Newsrooms should take the voice ecosystem seriously
The voice ecosystem has been drastically expanding in recent years. While smart speakers remain largely a North American phenomenon, the technology is evolving fast and will soon have implications for newsrooms across the world. What promise does voice AI hold for editors? Is there a sustainable business model for the voice economy?
GEN spoke with Matt McAlister, co-Founder and CEO of Kaleida Networks that recently unveiled its News Assistant Platform. Matt sees a lot of potential in voice AI that connects users to news organisations. While it is still too early to tell, he says ‘voice technologies are making a lot of things possible, and we need people thinking beyond platform distribution as a model for news’. You can download the new Kaleida report here.
GEN: You write in your report that personal assistants / smart speakers ‘have proven to be good at doing what the tech platforms do best. They provide data. Fast. Personalised. News orgs are not designed for that use case. They value stories over data. They prefer veracity to speed. And they talk to readers equally, not individually’. Are news organisations condemned to be the ‘hostages’ of Amazon, Google or Apple?
Matt McAlister: No, I don’t believe that at all. Distribution has moved to the platforms, but there is a lot of evidence that people value quality news brands. They want insight about what’s happening in the world, and news orgs are uniquely capable of serving that need. In terms of the voice ecosystem, it’s early days still. The current devices are not good at providing context or insight. That’s an opportunity.
According to you, ‘the bigger hurdle for news may not be distribution after all. The more pressing issue might be the gap in the market for insightful answers to more complicated questions’. Why do you think that news organisations are not prepared to answer users’ questions?
News organisations are dancing around this issue too much, in my opinion. They respond to users in comments and create new kinds of community services and engage on social platforms. But instead of developing that as a core competency, it’s done by renegade teams inside the organisation. Those teams understand how important it is to be part of the wider conversation. As soon as the wider organisation gets serious about answering questions, tons of digital opportunities will open up.
Voice platforms require comprehensive, well-structured and accessible content to answer end-user questions. Why is it so difficult for a radio or a newspaper to do it?
I think this is hard for anyone to get right. We did some tests with users and were surprised by the variety and complexity of the questions. Take this question, for example, ‘Why is Boris Johnson the favourite to become the UK’s next Prime Minister?’ There’s no platform out there that can answer that unless an article has already been written about that question. However, if you asked that question in the presence of editors and reporters in a newsroom you would get the context and insight that paints a very clear picture. It’s not about creating comprehensive or structured content. It’s about knowing what people want to know and responding to them.
You consider in your report that an ‘answer marketplace’ could be very effective and you take the example of an independent non-profit foundation, such as Wikipedia, which provides ready-made answers to users. How can the voice ecosystem be based on an online encyclopedia? How can Wikipedia open doors to third parties and news organisations?
I suggest in the report that a new ‘answer marketplace’ should be formed, perhaps collaboratively by news orgs. Wikipedia is very good at answering fact-based questions, but it’s not good at answering context and insight about what’s happening on a day to day basis. A new marketplace optimised for connecting people to news organisations would be better than Wikipedia. Perhaps a new Q&A marketplace could be formed with support from or in collaboration with Wikipedia, but Q&A is a fundamentally different service and capability from an encyclopedia.
Will podcasts be successful on smart speakers? Are they an entry point for a more sophisticated experience with voice AI?
It’s unclear what will happen with podcasts on smart speakers. Personally, I think it’s a brilliant use of technology, but I also suspect that use case will remain a little bit fringe. The revenue isn’t really there, yet, and the uptake seems to be limited. It feels a bit like zines in the early 2000s.
The business model for news organisations / third parties of the voice economy is not mentioned in your report. How can news organisations make money with smart speakers?
Currently the options are very limited. We asked a few news organisations about this, and even the more advanced publishers were unimpressed with what they had found, so far.
However, if you think about the scale of the opportunity the revenue picture starts to look really interesting. Voice will probably be a standard interface for Internet-connected devices of all types, not just smart speakers and mobile phones. There is probably scope for all kinds of new devices. And if news organisations and other media entities got behind some new standards and technologies that support different devices then they could reset the game. They could influence distribution again. They could optimise the market for advertising in a way that works for everybody.
Again, it’s early days. Voice technologies are making a lot of things possible, and we need people thinking beyond platform distribution as a model for news.
Your prototype / platform is based on Google tools: is it a trap or an opportunity for news organisations?
Google and Amazon both have increasingly impressive tools. It’s actually incredible what they can do. Some of those tools will create lock-in eventually, but the switching costs aren’t too bad still. So, my recommendation is to use their tools for now, in order to understand how everything works with an eye toward supporting industry standards and open source versions of competing tools.
Edited by Ana Lomtadze
Matt McAlister is the co-Founder and CEO of Kaleida Networks, a media research and data company develops tools for media executives and news consumers. He is also the co-Founder of Publish.org. Previously he worked at the Guardian, developing media businesses and new technologies for journalism. He has been involved in various aspects of the digital publishing ecosystem since 1994 — leading digital arms of print businesses, building platform services at large media companies and creating new digital businesses.
Download the new Kaleida report about its News Assistant Platform here.
