Digital Retail Innovations Q&A – Pixoneye
Following the launch of the Digital Retail Innovations 2017 Report (sponsored by Webloyalty) a series of Q&A’s with the leading figures behind some of the key entries in the report are being published.
Here is a Q&A with Ofri Ben Porat, co-founder & CEO, Pixoneye, which is placed at number 33 in the report.
1. How did the original idea come about?
While working as the Senior Marketing Advisor to the Minister of Tourism in Israel, I had to come up with a way to segment and target the two most substantial tourist communities; the LGBT and the Evangelical Christian community – without accidentally mixing our in-app offers between the two.
Realising that today’s user segmentation relies on basic and obsolete third-party data (including browsing patterns, purchase history, geo-location, and social media listening), I paired up with Nadav Israel (Pixoneye CTO and co-founder) in order to find a way to analyse an untapped resource to increase human understanding on mobile devices better than ever before; by automatically and anonymously analysing user’s personal photo and video galleries on their mobile devices.
2. What has been the level of interest/take-up of the service and what are the key ways retailers can/are using it?
As developing the innovative and unique product was the company’s primary focus in the past two years, Pixoneye began commercialising the product merely six months ago, soon after the commercial team had been set up In London in December 2016. Since then we have been featured in various lists of mention, and most recently- have been nominated for the BIMA Advance Award for Start-up of the Year.
We are currently in different stages of integration with a range of different clients – from Prof of Concept to paying customers, over various industries including financial services, utilities, and e-commerce to name a few. For example, for a fashion app looking to offer the most relevant inventory to their customers – by using image understanding – out of 150 different characteristics we can detect what the users’ main fashion style is.
By using their photo gallery as a dataset over a timeline, we can create contextual user behaviour understanding – to see if they have a summer-weather, beach vacation coming up (as their galleries show us that in July the past two years that was their preferred destination), and offer an in-app discount on the new summer apparel just in time for their upcoming vacation – thus increasing user engagement, in-app user experience, and revenue for our client.
3. What about competition in this area?
With some of the leading minds in computer vision and deep learning, Pixoneye has developed unique technology which is a clear and evident outlier from the current market. Our on-device deep learning and computer vision technology has been 2.5 years in the making. Today, we can identify intent at the same accuracy as Google but at over 30 times less the classifier weight (1.5MB vs 50MB at Google). Our proprietary dataset is unrivalled by other players such as RichRelevance or Facebook, as no other company has successfully tapped into users’ mobile photo galleries. We are also unique in that we provide both a DMP platform and collect the data first hand, with a product offering that can be adapted to the needs of businesses in a wide range of verticals, from brands to publishers to research companies.
4. How has the solution been developed / advanced over time?
When founded in January 2015, we were initially going to build a smart photo gallery app that allows you to search for photos based on their content. Instead, we decided to create an SDK and become tech enablers – as 99% of apps already have access to end-user photo galleries, and Pixoneye has been developing this technology ever since.
We had a second major pivot early this year when we repositioned ourselves from an AdTech company to a Core Tech company. Initially, we were using our technology to enable publishers to provide targeted advertisements within their apps. Although very successful – we were generating a x10 increase in engagement – we left AdTech and now offer full marketing architecture and automation solutions, improving user understanding for clients in the financial services, insurance and e-commerce sectors.
5. Have there been any surprises along the way?
With our product so well placed for retailers and advertisers, we did not expect such support from Telco’s. We are only the second company to have ever received follow-up strategic investment from Wayra, Telefonica’s start-up accelerator, and we are now in talks with O2.
Similarly, we did not anticipate so much interest from financial services firms. We have recently received strategic investment from Santander Innoventures, and have signed with a major European retail bank.
We also did not anticipate such large traction in Asian markets. We have recently signed a contract with a leading multinational conglomerate and have attended the Hey Startups event in Seoul, South Korea, and the Mobile World Congress in Shanghai, China. We went out there with little expectations but were positively surprised by the huge amount of interest expressed from these markets.
6. What can we expect from the business in the future?
Currently our main focus is capitalising on all the traction we have received, expanding into new markets and developing the next generation of our deep learning image recognition algorithm. In the future, we want to see Pixoneye on every device, and we look forward to perfecting our products by running millions of pictures through our system to improve our accuracy level for even more ground-breaking results.
7. Do you have any recommendations for entries in the next Digital Retail Innovations report?
We did not apply for an entry in this year’s Digital Retail Innovations report, which has been the same for all the other entries in the report. As such, we applaud the report’s commitment and skill in finding unique and ground-breaking businesses both nationally and internationally.
The Digital Retail Innovations report commends those in the industry who dare to push boundaries and motivates us to continue to search for new innovations that transform industries and redefine the way that consumers interact with technology.
By the time the next report comes out I’m sure we will experience even more radical technological transformations, so it is hard to say. As an outlining theme – people, and bettering inter-human communication and experience should remain at the heart of what we do, and what an innovative company should strive to achieve.
Webloyalty is a leading reward programme provider working with major online retailers to help them build stronger, more profitable relationships with their customers. Through our membership programmes we help our online retail partners’ customers save hundreds of pounds a year while providing the partner with an additional revenue stream. As well as incentivising customers to make repeat purchases at the partner’s site, members can also earn cashback and get great deals on everything from fashion to electronics to travel, at hundreds of top online stores.