EYE Views: can voice assistants overcome the creepy factor?

Realising how third-party data has been used to target consumers makes them feel wary. They find it creepy. The misuse of personal data (real or perceived) and overt advertising drives the creepy factor up and that’s a risk for brands, not least because it makes consumers less likely to engage with an ad.

But, as people become more and more comfortable talking to voice assistants, the conversational approach might be the best way to mitigate the creep-out issue.

And with business use of voice tech set to triple in the next 12 months, we asked industry players how companies, brands and consumers should respond to the challenges and opportunities.  Here’s what they said…


Nick Nguyen, VP of Product, Firefox, at Mozilla

From our research as well as our own experiences with voice assistants, we find that the engagement with these systems is far different from the web. For instance, users tend to use voice assistants for objective facts and actions, like asking for a sports score, checking on the weather, or turning on a smart light bulb. In these contexts, users aren’t looking for opinionated answers or anything that takes more than a couple seconds of their time.

The advantage of these systems is convenience and speed, a hands- and device-free interaction with technology that is especially useful for those who prize convenience or even the ability to do things when they are otherwise engaged. We have not seen a huge demand for longer interactions with in home voice assistants. This may change as technology evolves, and a new generation of users grows up with home assistants just as the previous one grew up with mobile computing, and these users may expect their assistants to know them better to provide better answers.

One position is that advertising platforms will evolve to fulfill this need, and they may provide utility to users, but what is most important to us is that users understand and agree to the tradeoffs they’re making for this convenience.

color-linkedin-128 color-twitter-128 color-link-128 

Alex Barclay, Partner & Head of Innovation Strategy, Big Radical

Targeted, personalised ads have a problem. For many people they’re like a weird uncle at a party – creepy, socially awkward and intrusive. Both make advances you’ll resist.

The reason we feel like this is even the best targeted ads, ones using the latest AI, ML and data analytics don’t understand how people work. We want interaction to feel natural, warm, useful or entertaining in some way, not one-sided. Most targeted ads feel limiting and unempathetic.

Conversational UI is working to change this. These interfaces work hard to simulate what Marshall McLuhan would call a ’hot’ medium of communication – a quicker, more responsive engagement. They aspire to be unselfish but useful conversationalists.

Of course, there’s a catch. Almost all automata and other human simulators around are good at some things, but hopeless at others. Conversational UIs need to be told what to do, how and when. They also lack any sense of spontaneity or creative spark. Work needs to be done, and it is. That said you might still want them at your party. They could certainly help with ordering the pizza. Unlike a weird uncle

color-linkedin-128 color-twitter-128 color-link-128 

Krupali Cescau, Head of Insight, Amplify

We are in a transition period where the enormity of how invasive AI has become in our lives is creating a sense of panic. We’ve been accepting and implementing ever more sophisticated tech in incremental pieces for years now.

And all of a sudden, the boom in virtual assistants, not on our phones where they are compartmentalised into controlled activation through a handset, but where they are ambiently available in our most private sanctums, has snapped us out of our auto-update bubble. It has made us think harder about who or what we want listening to our conversations, whether spoken or gleaned through our online behaviour. Voice activated search has come at a time when data is a hot topic and there will be a period of questioning and adjustment before we realise we are trying to protect something we gave up the rights to long ago.

color-linkedin-128 color-twitter-128 color-link-128

Nicolas Carey, Design Lead, Potato

When designing an experience for a voice platform, of central importance is trustworthiness. We must consider the ethics of the human-machine relationship. It is also critical we remember the power we have when we design these products and services.

If we’re creating a product or service, we probably want a human to think, feel or do something. Humans must always be aware of whom, or what, they are speaking to and the relationship they develop with any voice product or service must not be exploited. It’s critical the human feels in control of a conversation as opposed to being guided to a commercial gain such as being persuaded to make a purchase.

The ‘uncanny valley’ – where innovation meets the challenge of ‘creepiness’ – has been well documented from an object/robot perspective but it is set to be better tested withvoice assistants much sooner.privacy of that data is a question that should be taking priority at the board tables – not just another “to-do” on a compliance officer’s list.

color-linkedin-128 color-twitter-128 color-link-128

Douglas Suvalle, Big Data Mobile and Web Executive, Glassbox

It’s predicted that half of all online queries will use voice search by 2020. It’s only a matter of time before the mouse, as we know it, no longer exists as voice recognition takes over.

The challenge for many brands will be digitising dictation, and building it into their AI strategy in order to optimise website performance. Furthermore, brands will need technologies to capture data that helps them understand users’ emotions and behaviours in order to determine how people engage with content. Only with this insight can businesses deliver slick web experiences for customers in the digital age.

color-linkedin-128 color-twitter-128 color-link-128 

Sylvia Jensen, VP EMEA marketing, Acquia

From a marketing point of view, voice-enabled tech is just another channel with which marketers can reach their audiences — just like websites, social media, email marketing and in-store experiences. The challenge for marketers, therefore, will be integrating the data that voice devices will collect with existing buyer profile data from those various other digital touchpoints to build a more complete picture of their customers.
In my view, the only way to improve the customer experience across channels is to assess the journey customers make via websites, social media, voice-activated tech and more — and to use this data to personalise the experience. The argument that consumers find personalisation creepy only exists because brands often don’t communicate what they do with customer data. But if brands open up their data collection practices for scrutiny to prove that they are being responsible with data, this creepiness will evaporate quickly.

color-linkedin-128 color-twitter-128 color-link-128