While we are all aware photos have a massively dominating position on the web, we are also painfully aware of how little is done to tap into this massive source of data. What people photograph and share reveal a lot about what they can and will consume. Few startups offer solutions, mostly because the task is not simple. The massive volume, the separated ( and walled) locations, the current limitations of content recognition makes it extremely difficult to return meaningful and actionable results. However, some jump two feet in with the clear resolution to be the first to crack the code. Beautifeye is one of them and we sat down with founder and CEO Luca Marchesotti to learn more..
A little about you, what is your background ?
I fell in love with Computer Vision during my Ph.D. back in 2001, the early days of large-scale Machine Learning. Since then I never stopped being fascinated by images and visual content.
As a consequence of that, I spent almost a decade as a vision researcher in Xerox Labs, Grenoble and in 2014 I founded Beautifeye – Insights from images.
In between, I spent a lot of my time coding Computer Vision algorithms, writing scientific papers and filing 40+ patents related to imaging apps and solutions.
What does Beautifeye solve? How does it work ?
We are relieving the pains brands are currently experiencing while using traditional market research: slow results, sometimes not spontaneous, certainly disconnected from the real-time, multifaceted world of consumers.
Beautifeye flips the typical inquisitive approach of market research. Instead of asking questions to consumers, Beautifeye finds answers in the visual data spontaneously shared on the Internet.
To do that, we use a combination of Image and Text analysis: we call it Hybrid Recognition and it helps us to generate faster, more trusted and more reliable consumer insights. Unlike the many other platforms available in the market, Beautifeye puts at the center of the analysis visual content and images in particular.
How many images does Beautifeye analyze a day and one what platform (Instagram, Tumblr, Pinterest, Snapchat, Flickr,?)
Beautifeye crunches millions of images from a variety of sources: at the moment we are really focused on the most engaging visual social networks like Instagram, Tumblr or Twitter. But Beautifeye can also operate with data originated from an in-house video repository of a customer.
So I find pictures of my brand on social media, then what ?
There’s a number of use cases where Beautifeye hybrid recognition technology can be used: you are a brand and you want to organize a focus group, a consumers panel or you want to take advantage of more complex research tools like Need State Mapping.
Let’s take focus groups for instance. With Beautifeye you have the ability to deliver insights and recommendations in a matter of days instead of weeks – the typical timeframe of a focus group.
How do we do that ? Firstly, we extract a LOT of low-level signals with several image and text detectors: scene classification, face analysis, logo detection, text categorization and clustering, named entity recognition, textual topic classification. Secondly, we build a very high dimensional data superset. Finally, Beautifeye applies a unique workflow to distil marketing insights from the raw signals in the superset.
Can brands use/license the images they discover?
Is it more a marketing tool or an analytics platform?
Beautifeye’s users go beyond social media listening and transition towards a fully fledged market research engine. Several of our users are marketers seeking for inspiration and quick consumer insights. Despite that, the core of our customer base is represented by consumer analysts and insight managers.
In this sense, I would say that Beautifeye is more an analytics platform than a marketing tool.
How can a brand use Beautifeye to increase sales of their products?
Beautifeye supports brands in their mission to find golden insights about people’s behavior. What I am talking about here it is a very specific kind of insight: for instance, consumer discovery that optimizes your brand positioning in a fast growing but remote market. Or insights about a specific way to serve food particularly appreciated by your highest margin audience. Or a new taste that boosts your sales locally and generalizes across regions.
Who is your typical customer?
We are primarily working with large multinationals and global brands. For the moment, we are focusing on three main verticals: Food and Drinks, Tourism and Cosmetics. Having said that, we have an increasing number of agencies taking advantage of our SaaS platform and API to serve a larger number of markets.
What interesting patterns have you seen emerge ( like beers are more often consumed on beaches or sandals are more often worn in cities)
When it comes to food, for instance, one of the fascinating things it is the diversity and the creativity that consumers can exhibit in different markets. We recently studied how burgers are prepared and consumed across the UK and we were overwhelmed by the variety of food pairings and the serving styles .
Especially in London, the #filtyburger with all kinds of different meats and is now establishing as a trend among fit hipsters to boost their protein intake as well as relieving the pain of a long session at the gym with a tasty treat!
What would you like to see Beautifeye offer that technology cannot yet deliver
Busy marketers and analysts need better and faster ways to harvest insights. We are working hard to streamline the interaction workflows with our users: I envisage a future where the Beautifeye chatbot interacts with our users and helps them to further reduce time to insight.
Photo by left-hand
Author: Paul Melcher
Paul Melcher is the founder of Kaptur. He is an entrepreneur, advisor, and consultant with a rich background in visual tech, content licensing, business strategy, and technology with more than 20 years experience in developing world-renowned photo based companies with already two successful exits.