Monday, November 23, 2015

The Advertising Conundrum


When our blog kicked off in 2013 I wrote an entry about opting in or out, discussing some of the pros and cons. In that short time, cable advertising has started to mimic some of the internet ad methods to become better at targeting online viewers. While TV ad targeting lags a bit behind, cable companies are working hard at Dynamic Ad Insertion (DAI) to catch up, even using online behavior data for personalized ad selection. When I wrote the first blog post, I suggested I that wasn't too concerned about sharing my data if it helped focus the ads I saw, and generally I still agree with that statement. However, my behavior would suggest otherwise. The one shortcoming of online internet ads is the inability for it to know when you have made a purchase decision. I was getting ads for outdoor lighting over 3 months after I purchased the lights from that very company! Subsequently, I am now blocking ads in my browser.

A new measurement approach to online video banner ads has been in use for quite a while. Internet ads often allow skipping after a designated watch time. Some believe this provides a better value to advertisers by assuming a higher quality impression rate, by analyzing drop off rates and knowing the number of people who watched the full ad. In television, there are often contractual requirements to block ad skipping; this has fueled a growth in products that provide time shifting to circumvent the ads altogether, causing a legal uproar, and prompting a counter attack with software to block the blockers. However, it seems there might be value in similar on-line models for cable and advertisers to consider. Ad skipping could provide more data for determining how to target ads. Combining various data sets and using smart algorithms to place ads provides technical challenges, but has the resulting potential for better ad consumption. In the meantime, the only ads that definitely won't get skipped are the product placements inside the programs.

No comments:

Post a Comment