Will that video go viral?
Building software to predict engagement potential for online videos key to success
many elements of the scoring engine will need to be built from the ground up. This work requires a huge amount of resources. Silver’s current development team ranges from three to seven employees and contractors as needed, and he’s hoping to expand staff this fall to fill the gap. But such aggressive spending on research and development means the company will need a capital injection in the first quarter of 2014 to finish the work and bring the video-scoring product to market.
Luckily, Silver says he’s likely to get that injection from investors, and is convinced that finding his first clients will be easy once his software can aid them in better engaging customers through both text and video.
“Today, over 70 per cent of marketers are using video as part of their content marketing strategies,” adds Pulizzi. “The bad news is that most video is not very engaging and extremely self-serving.”
Giving companies the ability to change that, says Silver, is what will make his software a hit. He just has to get it working first.
"Tuning the engine,” he says, “remains a big challenge.” Bradley Silver knows whether your content will be successful.
He’s the founder of Atomic Reach, a Toronto startup with a “content scoring engine” that can predict the success of an article or video before it’s published online. Silver says the engine lets companies know whether content will resonate with an audience, how to improve it and how it engages customers before it’s published.
The software is a potential game changer for the content marketing industry. But while Atomic Reach’s program works well with written content, Silver has been unable to get the algorithms behind it predicting results for video.
“This could open a universe of potential,” he says, adding that the company has a lot of work to do before that universe is open to them.
Silver is a serial entrepreneur who founded a business protecting Fortune 1000 institutions from online fraud in 2001, and one of the world’s first social media monitoring companies in 2003. He launched Atomic Reach two years ago, at a time when brands were starting to notice the importance of engaging audiences through content.
The company started with a program that enabled companies to find and publish online content relevant to their audiences. But Silver quickly noticed it was difficult to tell what content would actually engage customers. He couldn’t find any reliable prediction software, so the business built its own.
That titular software works like this: after creating a profile on Atomic Reach’s website, publishers input individual links or RSS feeds to their content, which the engine’s algorithms analyze using 15 measures (ranging from paragraph density to language sophistication) of quality assessment. It then provides a rating out of 100 — the higher the number, the more engaging content will be to a particular audience — and also notifies publishers of spelling and grammar mistakes and faulty links.
It’s a powerful idea that caught on quickly, helping the company to raise $2.25 million from several venture funds and private investors. Approximately 20 businesses and 200 bloggers — they get to use the program for free in an effort to boost grassroots support — already use Atomic Reach, and Silver is in negotiations to sell his services to an array of Fortune 500 companies in Canada and the United States as well. But whether or not the content marketing industry as a whole will embrace the new tool with open arms remains to be seen.
“You’re disrupting an existing workflow, which is a challenge,” he explains. “When you’re asked to re- think how you approach a specific task, it can be uncomfortable. You’re often greeted with a certain level of trepidation.” Joe Pulizzi, founder of Content Marketing Institute, a content marketing consultancy based in Cleveland, is skeptical that the software will work. “I don’t know if I would ever say you can predict success of content,” he says. “Maybe they found the secret sauce, but even if you know everything about your audience and their pain points and you have the right contributor, sometimes it works and sometimes it doesn’t.” Silver believes he has in fact found the right recipe, and says this is evidenced by the results his clients have seen. By using his engine to improve content, they normally see a 5 to 12 per cent increase in audience engagement, which can take the form of comments, social media interaction or page views. Carrying the concept over to video, however, an increasingly important medium for brands and publishers, could be key to finding widespread adoption, but the process has been difficult. While metadata such as wording, paragraph density and use of quotation marks is easy to identify in text, determining the sophistication and effectiveness of video requires a much more complicated analysis. Elements such as audio and video quality, dialogue, and editing are significantly harder for an automated tool to dissect. Furthermore, the software needs several unique formulas for every different format video can take, whether it’s interviews, ads, movie previews or even slideshow presentations. To make the process easier, the company is using existing services, such as transcription programs that can dissect dialogue and measurement tools to analyze production quality, to help the process along. But