A new machine learning platform is shaking up the content marketing field and making it easier for companies to reach their target audiences.
Toronto-based technology company Atomic Reach has launched a new software product, Atomic AI, which allows “anyone publishing digital content to pinpoint the readability level of their audience,” according to the company, and helps them write appropriate content.
Using artificial intelligence, the platform delivers four times more Facebook conversations, three times the engagement and doubles pageviews, the company claims in a press release from Jan. 17.
“While completing research and development for [Atomic AI] over the last three years, we came to understand that there are 23 characteristics of language that influence an audience or reader’s decision,” Bradley Silver, founder and CEO of Atomic Reach, tells ITBusiness.ca. “The most important one is readability, because you can have something that looks crisp – meaning all the grammar and spelling is perfect – but if the message is either too complicated, or even too simple, for the target audience, the value creation is lost and they will go elsewhere for the information they are seeking.”
For example, he points to the difference in desired readability in the parenting and financial sectors.
“In the parenting industry, we’ve noticed that audiences look to educate themselves quite aggressively, so they may go from a general readability level to an advanced readability level very quickly because of that desire to understand what might be happening to them or their child,” Silver says. “Alternatively, in financial services, we see people prefer to read at a more general level because while health can be scary and complex, money is even scarier.”
Atomic AI includes a writing and editing platform that checks text documents in real time, giving the writer feedback to optimize style, structure and language for their target audience. It can also schedule social media posts for ideal times to increase engagement and visibility.
To develop the technology behind Atomic AI, Silver and his team had to address the complexities of teaching a machine to read language and understand the many different applications and meanings each word has in varied contexts.
“We use neural networks [a computational approach modeled after how human brains solve problems] to tag words by part of speech, then we use artificial intelligence to understand the context of how the word is being used. From there, we use AI to understand the context in which you’re using the word in the document,” he explains.
In terms of its own target audience, Silver says he expects the company’s clients to be any organization with a marketing communications department that is producing a fair amount of content, particularly focusing on sectors such as healthcare, digital publishing and financial services.
Atomic AI is now available to business of all sizes via mobile app, Google add-on, WordPress plugin and extensions for Google Chrome, Facebook, Twitter and LinkedIn, Silver says. Currently, it is only available in English, but with a number of different English dictionaries supported, the company sees potential applications across the globe, from North America and Europe to Australia.