Beyond HumanBig PictureCatalystsConnected WorldExchangeMarketing MixNew MoneyNew SchoolPeople SciencePulse

Using big data to deliver tasty snacks

hero image hero image
Photo credit:


Founded in 2008 in the United Kingdom, Graze.com launched in the United States in December 2013 and has already shipped more than 1 million graze boxes filled with nutritious snacks. Anthony Fletcher, its CEO explains how Graze’s use of big data helped it become such a success in the US.

Anthony Fletcher, CEO of graze.comSnacking reinvented

There is a fundamental shift occurring in the way Americans eat. More and more people are replacing traditional meals with snacks. Snacking is convenient and has become an important part of today’s busy urban lifestyle. The ability to have snacks delivered directly to their desks to satisfy immediate cravings is something that appeals to busy workers. At the same time, consumers have become more health conscious and expect their snacks to be more nutritious than the chocolate bars and chips one would typically find in a nearby vending machine. Nutritious snacking options (e.g. low calorie, high protein and nourishing) resonate with these consumers.

Our customers, who we call grazers, are looking for convenient nutritious snacking options and are open to trying new things. Most of them are urban early adop- ters that are tech savvy. From the extensive data we collect we detect that the majority of our rapidly growing U.S. customer base is located in urban centers like New York. We are see- ing strong growth in the urban office segment in particular, especially in New York. Our offer has a high appeal for professionals that are time constricted, but still want to eat healthily. We estimate that the size of the U.S. e-commerce snacking market to be about $500M.

Big data meets Darwin

In the U.S., graze.com currently offers more than 90 different snacks and each graze box delivered contains four individually packaged and portioned snack combinations that are personalised to each user’s taste and dietary specifications, including gluten free options. We customise each box by using a unique taste algorithm called DARWIN (algorithm rating what ingredient’s next).

DARWIN determines how snack portions are sorted, how boxes are labeled and most importantly how each box is tailored to each customer based on customer feedback and behavior. We get more than 15,000 new customer ratings per hour and uses this data to create, package and market new innovative products. This means our customers don’t specifically choose their snacks, but rather are sent snacks based on their previous product ratings and collected data. This data enables us to do what other companies shy away from, which is to take risks and surprise grazers with a product selection they don’t personally choose but might not try otherwise. Our algorithm is so sophisticated that it can even capture quite abstract ‘clusters’ of data to spot whether a grazer is lactose intolerant by their behavior when selecting product categories.

The big data approach also allows us to react quickly to our customer’s taste. Since the initial launch in the U.S. last December, we have already redeveloped approximately 30 percent of our product range to suit our U.S. grazers. Our data indicated that the U.S. customers did not like certain products that our UK customers love – our joined up approach allows us to react quickly and adapt to these local preferences.

David vs. Goliath

It is proven that companies, no matter the size, who intelligently utilise their big data are becoming more competitive and increasingly more successful. A small tech-enabled business, like graze.com, can outmaneuver industry giants by being fast, innovative and closer to our customer’s needs. Continually improving and innovating DARWIN is important to further advancing our business. Our data scientists are looking at improvements like capturing additional clusters so we can determine our customers’ preferences even more precisely.

For entry into the U.S. market, we also developed a self-learning logistics solution which we named ‘the brain.’ This system crunches millions of historical and live data points and helps us determine the best and most efficient route direct-to-grazers utilising FedEx and the U.S. Postal Service. The system is already proving to be incredibly successful and during this past winter 99 percent of boxes were delivered despite the adverse weather conditions in the U.S. As a result, larger more established companies are approaching us for inspiration on how better to utilise their data to improve their systems.

We recently announced our commitment to spend more than $75M on the U.S. business over the next two years. We are focused on finding new avenues for growth and using our one-of-a-kind logistics solution to deliver a high-quality, reliable proposition to our grazers at a price point that’s appealing. Owning the whole integrated supply chain enables us to use grazers’ feedback to tailor product offerings, ensure the right products are getting to the right markets, to be flexible and fast, and to be profitable.

Anthony Fletcher, CEO of Graze.com