No More Eyeballs — The Impact of Voice on Digital Content Providers

Nicolas Hinternesch
AT Internet
Published in
7 min readMay 9, 2019


Ever since I picked up my first Echo, the flashing ring of blue light has kept me on the edge of my seat. And after developing a handful of insanely meaningful Alexa skills and turning my lights on & off a few times, I started thinking about the bigger picture. How does the recent outburst of voice user interface adoption impact our digital ecosystem?

From the perspective of a passionate Alexa-for-everything-user, Senior Digital Analytics Consultant for the publishing industry, and AWS certified skill developer — Here are my two cents for digital content providers:
Why the concept of online content is going through drastic changes, how to successfully respond to the voice trend, and what to account for when it comes to analytics & tracking.

Change A: The Interaction of Users With The Digital World

The way we interact with content on the internet is constantly evolving and basically boils down to the evolution of interfaces. Before we ultimately reach full-on singularity (right, Elon?), it seems like we are finally getting over the single most consistent middle man between us and the internet: The screen. The step only seems logical, considering that our voice is our most natural tool for communication. This inescapably leads to…

Change B: The Way of Accessing and Consuming Information

Picture how users go about finding content on the internet. In a screen based world, the entry point for navigating the depths of the world wide web used to be a search engine. The results (as biased, ad-influenced, or skewed as they might have become), then distributed the traffic to different content providers. At that point, users could make an informed decision on where to continue their search for information and they would knowingly enter the universe of a particular site and be exposed to everything on it.

Voice user interfaces (VUIs) disrupt this entire process. They redefine the entry point to the digital world. When they receive a request, they can generate a response from multiple different sources. They rely on built in functionalities or pull content from other web sources. However, all of this happens behind closed doors and stays within the scope of the voice assistant. There is no more distribution of requests to different sources. Instead, voice based technology creates a world of generic one shot responses. Information and content from a black box (quite literally actually).

There is no ambiguity, no bias, no clear visibility of sources. While this generally has a huge impact on our society’s content consumption as a whole, the biggest drawback for content providers is that their platforms are no longer fully exposed to the users. It is their information at best. This has a huge impact on the business model and basically results in…

Change C: The End of The Classic Advertising Model

Let’s assume your piece of information does end up in a voice response. (Remember, there is only one spot in each response. There are no search results pages with multiple options, which makes voice searches much harder to rank for.) Even if content from a certain producer is being presented in a response, the user never actually lands on the provider’s platform. This implies that content platforms are stripped of all means that help them drive essential goals: There is no way to trigger further consumption, increase engagement, drive recirculation and visit frequency, or actively improve user retention. But most importantly, they are not getting the users’ eyeballs on their platform. Which is basically their currency. No continuous user attention = no ad revenue.

The Path Ahead

All of this sounds a little dramatic. And we are not quite there yet. In spite of the increasing integration of voice technology in many households and industries, not all content-based digital products face the threat of being cannibalized. However, it is still worth evaluating now, whether certain business models might have to adapt. Being a content provider, this could go down two basic routes:

1) Isolation

One approach would be to block voice assistants from accessing your data, i.e. not making your content programmatically readable for them to scrape when they receive a suitable user request. You would rely on people to access your content the ‘regular old-fashioned’ way. However, would you really want to neglect the trend entirely? You would willingly agree to being digitally invisible. Invisible to everyone who interacts with the digital world via voice.

2) Embrace

If your product can be fully used & experienced via a voice interface: Jump on board. Build an Alexa skill. Develop an action for the Google Assistant. At the current stage, the major operators of voice assistants are highly receptive to third party applications, as they extend the functionalities of the devices and drive adoption. That way, you are making your product available for this new technology, while simultaneously keeping users in your very own palm.

When it comes to voice application development, it is critically important:
a) to be relevant (how does your content benefit a user’s voice experience)
b) to effectively design for VUIs (an entirely new field not to be underestimated)
c) to add unique customer value to your skill or action

Added value will result in higher retention and will prevent your third party application from being replaced by a built-in feature of the voice assistant. My personal experience says: If installing a third party application does not come with a unique added value proposition (that only you as the third party can deliver), its functionality might sooner or later be integrated as a native feature, which will make your skill/action obsolete.

A great success story from the UK is the BBC. They have taken a detailed look at their huge portfolio and singled out core products and services that transfer well to the voice technology. As a result, they have released a wide range of such services as stand-alone Alexa skills, one of which is the much acclaimed BBC Kids skill.

Why has this approach been successful? In a nutshell — by addressing the three core success factors:

a) They are using their name, weight, popularity, and leverage to add to the experience on the platform
b) They adapted the content and its presentation to the new environment. The skill is not a like-for-like reflection of the website. Instead, it focuses on content that works well via voice (storytelling & gaming) and uses a well designed multi-turn dialog flow, which is tailored to the audience
Effective VUI design
c) They add specific user value that incentivizes usage and is unique: BBC-exclusive content, account linking possibility with the BBC ID, possible content personalization etc.
Added unique customer value

The Challenges of Voice Analytics

Adding voice activity as a new component of the digital portfolio (i.e. managing and maintaining a skill or an action) means adding another layer of complexity to a company’s analytics setup.
As a Senior Digital Analytics Consultant at a leading European analytics vendor, I am working with various clients on their voice measurement strategy. Here is my breakdown of what I believe to be the three keys of sophisticated voice analytics:

  1. Strategic Integration
    Voice is an entirely new platform, which goes far beyond pages, screens, and clicks. As a result, measuring a voice application will lead to unique new data-points that differ considerably from the regular web- and app-tracking data. The key here is to embed voice data in your current tracking architecture and make it comparable with your other analytics data. Only that way, you can get an aggregated view of your digital content’s performance across all platforms.
  2. Customization
    Voice applications are usually built on a combination of unique components that are exclusive to the platform (user-intents, utterances, variables, or multi-turn dialogs). In order to capture these unstructured user interactions and turn them into meaningful data, customized tracking is essential.
    While Amazon, Google & Co. do provide standard analytics for voice-developers, these reports are quite limited. The data tends to be kept behind closed doors and is only made available in top level reports of aggregated data. It is usually not exposed as raw data via API and cannot be edited, extended, or customized.
    Working with a third party analytics solution (like AT Internet or Google Analytics) allows for a much more sophisticated tracking implementation based on your own custom dimensions, events, and metrics.
  3. Technical Implementation
    The practical integration of enhanced tracking depends on an application’s technology. It usually requires sending individually crafted tracking requests for the specific events to be measured.

Bottom Line

— Understanding the shift in user behavior
— Reassessing the business model
— Addressing the unique challenges of VUI-development
— Implementing new forms of analytics

Digital content providers are required to navigate uncharted waters.
But the time to start is now. Voice is an opportunity rather than a threat.

— If you’re up for talking voice strategy, need help with skill development, are looking for analytics advice, or want to discuss all things Alexa:
Feel free to
reach out.