SpeechKit is competing in the ‘New ways to reach your audiences’ battle of our Startups for News competition to be in with a chance to pitch at the GEN Summit 2018. SpeechKit uses artificial intelligence to allow publishers to automatically generate podcasts from their written news articles and monetise them. If you would like to vote for SpeechKit, you can do so here.
What did you do prior to launching SpeechKit?
The idea to put news into audio comes from co-founder James’ first company — Clippet — an app that summarises the main news issues of the day into quick-fire human-read audio clips.
After having sold his first company, co-founder Patrick, took a python machine learning course on Udacity. At the same time, he was faced with an ever-growing list of unread news articles. He therefore had the idea of using artificial intelligence to turn written news stories into audio, allowing people to listen to articles instead of reading them.
Along with co-founder Kostas, a digital nomad and engineer fresh from Silicon Valley, the team got together to bring audio to the news in a way that is fast, economical, and scalable for news publishers.
What problem are you trying to address for newsrooms?
Faster news cycles mean that many people can’t find the time to read beyond the headlines. Mobile phone penetration is at an all time high, but for many, reading on mobile can be difficult. Wearable or voice-activated devices have little or no screens, making reading on them practically impossible for most people.
Audio, on the other hand, is less demanding, more convenient, and perfect for screen-less devices. Never before have we seen such volumes of audio being created, streamed, and shared. But many news publishers are struggling to adopt this medium. Producing new audio content, like podcasts, can be time consuming and expensive. Furthermore, producing human-read audio for all news stories is pretty much impossible.
How are you attempting to solve the problems described above?
SpeechKit helps publishers turn their news articles into podcasts using artificial intelligence, and monetise them with audio ads. SpeechKit works in over 20 languages and dialects.
The audio takes seconds to produce, and once it is ready, news readers can use the audio player that sits beneath the headline to listen to a news story. Increasing both engagement and time spent.
Using SpeechKit, news publishers can also monetise the audio clips with brand sponsorships in the form of short audio advertisements. This can increase publisher revenue without significant investment in producing new content.
What sets you apart from your competitors? List three elements.
- Better return on investment — SpeechKit requires minimal investment into audio production compared to traditional podcasts or human-read audio news stories. It can also be monetised far more easily.
- Better distribution than podcasts — the SpeechKit audio player augments with online new stories. This means that there is no need to grow distribution, as is the case for traditional podcasts, because it benefits from the news publishers existing news traffic.
- Better quality — SpeechKit offers better pronunciation than just using text-to-speech APIs like Amazon Polly or Google WaveNet, because we do proprietary pre-processing and SSML text enrichment prior to speech synthesis.
What is your business model?
We make money by charging news publishers a monthly fee for the service and splitting audio sponsorship fees with them.
What are your next steps?
Apart from continuing to build more relationships with publishers and brands from around the world, we’re really excited about the rise in voice-activated devices from Amazon, Google, and Apple. We want to make it easier for news publishers to create and manage their content across all of these devices.