Search Engine Wish Lists, Part 2

Last week’s search engine wish list ended with some video-related wishes. This week, I’ll finish up the video wishes, then continue with both my wish list and wishes expressed by readers. Thanks so much to all who provided feedback.

Video Ads

I’d like to see pre- and post-roll video ads that are short (perhaps :05) targeted against geography, daypart, user profile (behavioral or volunteered demographic profile), or content channel. Content channel targeting is too broad, but I understand the challenges in developing a video ad platform that splices ads in dynamically, rather than in a one-size-fits-all fashion so all viewers of a particular clip see the same ads.

The key to successful video advertising is relevance. Consumers tune out irrelevant ads unless they’re funny or entertaining. Not every marketer can produce clever, witty video ads like Apple does. And with only five seconds in the proposed pre- or post-roll, being relevant is easier than being funny.


The engines need more accurate geotargeting options, preferably using both IP address and voluntary profile data. As I travel to conferences and client meetings, or stay in hotels, I occasionally do local searches without specifying locality. This allows me to see which geotargeting ads the engines are running. Too often, they’re wrong. The engines need better geotargeting accuracy.

With increased accuracy, the engines can also provide better precision. Right now, Google provides precision without sufficient accuracy. I can draw a polygon in Google around a small, densely populated area such as Manhattan and still get lots of clicks from folks in the outer boroughs. This means the system allows for a precise targeting area but can’t back it up with accuracy.

Even with poor accuracy, Google leads the way in precision, with several great options for defining a geographic area. MSN and Yahoo have to step up, particularly if they want to fully exploit the local market. Local advertisers need a small radius of targetability or a Zip Code level of targeting.

Yahoo-Specific Wishes

Yahoo should immediately merge its two sales teams and ad support structures. Yahoo’s powerful combined delivery value proposition can’t be delivered by two separate sales teams. Also, with the merging of behavioral display and dynamic text advertising (on last week’s wish list), the lines between display and search advertising are significantly blurring. Having one sales team is better for Yahoo, better for media planners, and better for marketers. Those buying media want to get more involved in buying media without having to establish and manage two entirely different relationships, each with its own churn rate.

Many of you agreed it’s time for Yahoo to allow advertisers to select pure Yahoo traffic and opt-in to the remainder of its search network. Some advertisers don’t have the same level of success with non-Yahoo clicks and therefore bid lower than they would if they could select pure Yahoo traffic. The ability to set up campaigns with such an option would facilitate higher prices on Yahoo and still allow advertisers interested in broad reach to target the broad network. In addition, some advertiser (myself included) favor rolling out bid boosts or depressions by traffic type. Some clients don’t do well with domain parking traffic, for example, and would prefer to lower bids there.

Network Opt-Outs/Domain Filters

How a bout Yahoo and Google providing advertisers with the ability to opt out of certain domains in the network or to opt out of any domain parking traffic (also called direct navigation)? The problem with any network is that some sites don’t work as well for a segment of advertisers. Forcing advertisers to buy these clicks harms the network as a whole, not to mention angering the advertiser.

Easier Campaign Cloning

With the arrival of geotargeting in all the engines, many advertisers will want to set up cloned campaigns for certain geographies in which the conversion rates are higher. Unless you use the APIs (define), like some of the bigger agencies, cloning a campaign is a real hassle. Imagine having to clone a campaign 210 times to address the designated market areas (DMAs) in the U.S., then doing it again for Canada an additional 43 times. It’s just not feasible.

Thanks to your submissions and the help of a bunch of folks who have spoken to me at events, there are still more items on the engine wish list. Next, I’ll devote at least part of my column to part three of the search engine wish list.

Related reading