Fill in the form and the white paper will be sent to you automatically!
At Datlinq we are constantly working on ways to discover how places, consumers, products and brands interact. We do this because we know that this is of great importance to any company acting in the foodservice market. Especially when launching new products or if they want to grow their existing sales efficiently, it is very helpful to have insights that help find locations where their specific product would fit well. More specifically, access to rich profiles of foodservice locations is a very valuable asset in this quest.
The traditional way of building such location profiles is by manually collecting all sorts of characteristics, for instance by visiting or calling the outlet, or by searching for it online. Obviously, this requires a massive amount of manual labor including continuous updating. This simply becomes infeasible when expanding to multiple countries, or even most of Western Europe, like Datlinq does.
To overcome the issues of scalability, expense and maintenance, at Datlinq we leverage recent advancements in machine learning and natural language processing – more specifically a technique called word embedding – to automatically train semantic foodservice models from our data. These models (one for each country), represent a lot of knowledge about the relationships of all sorts of concepts in the foodservice universe, like brands, dishes, drinks, ingredients, cuisines, and flavours.
So, how exactly do we use these word embedding models to recommend foodservice locations? Well, if our model understands the relationships between the words in our foodservice data, it should also be able to understand the relationships between foodservice locations. Consequently, we can use this technology to make sure that our customers are, either with marketing or sales, always targeting locations that have a good fit with their specific product. Already numerous successful projects have shown that this way of working significantly improves conversion ratio’s across channels, thus increasing revenue and margin.
Want to understand how this could work for your company? Contact us!
Want to know more? Our lead data scientist Martijn Spitters wrote a whitepaper on this subject! Fill in the form!