Amazon Polly: Breathing New Life into TTS
Mar 10, 2020
Voice is often an underrated tool in a brand’s locker when it comes to showcasing identity. After all, in the hustle and bustle of the modern world, we don’t always have time to stop and watch an advert, read a billboard, or try out a product – but aural advertising manages to strike the perfect balance between passive consumption and emotional connection, making it extremely powerful. This is best illustrated right now in the entertainment industry, with shows like The Voice and The Masked Singer growing to become runaway successes on both sides of the Atlantic – popularising the notion of identity and success being associated solely with audio cues. This idea is important to the shows, but even more important to brands – and a new development from Amazon has made it easier than ever for companies to showcase their identity through the power of voice.
The Amazon Polly is a Text-to-Speech (TTS) service that uses advanced learning technologies to synthesize speech, allowing innovators to ‘create applications that talk, and build entirely new categories of speech-enabled products’. That sounds exciting enough on its own, but Amazon’s latest advancements in Neural TTS (NTTS) are set to really change the game.
Polly users in the UK currently get to choose from one of three voices for their programmes – either Brian, Amy or Emma – and realistically, there isn’t much variety between the options. This is a problem for a brand with ambition, as reducing things to a simple ‘one of three’ choice severely limits any opportunity to carve out a strong brand identity. But NTTS has enabled Amazon to roll out a brand-new feature called ‘Polly Brand Voices’, giving companies the ability to create a custom voice specific to their brand. Alongside the big launch news, KFC Canada provided examples of how they used Polly Brand Voice to recreate the voice of their world-famous founder and long standing mascot, Colonel Sanders. In an Alexa Skill that allows you to reorder food, KFC has allowed its customers to talk directly to the Colonel himself – simultaneously doubling down on the familiarity of the classic figurehead of the company, and producing dynamic spoken content on multiple platforms across the world.
On the surface, this is brilliant for audio branding, as well as brands in general – the rising popularity of custom voices over presets shows that companies are taking new strides to make their customers’ experiences as authentic as possible – but what does this new tech mean for the professional voiceover artist? Many brands are currently using voiceover artists or celebrities in their Alexa Skill audio content, and in what seems like only the latest case of AI taking over the jobs of real people, their positions could well become filled in the near future. On the other hand, whilst the Polly is undeniably an impressive advancement in TTS, the emotion and viscerality of the human voice isn’t something that can be replicated by a computer – at least not yet.
The warmth, intonation and passion of the human voice artist can’t be rivalled – which is why at PHMG, we pride ourselves on having built a portfolio of the finest voiceover artists in the industry. These talented individuals deliver a human quality that showcases the real side of your business. But whether computer-generated or otherwise, developments like those with Polly highlight the overall importance of having a strong voice that represents your brand – and how voice is the key to showcasing identity.