Customizing Campaigns: Are We All Using the Same Data?

Our CTO Brian Malone highlights three ways Dedicated Media customizes our approach to optimizing campaign data.

Fifty years ago, stock market investors struggled with a new concept called the Efficient Market Hypothesis. It suggests that a stock’s price reflects all publicly-known information about it. Thus, no one should be able to beat the market, in the long run.

It seems we have reached this point with ad exchanges. Most inventory is available for real-time bidding (RTB), and data segments from third-party providers are ubiquitous. Can a company really expect to outperform its competitors using the same platform controls as everyone else?

Certainly, the ad market is different from the stock market. An outside party can’t arbitrage impressions in the same way. Impressions are immediately perishable. This provides some temporary advantages to companies bidding at the right time. Plus, there is private data such as past browsing behavior and previous campaign performance that can improve results. But when it comes to finding a new customer, most companies use the same inventory and targeting data as their competitor.

So how did Warren Buffett get so rich with these efficient markets? I don’t know, but it certainly wasn’t by doing the same thing as everyone else.

I’m not sure how much we’ve learned from stock market dynamics, but at Dedicated, our view of data has changed. We assume that the information about a user is public knowledge. If we know someone is in-market for a car, so do our competitors. The price of that data segment should increase to the point of unprofitable returns.

To overcome this “efficiency,” data needs to be customized to the company. There are three ways we are exploring this customization approach with our current campaigns:

1. Second-Party Data: Exclusive relationships with content providers where we know where the data comes from, much like selecting certain grapes for a wine rather than buying a bottle off the shelf.

2. Data Modeling: Tailoring public data to find significant trends. For example, we know that automotive conversion rates improve when the stock market goes up, but only if the stock market goes up three days in a row.

3. Unique Corporate Datasets: Social data is both public and informative for companies. The volume and sentiment of social posts can produce conversion rate lift in all digital campaigns, not just social campaigns.

An underlying theme to the three points above is people. You still need the attention of informed, insightful people for data and algorithms to evolve. Thus, a hybrid approach of advanced technology and dedicated people will always be at the core of our optimization strategies.