Technology Review: Blogs: TR Editors' blog: Google's Play for Radio Advertising

About Us

Insights, opinions, and our editors' analysis of the latest in emerging technologies.

Google's Play for Radio Advertising

We're all used to context-sensitive, custom-delivered ads on our search result pages. Now that kind of targeting could come to radio.

Tuesday, January 17, 2006

By Wade Roush

Google announced today that it plans to buy Dmarc Broadcasting, a company in Newport Beach, CA, that automates the insertion of ads into daily radio programming across the country according to the advertiser's preferred geographic area, target audience, time of day, and so forth.

Dmarc's advertising network is, in some ways, the radio equivalent of Google's AdWords program, which supplements search results on Google and e-mail messages on Gmail with ads selected to match the topic of the current page. Indeed, Google says it plans to integrate Dmarc's technology into its AdWords platform, "creating a new radio ad distribution channel for Google advertisers," in the words of Google's press release.

But there's one big difference between Dmarc's system and AdWords -- a difference that makes Google's announcement either puzzling or intriguing, depending on how you look at it. AdWords is all about context: Google's software automatically analyzes the words on a search page or in an e-mail message, tries to pick the most important ones, and retrieves ads related to those words. That's why they call it AdWords.

Dmarc's system, on the other hand, simply automates the radio industry's existing ad-insertion process. It lets advertisers upload new radio spots, choose one market or multiple markets in which to broadcast them, and see instant reports about which ads have been played. The system is sensitive to context only in the sense that advertisers can manually specify which stations should play their ads, based on the stations' demographics.

In other words, Dmarc's network is a nice hack that simplifies radio advertising. There's nothing clever about it, in the way that AdWords cleverly reads a Web page and divines its subject matter, or in the way that Google's original search-engine service draws on the collective wisdom of the Web, by giving the highest rank to pages with the most incoming links. One wonders why Google's brainiacs would be interested.

But here's the intriguing possibility: Perhaps Google is working on technology that would "listen" to the word-stream in a radio program, parse its meaning, and insert ads related to those subjects. The company employs plenty of PhDs who study natural language processing, including machine translation, so this possibility isn't so far-fetched. How would the system work? If your drive-time AM station broadcasts a news report about chinchilla farming in Estonia, the report might be immediately followed by an ad for fur coats (to use a politically incorrect example).

And if such a system works for radio, what's to stop Google from entering the TV advertising market, or even outdoor advertising? If there's a context to work with, the distribution of any ad can be tailored. On the Web, Google -- and, to be fair, Overture, which was acquired several years ago by Yahoo -- have already helped advertisers solve the age-old problem of figuring out which half of their ad spending they're wasting. If the art of radio and TV advertising could also be turned into a science, Google could soon see its profits piling up even faster.

Comments

	Possibly a Simpler Reason? Guest (Mark C) on 01/17/2006 at 12:00 AM
Posts: 1	There's a third possibility. Maybe Google wants Dmarc to combine radio ads with its Adwords. Advertisers add metadata to each radio spot. When a user searches on a specific word, they may have the option to listen to a radio spot or click on a link. If this makes any sense, then there's no reason to believe that video ads couldn't be offered the same way. It will be interesting to see... Rate this comment: (Reply)

	Video and Radio ads on Google Guest (Nigel F.) on 01/19/2006 at 12:00 AM
Posts: 1	I think the immediate opportunity is what Marc C stated - it would be totally logical to use the vast traffic on G to promote radio ads and later on, video. In terms of Google ads on radio - yes, but the low hanging fruit simply has to be increasing the distribution of radio advertisers on Google as opposed to Adwords ads on radio. The question in my mind is, does Dmarc bring to the table a whole raft of radio advertiser relationships - I would imagine it does. Rate this comment: (Reply)

	Think about the future... Guest (Gary Delooze) on 01/18/2006 at 12:00 AM
Posts: 1	... where Radio comes over an IP transport, and content can be TIVO'd to create a tailored music programme, or where the hundreds of channels explodes into many thousands of channels due to the increased 'bandwidth'. In that world, context-sensitive advertising isn't too far-fetched for a radio station, as there will be a 1:1 or 1:few relationship with the listener that doesn't exist today. Commercially, it makes sense in that context to take a piece of the advertising pie. Of course the alternative is the realisation that Google (in spite of the 'cool' branding and deep technology) is just an advertising company after all, and wants to spread it's wings across multiple channels... Rate this comment: (Reply)

	Not so easy Guest (Name Witheld) on 01/18/2006 at 12:00 AM
Posts: 1	While theoretically possible, the technical, business and human challenges, at the Radio Stations, of overcoming the reasons why "logs" which list each commercial spot are generated at least 24 hours in advance are huge. To dynamically react to content, and on the fly determine which spot to insert you not only have to overcome the technical difficulty of the natural language parsing, etc, which I agree Google could probably do in time, but you might also find out that making this a reality in enough radio stations to gain momentum is a lot harder than you think... It would be an interesting thing, though, to see that pulled off anytime soon. Rate this comment: (Reply)

	Feasible, but not a good idea Guest (Thomas Rasmussen) on 01/19/2006 at 12:00 AM
Posts: 1	Interesting idea. However, why use all the resources on this - and while it certainly could be done eventually, it would require a lot of resources - when you could achieve virtually the same results simply from content-categorizing the various shows and programmes? The scenario you put forward is not really parallel to Adwords, in that it is not user-context driven (when they figure out how to insert radio ads that relate to your ongoing conversations, then you have it). While creating a system of real time content-driven ads on radio might be somewhat interesting, I hardly think the results - when compared to simple content-tagging - would merit using the resources. In other words, I seriously doubt this would be a good investment. If Google is actually doing this, somebody call them and tell'em :-) Rate this comment: (Reply)

	ad-spend/ad-campaign management Guest (Unmesh) on 01/24/2006 at 12:00 AM
Posts: 1	A simpler possibility is that G intends to provide an integrated ad-spend service (print, radio, web , etc) with a very useful ability to monitor/auto-adjust (algorithmically) the per-medium viewership(click)/ad-spend as the campaign progresses. Rate this comment: (Reply)

emerging technology
Guest (crystal williams) on 01/25/2006 at 12:00 AM

Posts:
1
how does emerging technology affect radio today?
Rate this comment:

(Reply)

	G should get into set top boxes Guest (Kiran) on 02/21/2006 at 12:00 AM
Posts: 1	...that can look at multi-channel data (radio programmming, radio ads, user context) and be able to mix and match according to users' needs...with the occasional irrelevant pop-up of course :) Rate this comment: (Reply)