A screen affords space for multiple ads. However, with voice, we want one answer. We want the answer. A consumer looking for a new TV via a voice search isn't looking for ads from different TV brands or retail stores -- the types of ads that appear on search results pages now. With voice search, consumers are looking for a concise guide on the best TV to buy.
If voice search does find this one appealing answer, and the consumer says, "read it to me," the content acts like a mini-podcast. Podcasts are only becoming more popular, with 42 million Americans listening to podcasts weekly. And studies have shown that ads played during podcasts have good recall among listeners. Considering the "read it to me" option, we can see how portions of time spent with media are transitioning from in-content to linear media, a format that requires interruptive advertising.
But this transition may bring about two problems for advertisers. First, we listen to or watch content slower than we read and skim it. Since there are still only 24 hours in a day, this may lead to a decrease in overall ad-impression opportunities. Second, we don't tolerate interruptive advertising as well as in-content, as is demonstrated by cord-cutting, DVR ad-skipping, and even the :30 skip button in podcast apps. All of a sudden, an article that may have displayed five ads, now contains one audio ad -- one ad that is susceptible to being skipped. Publishers can't afford to make less money per unit of consumption, so get ready for CPMs to head north, and for "audibility" to become an important metric.
The potential transition to just one paid search ad for a voice search may also dramatically impact attribution. If there is only room for one paid search placement, Google may not be able to afford to continue its CPC model. In fact, Google is already shifting to multi-touch attribution. After all, if an advertiser owns 100% of the impression share during that voice search, there is value to that impression, and that impression may prove to have an impact on an eventual purchase even if it's not acted on at the moment. The need to tie together more devices than ever -- devices that process and deliver audio requests -- makes attribution all that much harder.
For publishers, voice search will introduce an entirely new content production division. In the beginning, the "read it to me" option may suffice. BUT soon, Alexa's and Google's robotic voices may lose their appeal, and I predict publishers will begin producing content that is specifically designed for voice-enabled devices. Intriguing voices and new ways of drawing in the user (without visuals) will bring us back to the radio glory days, but with a digital twist.
In addition to the way content is produced, publishers will also need to rework monetization strategies. ComScore predicts that by 2020, 50% of all searches will be voice searches. To ensure that publishers are making money on these searches, they will need to create new ways of structuring subscriptions, adapt to new ad formats and prepare for a new ad model in which there are far fewer ads. The industry and our governing bodies should begin acting now to experiment with, and standardize, ad formats, so that these ads may be bought addressibly, programmatically and measurably.
Voice search will change how we create content, optimize content, deliver content, and will definitely change how we think about buying, delivering and measuring ads. Advertisers, agencies and publishers must operate at the pace of the consumer, not at the pace of their own approval processes.