How to give voice to your analytics
A view from Sarah Salter

How to give voice to your analytics

The rise of voice assistants and smart speakers has brought about a fundamental shift in consumer behaviour.

Growing numbers of users are happy to search - and, increasingly, shop - via voice, rather than screen. As with any behavioural change, understanding the impact on the customer journey is critical. So to do so, brand owners and agencies must also understand the challenges and opportunities of voice metrics.

The importance of voice as a user interface cannot be underestimated. Consider this year’s Consumer Electronics Show, which took place in Las Vegas in January, where voice control/voice recognition was ubiquitous and multi-device – applied to everything from controlling cars to showers, mirrors, toilets and even medicine. It proved how far things have come in just 12 months since CES 2017, when attention revolved around Amazon’s Alexa.  

In terms of understanding the full consumer journey, voice analytics are vital

The message of CES 2018 was simple: very soon, voice interaction won’t be a premium feature in our homes or cars, but a standard option. Yet it was intelligent personal assistants, such as Apple’s Siri, that established the voice marketplace. Moreover, when you consider that combined voice-assistant distribution now stands at 400 million Android and 600 million Apple devices, it’s clear that voice is already a mature proposition, according to Ravi Lal, marketing and commercial director of voice platform Opearlo . 

All too often, however, brands and agencies fail to fully integrate voice across the user journey and, when they think in terms of analytics, it is often left out. In light of this, Wavemaker is identifying where voice has a significant impact and can be most useful for brands – both in the priming and active stages.

In terms of understanding the full consumer journey, voice analytics are vital. Emphasis to date has been focused mainly on measuring what’s happening within a voice application. This means measuring two broad collections of variables: use of a service – new users, active users, retention and so on; and voice-event tracking – the specialist area within voice analytics. 

Voice-event tracking measures which parts of the voice app are being used, how often and how successfully users are navigating through various sections and how many complete the full journey.  

Gathering user data – email addresses, for example, which are necessary for re-targeting and other CRM-based marketing – is not simple, however. Explaining this need and making it transparent and easy for users to do is a key design challenge. 

Looking ahead, the next generation of voice analytics will need to link together the full user journey – so-called "attribution tracking". Starting with paid and free traffic, the first question should be: where are we getting voice service activation? The next is: what’s happening within the voice app, up to the point when users leave that app and purchase – for example, via Amazon?

The goal for brand owners and agencies should be to know who clicked on their display ad, then clicked through to the landing page and downloaded the voice app, then searched and, lastly, acted – the kind of end-to-end thinking that is critical in terms of proving ROI. 

This is likely to be especially crucial, given the work and investment that’s going into linking voice-activation with screen technology, not only from Amazon with its Echo Show device, complete with a screen, but also major consumer electronics brands. The latter includes LG, which showcased a smart display powered by Google Assistant at CES.

Analytics platforms currently available have both strengths and weaknesses, which is why we will support several of them, according to different clients’ particular wants and needs. 

Certain elements will be essential for any package, such as: event tracking; attribution linking (tracking inbound and outbound traffic to and from the voice app); per-user level tracking (to provide one-to-one personalisation within an app); and ability to sync with CRM and other customer marketing systems. 

Voice analytics will always provide product insights via event and phrase tracking. But voice analytics needs to rapidly evolve from being a standalone, siloed dataset, into something that feeds a joined-up view of the customer, across all online and offline touchpoints throughout the user journey.


Sarah Salter is innovation director at Wavemaker 
@SarahSalts