Data food in action: using data to enhance efficiency and reliability with Marleenkookt
If you follow me on LinkedIn or Twitter, by now you’ll have noticed I’m nature driven and data obsessed. A lot of the work I’m doing these days at JADS — through hackathons, our Data Entrepreneurship in Action (DEiA) program, and recently, development of my proof of concept lab — has a focus on agriculture, farming and food.
DEiA students have been working on real world case studies, and the latest business to benefit from their equally data obsessed minds is Marleenkookt — healthy meals made and delivered to your door. Marleenkookt are looking to innovate their business, and with it, their short supply chain.
Short supply chains involve as few intermediaries as possible, aiming to connect local growers with local consumers. While I was at a 36-hour FarmHack hackathon last Friday (working to change the future of food, one prototype at a time!), my sidekick Jai Morton caught up with Joris Keijzer of Marleenkookt. Read on to hear how they’re learning to harness their data, further contributing to the future of food.
With a quantitative background in organisational research, you’ve always been interested in data. But what drew you to data science as a way to grow and explore Marleenkookt?
“An article that I read about a year ago about data science peaked my interest. It said AI was quickly becoming a part of the standard strategic toolbox for companies, similar to how spreadsheets in the 80/90’s revolutionized the way businesses used data. After reading this, I remember not wanting to be left behind and started exploring how we could use data science,” he says.
“Data science isn’t just for big corporations.”
But it would seem Joris hasn’t been left behind. In fact, he could be ahead of the curve, especially for a small business.
Data science isn’t just for big corporations. It’s accessible for everyone — it’s about knowing where to find data in your company and how to harness it. Marleenkookt is doing just that, taking steps to integrate data into their business toolkit.
Do you think the idea of AI as part of the new standard toolbox is catching on easily and quickly, or is there confusion or resistance?
“No, I don’t think it’s a standard part of the toolkit yet, but I wouldn’t say there is resistance. I think many of those in senior or middle management now have not grown up with the idea of data. So they may question or challenge its relevance, but I think if there is a lag in uptake, it’s because they’re yet to realise the potential of big data and how to use it,” he says.
“I think a change in perception and way of thinking is coming.”
“I think a change in perception and way of thinking is coming. It’s about keeping informed on current trends and also having the guidance and education to move forward.”
With the DEiA program and JADS students, you’re working to optimise Marleenkookt delivery routes, and predicting customer’s menu choices. How are you leveraging the data to do this?
“We’ve been working with JADS for the last 3–4 weeks, as one of the DEiA students’ case studies. It’s been fascinating to work with the students, as they have fresh mindsets and this sort of work comes to them naturally. Leveraging the data starts with knowing what you need. To them, it seems obvious what sort of data would be helpful and in which format, and how they might use it. Whereas I can see why it is useful, but wouldn’t have thought of it myself.”
“We’re looking at enhancing efficiency and reliability, as our supply chain is very ‘just in time’. People order in the morning before 11am and choose a half hour window the same evening to receive delivery. This has led us to take a closer look at customer ordering patterns and behaviours.
“These insights will help us to more accurately predict how many meals we can sell on any given day. The more accurate the forecast, the more accurate the produce order, meaning less food waste. We can also better manage staffing needs with this data at hand, knowing when busiest preparation and delivery times are. And on-time delivery means upholding customer satisfaction.”
Marleenkookt are also aiming to use data to guide their recipes.
“We think the varieties and combinations we offer now are good, but we don’t know if certain groups of customers aren’t enjoying them or aren’t being accounted for. We’re hoping data, such as reviews on certain products or time of visiting recipe page, can give us guidance into customer preferences so that we can take them into consideration. The students are also looking to create taste profiles based on ingredients that go into certain meals.”
“Overall, the experience so far has really helped us to adopt the mindset that we can and should use data in our business daily.”
“You can apply this to many industries, not just the food industry.”
Lastly, what’s your dream for data science and the food and agriculture industry?
“Ideally, what we should be able to do is forecast three months ahead, and plan menus three months ahead, because this will allow us to shorten the supply chain. We can tell farmers we will sell a certain amount of kale, ask them to grow it to be ready for a certain date — demand driven farming (growing for and not above it).”
“You can apply this to many industries, not just the food industry. The more companies that are working to predict their consumers’ wants and needs, and when they are likely to purchase, the more efficient supply chains can become, resulting in less waste.”
This dream is, in my opinion, well on its way to becoming a reality, and the agriculture industry is rapidly showing us how. See my latest piece on Medium, a recap of the FarmHack, for just one example of a company thinking and moving towards the same direction as Joris and Marleenkookt.
Nature driven, data obsessed. It’s the way to go.
It’s also why my PoC lab (which is progressing nicely) will certainly focus on data food and smart farming. Keep your eyes peeled.
