich media) interaction rates and/or directly on conversion. (Conversion events can be so statistically sparse, that practically speaking you do need other measures of performance as proxies.)

Endless possibilities of display do provide a challenge, but the right way to address is to combine dynamic creative with automated optimization algorithms that relieve much of the quantitative and analytical burden of the otherwise daunting permutation counts.

“This is not to say I’m not a fan of dynamic display – i’ve worked quite a bit in that area – but it’s not a panacea. You can get amazing results from a 3×2 MVT on a 300 x 250 gifs but how many people even do that? Hardly anyone.”

I’m a fan of dynamic display too – particularly when it’s paired with automated optimization algorithms to address some of the very issues you highlight. In fact, that’s my day job – working with these types of technologies is pretty much all I do. And you might be surprised by the strides in adoption made by these technologies in the last 6 months, and not just with 3×2 MVTs – but with far more complex setups that often entail literally millions of permutations. (Again, only with automation can complexity like this be realistically handled, but it’s just beautiful when applied correctly.)

Great comments all, thanks!

The creative aspect of search is incredibly undervalued. Your gut is similar to many others but search creative has huge influence. There is plenty to optimize on both conversion & CTR which influences Quality Score and ultimately bids and ROI. I’ve done a large number of multivariate tests on search creative and 1) it has a huge impact on conversion 2) that influence can vary by different elements – title, d1, d3, url 3) performance can vary greatly. Here’s one case study I presented at SES three years ago: http://bit.ly/2VFjqw

The endless possibilities of dynamic display are actually not so great for testing. The optimization has to been done on click through rates that are really low are rarely factor in network/publisher data (where the ad was running, what position on the page, what time of day, etc) thus bringing lots of noise into the result sets. This is not to say I’m not a fan of dynamic display – i’ve worked quite a bit in that area – but it’s not a panacea. You can get amazing results from a 3×2 MVT on a 300 x 250 gifs but how many people even do that? Hardly anyone.

One last point to keep in mind — and this is just a function of cross testing and based on lots of personal experience. The ad generally is the most influential factor in conversion rates no matter what you are testing in the stack however the bid and the landing page ultimately control the overall performance results. Ain’t that a trip? Happy optimizations!

Actually I think that a big difference between display and search will be the creative.

What’s the character limit on search creative? 125 characters? I like words as much as the next guy, but there’s only so much you can do with less than a tweet. I’m sure that others will object, but my gut is that there’s just not that much to optimize.

Compare that to the sheer permutation loads that can be found in display advertising – particularly when you bring both rich media and dynamic creative generation into the mix. It’s no cliche to say the possibilities are endless.

Looking at Search it’s easy to echo Darren’s ascertation about landing pages and the complexities of service channel conflict.

So what happens? Bid management gets commoditized, ad creative get little attention and landing pages are all but forgotten. All the budget goes to media.

Smart marketers will always do well in either channel but they are few. In the end nothing changes here. We need new paradigms.

I agree with your assessment on landing pages. It will be a long time, if ever, before advertisers give up control there.

Second, I think that the majority of the advertisers will not give up the full stack if it includes the landing page – though for the ones that do, will see tremendous return.

The reason for this is because the process of media optimization, creative optimization, and landing page optimization generally is split across multiple parties. For big brand advertisers with many agencies, it’s going to be hard to get everyone to give up control unless the brand really pushes for this and that’s going to take a while.