Wibbitz, an Israeli text-to-video startup, has raised an $8 million Series B round led by NantMobile, with participation by existing investors. The funds will be used to expand into the U.S. via a new New York City office, headed by CEO and co-founder Zohar Dayan, who brought me up to speed on the company yesterday.
Wibbitz uses natural language processing to quickly turn publishers' text articles into short videos. The process begins with Wibbitz's technology digesting the article via algorithms meant to emulate a human reader's behavior, identifying key people, main points and the theme, resulting in a summary of about 20% of the full article.
Wibbitz then takes that information and repackages it into a script for the video. Next, Wibbitz taps into a catalog of licensed content from Getty, Reuters and others to add visual components as well as lower thirds, splash screens, etc. All of this takes approximately 10 seconds.
At this point Wibbitz can either provide an automated voiceover to read the script, or farm out the voiceover work to its 24/7 network of voiceover artists, which Zohar said the vast majority of publishers choose to do. Publishers can decide on any or all of four languages, English, Spanish, French or German, as well as options like male/female, etc. These voiceovers typically take 10 minutes to complete and then the video is ready to be published.
Publishers also have access to a Control Room where they can monitor videos being created and modify them accordingly.
Below is an example of a Wibbitz video, using its own press release today as the text article.
To date, Wibbitz has focused mainly in EMEA and Asia, with publisher customers such as Times of India, FOCUS Online, Times of Israel, Costa Rica Today and others. These publishers typically embed the Wibbitz video summaries with the original text articles. Zohar said in the past 30 days, embeds have occurred on over 80 million pages.
Wibbitz has two big value propositions to publishers: easily and inexpensively create high-quality videos and generate premium pre-roll video ad inventory. Since video CPMs are higher than display CPMs, substituting or augmenting text articles with video views is compelling for publishers. In addition, the videos can be distributed through social channels to drive incremental views and monetization, especially on mobile.
There's also an interesting opportunity for customized ads to be programmatically inserted based on the content. For example, a weather report indicating snow in a certain geographic area could be a data signal to trigger an ad for snow shovels from a local retailer.
Wibbitz's current business model is not to charge for the platform, but rather to do a revenue share with publishers, based on it selling the ads mainly through ad networks and exchanges. No doubt as larger publishers in the U.S. which have their own video sales teams adopt Wibbitz the model will shift to usage-based or licensing.