We swim in a sea of data. Focus groups, transactions, all sorts of surveys, Web traffic, assorted non-sale responses, TV meters, portable meters, mall intercepts, and contests - and that's just to name a few. But what matters, and what's the value? These are becoming strategic questions for many companies.
Answers lie in what metrics can be used to run your business. Management uses key business indicators that are as close as possible to the operational decisions they can control.
Imagine if your metrics guaranteed specific sales for each media unit purchased. The decision of how much to spend on that media unit becomes academic. On the flip side, with sufficient information across industry verticals, these metrics also illuminate the value of the media unit to the broadcasters.
The Google model invites prospective customers to bid for responses to keywords. Google makes dollars per response delivered, using Ad Sense algorithms to estimate how many responses it can deliver per bid to decide who wins the bid. But just because company A bids $2 per response and company B bids $1 does not mean that company A wins. If Ad Sense predicts that company B would get three times as many responses as A, then company B wins. And so does Google, by optimizing the monetary value of its search words.
Imagine the capitalization of the company that replaces Google's pay-per-click response model with a pay-per-inquiry/visit/sale response model. It's just a matter of time before someone corrals our increasingly electronic world and delivers this.
The simple solution is to collect everything in a single panel of consumers. The traditional barriers have always been the cost of specialized measurements to track different things and their consequent strains on maintaining cooperation in a representative panel. The traditional answer has been vertical specialization to manage for panel "burn-out" and measurement costs.
However, as technology evolves, these economics may change. Measurements may become increasingly passive, minimizing cooperation burn-out, and more similar in nature, driving costs down.
A key in the evolution of measurement might be audio signals. They can emanate from anything. The Apollo Project, an ROI measurement system from Arbitron and Nielsen, measures proprietary audio signals encoded by broadcasters of TV, radio, and Internet video. Wal-Mart has been testing universal audio signals emanating from RFID chips to manage inventory and expedite checkout without having to scan UPC barcodes. If privacy issues subside, imagine unique RFID signatures emitting from cash registers at all sorts of retailers. Not only do you know that Johnny with his personal RFID listening device went to McDonald's, but by synching up with the store's time-stamped data, you know he bought a Big Mac, large order of fries, and small Diet Coke.
What do manufacturers, retailers, and service companies want responses to? If McDonald's wants store traffic or even specific purchases, the innovative Data Company could provide that by synchronizing its panel of RFID listeners with all the retailer's cash register databases. In each case of goods, there will be an abundant number of chips to listen for, as many predict RFID chips will replace barcodes. In the interim, much can be made by synchronizing RFID listening data with cash register data.
This evolution in response data may finally break the age-old CPM model, as a new Google model could transcend broadcast media. Advertisers would bid for inventory and pay-per-response. Google-like intermediaries or the broadcasters themselves will pick winners based on the value of the bid times the anticipated response rate. Optimizing the value of media units by predicting response rates will be the next war of algorithms.
Google is aggressive and ambitious. It wants to organize the world's information, and it's actively experimenting in managing ad placement to finance its ambition.
Will broadcasters become serious about optimizing the monetary value of their media units, or will they let themselves be co-opted by distributors such as Google, Apple, Microsoft, or possibly Comcast? Or will broadcasters and advertisers hire their own algorithm specialists and work through an independent exchange to process these transactions? Only time will tell.