A few months ago, I wrote that paid search won't scale for brand advertisers. My basic premise was there isn't enough available paid search inventory to let a large advertiser spend a significant portion of its budget. Most paid search keyword buys are direct response (DR) and drive high eCPM (define) for the inventory owner (the search engine) because this inventory is scarce and in high demand. It isn't possible to spend hundreds of millions of dollars a year on search as a single advertiser because there isn't enough inventory to buy.
I further laid out a taxonomy for various types of inventory:
But one thing nagged me: shouldn't there be such a thing as low eCPM brand inventory? Brand advertisers certainly buy cheap remnant inventory on occasion, so the low eCPM DR inventory category is a bit fuzzy anyway. Then it came to me while I was on a search brand advertising panel at Search Engine Strategies -- where I planned to be the contrarian, due to my comments about lack of sufficient inventory.
Some of my co-panelists described scenarios in which they'd used search to achieve branding goals. They specifically chose keywords that wouldn't have many competitors or achieve a high click rate but that still drove branding success metrics. They took advantage of the fact there are keywords that don't drive great performance from a CPC (define) or even CPA (define) standpoint, but that represent a point in time when the audience is open to a branding moment.
For years, I've talked about modality when describing the moment the audience is engaged in when we choose to advertise to them. Great ad opportunities typically coincide with moments where the audience is in absorption mode rather than in action mode. If I'm on a portal home page, for example, I'm probably open to seeing an ad. If I'm editing my MySpace page, I probably don't want to see an ad. If I'm reading e-mail, I'm open to the ad, but if I'm writing one, I'm not.
We've known since its inception that search is a great medium for advertising because the audience is very far down a specific goal path when we show an ad. If the search has commercial intent, clearly the searcher is very far down the purchase funnel and is receptive to see ads related to the search.
But what about all those searches without commercial intent? Are great branding opportunities there? Despite the lack of large amounts of inventory and given automation, couldn't we build scalable infrastructure to enable high eCPM brand inventory on search? Today that inventory is good for branding, but in low supply. And because it's too granular to buy a large reach campaign via search; it just doesn't tend to happen.
But what if we started buying keywords or keyword phrases that matched brand characteristics we wanted to leave with the audience? What if we took this a step further and made non-commercial search queries brand inventory only, with branding-appropriate ad formats?
Say the advertiser is Downy Fabric Softener. What if it bought keywords that were aspirational of brand, like "soft," "softer," "fresh," "breezy," "comfortable," and "cozy"?
And when someone searched on those keywords, Downy would have a 300 x 250 display ad in the center of the SERP (define), or at least a skyscraper replacing the paid ad spots. Heck, let's take it a step further and let Downy buy that inventory as a CPM. This would benefit everyone in the mix and build new types of brand inventory that would likely be very effective.
It still brings us back to the problem of insufficient inventory, but at least it helps monetize the other parts of search better while providing brand-friendly vehicles.
May 22, 2013
1:00pm ET / 10:00am PT
June 5, 2013
1:00pm ET / 10:00am PT