Key Takeaway
Our online breadcrumbs are becoming increasingly unstructured. We leave trails of external data in the form of emojis, images, GIFs, voice, video and more, as well as text. Google’s recent acquisition of GIF search tool Tenor underlines its confidence in the future of GIF-sharing and other new forms of media as the way forward in online conversations. As a result, the need for tools that can interpret and analyze this data is reaching an all-time high.
Have you ever tried to have an entire conversation using only emoticons? Wall Street Journal personal technology columnist Joanna Stern fell in love back in 2015 as emoji-based conversation was really ramping up.
Thanks to technology, today we can choose to express our emotions in a myriad of ways. When words fail, an emoji won’t cut it, and a meme is not quite right, many turn to the GIF. While competitor Giphy is well-backed with $150M in funding, it’s the Tenor GIF keyboard that caught the eye of search giant Google, looking at its mission of becoming the ultimate “emotional graph.”
Boasting 400M+ GIF searches daily, the startup aims to become an emotion-based search engine, “matching people’s thoughts and emotions with a digital object — the GIF.” CEO David McIntosh tells Forbes, “GIFs are a better way to convey thoughts, feelings and emotions at a time when attention has shrunk.”
The tool was recently selected by Samsung as its preferred GIF partner.
Google acquires Tenor, gains data from its global Emotional Graph
Tenor’s search engine is heavily powered by machine learning, with a proprietary Emotional Graph that maps common search terms to GIFs that capture every sentiment, as well as the relationships between each emotion. “The Emotional Graph maps more than five billion distinct human feelings and reactions, and evolves every day in response to news, memes, culture,” the company said in a recent press release.
Beyond its native app, the tool is built into smartphone keyboards and accessible via most major social and messaging services.
Consumer research firm Magid Advisors claim that mobile shares of GIFs could grow by 1,000%, generating $1 billion or $2 billion in total annual advertising revenue. This week, Google put their money where their unstructured data analysis is, announcing acquisition of Tenor for an undisclosed amount.
In a blog post on the topic, Google’s Cathy Edwards, Director of Engineering for Google Images, spoke about changing consumer behavior as more and more look to search results via Google Images over text. “Most people now use Google Images to find more information about a topic, and to help them communicate and express themselves—case in point, we see millions of searches for GIFs every day.” The brand intends to use Tenor’s tool to improve their GIF search capabilities.
“With Tenor Insights, we’re making it easy for marketers to explore the Emotional Graph and better connect with consumers around a new dimension – emotion.”
– Tenor CEO David McIntosh
Unstructured data and the need for advanced AI
Research firm Gartner estimates that more than 80% of enterprise data today is unstructured. Unstructured data is data that is unorganized and raw. It can be non-textual or textual and can include emails, PDFs, social media, images, videos, GIFs, etc. In other words, the more GIFs we send, the more unstructured chaos we create.
Luckily, there is an AI for that.
Startup Gyfcat is looking to improve the quality of your GIFs and analyze text within GIF frames. The team is using AI and machine learning to better interpret text, find the original videos from which GIFs have been extracted to improve the image quality, and make your GIFs more discoverable when searching.
Other tools like Affectiva claim to provide emotion AI by analyzing human emotions in face and voice. Their deep learning methods train algorithms to recognize particular emotions.
As our avenues for emotional expression continue to evolve, so too will those tools that help us search, share and analyze the billions of conversations that continue to take place online. Brands that aren’t looking out at insights from these online conversations will soon find themselves at a significant disadvantage.