Hey Siri, how are news publishers tackling voice devices?

Last week I posed the question ‘How can publishers shine on a platform with no interface?’

This week I got the opportunity to ask this very question to a room full of media innovators at Newsgeist in Phoenix. It turns out that people have been thinking about content for Amazon Echo, Google Home and similar devices. Here is what I learnt:

How do you handle discovery with no visualisation?

To kick off with a commerce example, which I think we can learn from, if you ask a device: ‘order me a pizza to be delivered’, it doesn’t have enough information to complete the command: which pizza do you want? Where from? You’re unlikely to want a list of pizza places and then an entire menu read to you. But if this is handed off onto another screen to show you options to choose from, the experience can be neat.

News may have a two step approach too: simple questions from users will likely to be answered by the gateway software such as Google with search e.g. what are the top headlines today? can bring up the top headlines on Google News. If the user asks for more, they’ll need to go ‘off platform’, linking into publishers.

What type of content should publishers be creating?

So far the most popular applications of these home devices are (unsurprisingly): music, news and weather. There is a huge opportunity to capitalise on this as a new medium that is growing (by 2023 a whopping 50% of computer interactions are expected to be via voice).

Looking through all the notes from the session it seems that content falls into three categories:

1. Top of mind/specific questions. Examples here are the Fact Checker, already built by Duke’s Reporter Lab (and which is building out it’s content with partners such as the AP); and has opportunities for local news such as children waking up to heavy snow asking “is my school open today”?
2. Regular interactions. Examples include the Flash Briefing on Alexa where a user listens to their morning, or evening news, as part of their daily routine.
3. Thoughtful/intentional listening. For example, ‘Alexa play the latest episode of ‘Reply All’ by Gimlet media

Text to speech is improving — yay — so there may be opportunities to use content that already exists. Editors will need to think about whether each piece is conversational enough for it to work on audio plus figure out how to deal with emojis (try it just for fun to see how Siri handles it now). However for the long term content works best when it’s adapted specifically for the platform it resides on. Media companies should start thinking about how they store and tag all their content now so that it can be resurfaced in the future as the opportunity arises.

One advantage of text based content is that people can browse and skim read, perhaps there is a way we can help consumers do this using audio?

Equally many people are multi tasking when they are listening: brushing their teeth, getting dressed, cooking dinner. So the cadence may need to be a little slower to adjust for this.

Predictability is important. Users will need to know what content is available when so consider regular updates. And don’t forget weekend! These devices live in people’s homes so they will want weekend content too.

The devices on the market give us the possibility to interact with the audio.

Simple commands such as ‘next’ ‘skip’ and ‘repeat’ work and the interactive nature of this medium means collecting data by asking users to answer a short survey could easily be woven in, in the same way Google Surveys work now. Thinking ahead, as NLP and machine learning improve, maybe there is a way we can create a chatbot type experience on these devices.

So how can individual brands shine through?

Actually this is a lot more simple that I’d thought. Here are a few ways that brand can be reinforced in content:

• The brand name can be mentioned as frequently, or not, as your narrator and audience can bear.
• Most companies that use audio — radio, tv & podcasts — already use original opening sequences to recognise a brand. For an example just listen to The Daily which has a unique, recognisable identity and has quickly risen to success.
• Bespoke sounds & auto cues will reinforce the brand experience

What about ethics and privacy?

The devices will recognise individual voices but how do you know if children are in the room and sensitive content should be screened?

One person told the room that on one device their music list has had all the swear words bleeped out. How does a provider choose which words are chosen? Could this be considered censorship?

And what about privacy? The answers in the room seemed to conclude ‘that ship has sailed’…. which gave me a cold shiver.

By default Alexa and her competitors are always on. Technically they don’t react until the appropriate trigger word, such as Alexa or Hey Siri, is mentioned. However I’ve heard a great deal of anecdotal evidence (including this blog post) of the sudden appearance of advertising following a (private) conversation of a product or service at home. Aside from Apple, who have taken a strong stance on user data, I don’t trust other platforms. Especially not with casual and private conversations at home (which is why my Alexa Dot sits in my bathroom and not the living room!)

The connected device ecosystem

In the pizza analogy above there is handoff to a screen. In the case of Siri it’s to a mobile phone but it could also be a TV or other device with screen. There are multiple examples where this may happen so we’ll need to think about continued and smooth transitions from one to the other.

Someone bought up the idea of a story developing over multiple platforms and devices throughout a day. For example in the morning you may be rushing and just want bite sized news but at lunchtime you may want to watch a 15min update whilst eating a sandwich, and in the evening get an in depth run down of the days events. At the moment we treat each platform in isolation but perhaps we have an opportunity to give users a richer experience.

And are there opportunities to have individual devices? For example a New York Times device that is given away to their premium subscribers with features already built in?

The all important money question

So how will all this be sustained? I haven’t seen any yet seen any paid content and not sure it would be different what what we are seeing on other digital platforms: users only pay for niche content. Some people in the room argued that content should be paid from the offset. Others thought a freemium model could work. Perhaps this the ideal use case for micro-payments (finally)?

This doesn’t seem to be a medium for traditional advertising (cue audience cheers). However podcasting has proven out the value of native advertising which can, relatively seamlessly, be sewn into content (cue publisher cheers).

Conclusion

And so, I walked away feeling buoyant that my Echo Dot could be filled with exciting content that may just be profitable for news. The big question is: how much control will the gatekeepers retain?

— — —

If you are a media exec looking for inspiration and practical takeaways around innovations that affect your business now, consider joining me on an innovation tour in California. At FORE:media we give unparalleled insight to technology pioneers. By connecting and collaborating in small intimate group you can ask the questions the really matter. Right now we have focus around all things video, monetisation and growth. Plus we’re enjoying the VR, AR and block chain revolutions first hand. More here: www.fore.media.