Three questions marketers should be asking about voice – the unseen channel

Daniel Benton
By Daniel Benton | 9 July 2020

Daniel Benton is GM of Neo Media World Australia.

Australian households have adopted voice enabled smart devices at a world leading rate according to research released in 2019. This rapid acceptance of voice technology by Australians has many marketers eyeing off the opportunity to reach potential customers through a new channel that is rapidly reaching critical mass. So, what is a marketer to do?

Given the relative infancy and complexity of the voice ecosystems it’s crucial for marketers to understand the fundamentals and answer the following questions before locking in a strategy.

Which ecosystem am I solving for?
The past few years have seen a fierce land grab by the key voice players to win market share locally and globally. While there is currently no winner globally there are some very clear contenders jostling for dominance with Amazon, Google and Baidu capturing +60% of Q4 2019 global sales.

However, there is a significant variance of what this looks like at a country level. Within Australia, Google has managed to carve out a commanding lead with an estimated 72% to 79% share, with Amazon capturing the remainder. Google’s dominant position has been achieved through a combination of aggressive pricing and partnership with brands like Woolworths, Bunnings and Spotify. The landscape looks completely different in the US where Amazon has been the clear leader, initially defining the category with Google fighting hard to regain ground in second spot.

Understanding these nuances are important as each ecosystem uses different search engines and data sources to surface answers; Google for Google home devices, and Amazon A9 and Bing for Amazon devices. They also have different development frameworks to build applications; it’s “actions” for Google and “skills” for Amazon. These factors mean that trying to address multiple ecosystems can mean a duplication of effort and resources. It also means global brands can’t simply roll out a uniform approach across all markets at the same time.

Given its dominance in this market, when it comes to reaching more consumers through voice – the Google ecosystem should be treated as a priority. Additionally, voice in the Google OS provides secondary side benefits of improving voice search and core search coverage.

However, there are secondary questions that need answering if your brand is ecommerce enabled and currently selling or planning to sell via Amazon. If so, Amazon should be strongly considered as it will drive growth as it makes more inroads into Australians homes. The nuance is that not all categories lend themselves to voice commerce in its current form. Think of the difference in consumer path to purchase between restocking a kitchen staple item versus researching and buying a pair of high tech Bluetooth headphones.

Should I build a skill or action?
The emergence of voice has distinct parallels with the early days of growth in mobile operating systems- and Skills and Actions are the equivalent of mobile apps.

Like it was then, the idea of developing an app to engage potential customers is quite seductive to marketers as we’re in the early days of a new channel that has seen mass consumer adoption. It can be a relevant and effective strategy but the reality of whether an investment in developing a skill or action will yield a return is sobering. Analysis by Neo in the US showed a limited adoption of both skills and actions;

  • The research found there were over 60,000 Alexa Skills for Amazon devices, but 61% didn’t have any user ratings.
  • The audit of Alexa skills deployed by several B2B and B2C brands found that the daily active users were typically less than 100.
  • The Google ecosystem didn’t fare any better with only 4,000 Actions enabled in the U.S., and most of them were games or educational based. They were difficult to find, and significant growth was not anticipated.

My suggestion is that any brand planning on building a voice app needs two things to make it a successful strategy;

  • Be confident the app is addressing a use case that can be solved uniquely via voice.
  • Be prepared to unlock investment to promote the app to potential users and drive growth. Without it, it’s highly unlikely users will organically discover the app. 

Given these parameters a better starting point is to understand how a brand is showing up in organic voice search results. This insight can then be used to develop a roadmap to address any gaps and ideally to build a competitive advantage.

How is my brand showing up?
Understanding how a brand is being presented within voice search is tricky, with no results to visually review (and potentially scrape) it’s extremely difficult to audit a brand’s “visibility” and understand the competitive landscape.

At Neo we’ve addressed this issue and have developed technology underpinned by AI and the cloud that allows us to programmatically mine voice results to provide a view into the results being served to users across all key operating systems.

The interesting trend we’re observing both locally and globally is that over 50% of the time devices are unable to provide an answer to common questions – including branded questions. As an example, we found that for 35% of branded queries for one of Australia’s top health insurers the queries were being answered by other sources. Additionally, we uncovered the brand had zero visibility for high value generic queries. These results highlight missed opportunities on both counts and demonstrates both the challenge and opportunity for brands to consider and plan for the presentation of content via voice.

Undoubtedly, voice presents an exciting, yet nascent opportunity to extend a brand’s reach. However, it’s a nuanced channel with a unique set of complexities. It’s also a rapidly evolving customer interface with myriad potential future applications. As an example recent Australian research from WPP AUNZ’s latest Secrets & Lies report Humanity & The Machine, suggests that the next horizon in voice is tech that can infer meaning and intent from a user’s tone. Potential applications of this technology would allow brands to create more tailored customer experiences and surface product recommendations based on a user’s inferred underlying mood. That’s the future, but it’s not far off.

Ultimately Australians are asking more questions of devices – is your brand listening?


comments powered by Disqus