At the end of 2017, Google’s NLP (natural language processing) project achieved a word accuracy rate of 95% for the English language. This threshold represents a recognition parity with that of a native English adult speaker. NLP is a form of unsupervised machine learning which broadly refers to the study and development of computer systems that can interpret natural human speech and text. Human communication is frustratingly vague at times; we use colloquialisms and abbreviations, and often, don’t bother to correct misspellings. These inconsistencies make computer analysis of natural language remarkably complicated - especially for machines. As such, reaching the 95% threshold adduced above is an especially consequential milestone.
There are myriad implications of NLP’s arrival at this threshold of competency. The linchpin statistic is that in 2019, 65% of all Google searches will be on mobile.
Google Ads’ mantra to 'Deliver the right ad to the right person at the right time' is so conceptually brilliant because its brevity and simplicity embodies the logical end point of an extraordinarily dense network of precursor and conditional technologies, as well as referencing the changing economics of scale associated with the internet. These factors have combined to render Google’s mantra anything other than an extension of blue sky aspiration.
In many ways, Google has covered a significant amount of ground towards the actualisation of this mantra - as anyone who has been in a foreign city and searched for something like, “cocktail bars near me,” can attest. However, until the 95% accuracy threshold was reached, information about nearby cocktail bars was more efficiently retrieved by typing the question into the search bar.
Because cogent prose and well-formed sentences are much harder to master than colloquial conversation, most people are more comfortable expressing themselves in speech than in written word. For example, WhatsApp is increasingly an exchange of voice notes for on-the-go urban citizens rather than a platform for sending text messages. Even when loaded on a desktop, WhatsApp exchanges consist predominantly of visual or video material. Speech separates man from all other animals, and Voice Search will usher in a distinctive new era in digital marketing.
Let’s turn to a summary of the major practical implications of high functional Voice Search for campaign managers.
1. Accelerated Mobile Pages (AMP)
Google Search is primarily used on mobile devices. Given that 65% of all searches are on mobile and that page load speed is a ranking factor for mobile searches, campaign managers should target their voice search ads to mobile devices and encourage clients to use AMP to speed up load times on their sites. If you encounter any institutional arthritis in this regard, remember to deploy the following information politely: 53% of users will leave a site that takes longer than three seconds to load.
2. Dynamic Search Ads (DSAs)
Since the advent of AdWords and now Google Ads, and despite the utility of regular search query reports, campaign managers have faced perennial challenges achieving adequate and, more importantly, relevant keyword coverage. Amazingly enough, it was recently found that 15% of searches seen on Google.com are brand new - every single day.
In addition, websites are larger and more dynamic than they have ever been. Smartphones combined with voice search are ideally placed to respond to this proliferation of additional search moments in new contexts.
Campaign managers should make wide use of DSAs in voice search campaigns because, on average, 86% of DSA traffic is incremental. DSAs will make sure you don’t miss any relevant queries and ensure your ads will show when relevant. DSA functionality still allows full transparency via the search query report, equivalent control to that available on ETAs through bids at the dynamic target level and exclusion lists for irrelevant traffic.
3. Capture attention through automated creatives
Automated creatives will render it easier for campaign managers to create more ads per ad group. Automation will make it simple to create compelling ads that will draw in highly distractible voice searchers.
Here are some of the new automated creative solutions available.
- Ad Suggestions: use a combination of machine learning and human review to create high quality ad suggestions from your existing ads, extensions and landing pages.
- Responsive Search Ads: are generated by mixing and matching headlines and description lines that campaign managers provide, simplifying the creation and management of text ads.
- Fully Automated extensions: are directly populated by Google. These are shown automatically when Google predicts they will improve ad performance.
Andrew Alexander is the digital marketing and business development manager at The Media Image.