Coast & Glen is a fish merchant who sells online subscriptions to a home delivered ‘Fishbox’.
Fishbox adopts the basic idea of a vegetable box delivered directly to customers, but unlike the traditional vegetable box business, their ‘Fishboxers’ will receive their order anywhere in the UK and can subscribe and manage their account online. Customers sign up to a regular weekly, fortnightly or monthly box of three different sizes. They pick their preferences of types of fish once signed up and rate their products to love, like and dislike so Coast & Glen know how frequently to send out different fish and seafood options.
More than 800 people are currently signed up to receive a regular Fishbox and this number is growing daily. There are 70 fish and shellfish products offered to customers, but their availability and yield is seasonal and the subscribers’ delivery dates vary as do their product preferences. Currently buying and box packing decisions are made based on expert knowledge and experience but as the customer base grows, these decisions will be much more difficult to manage.
Coast and Glen were looking for an academic collaboration to develop algorithms to ensure that as they scale up their operation and number of customers, they are able to produce the same high quality product, with freshness and fast turnover of stock being the key. This algorithm would 'virtually' prepack customers’ orders in advance and give an order list for when the company is buying fish at the market.
If, for example, they had 200 fishboxes due for despatch on a Wednesday, the algorithm would enable them to predict, based on those customers preferences and box size, what type of fish, and how much of each type of fish they would need to buy at Tuesday's fish markets. The difficulty they had is that they deal with wild fish, which means the markets change day to day. So, in order to produce an order form for the markets, they needed a priority list. For example, if they intended to order boxes of pollock but no pollock was available, then the list would have them order boxes of plaice instead.
The other challenge was that all the species have a different yield of fillet. So e.g., from pollock there is a 50% yield, whereas plaice it is around 30%. Therefore, in order to service the same number of orders, more whole plaice would be needed from the market than pollock. Yield is a value that would be input and changed as often as required.
As a result, the ordering process would have to be quite flexible to allow for number of customer orders, preferences, box size, type of seafood species and varying fish yields.
As this project is both innovative and unique, Coast & Glen wanted to partner with a university to utilise the broad thinking of lecturers and students alike. They had enquired with software developers who were involved in the fishing industry, but it was felt they were unable to competently complete such a project. The project required statistical analysis of customer data and a robust mathematical framework which would use the results of these analyses and combine them with information on availability and price. This was a complex system, requiring high level mathematical skills.
The company approached Interface who matched them with Professor Norman at the University of Stirling to develop an algorithm which would translate their fish box subscribers’ ‘live’ shopping list into the quantity and type of fish the company were required to source on a weekly basis. This mathematical framework, built using expert academic knowledge, would make the decisions automatically and in real time. This would allow the company’s ordering process to become much more automated and reduce errors in the supply chain process.
The project was funded by a Scottish Funding Council Innovation Voucher.
This new process ensures that the company can fulfill fishbox orders whilst still utilising all species available at the market, thus reducing pressure on any particular fish species, and that their fishboxes are sustainable for the future. It will also minimise any waste from over-ordering. Measures of customer satisfaction and profit will be taken into account.
With the ordering processes made simple and sustainable, the company will be able to scale up and become the market leaders for home delivered, sustainable seafood in the UK.
Academic - The impact for the academic team is threefold. Firstly, it has provided a training opportunity for staff, secondly, it has allowed them to work on an industry led project, and thirdly, it has enabled the development of new mathematical ideas for solving a real life problem. This provides transferable and employability skills.
The project also provides a potential impact case study for Ref2020. The University is currently recording activities/interactions/publicity arising from this project and the impact of this research on Coast and Glen. This would then feed directly into an Impact case study for the mathematics panel of ref2020. It is believed that impact will carry more weight in the next Ref and so this project could have long term benefit to the University of Stirling, both financially and in terms of reputation and prestige.
Finally, the University has benefited in terms of potential future collaborations as there is the potential for more large scale uses of the algorithm.
Scottish Economy -The Scottish economy benefits in that the more fishboxes the company are able to sell in the UK, the less Scottish seafood will be exported to Europe and the Far East, allowing;
a) a better price for fishermen by keeping the catch local and
b) offering the British people the chance to eat Scottish seafood of all shapes, sizes and varieties!
“My interactions with Interface have been very positive. The Innovation voucher is for a modest sum of money (compared to RCUK grants), but gave us the opportunity to do to do some really interesting research which is both mathematically challenging and which solves a real world problem. Working with Coast and Glen has been extremely enjoyable and I have gained a real buzz from working with such an energetic, creative and enthusiastic team. There is now the possibility of taking this work further and I hope that this will be the first step along a very productive path.” Rachel Norman, Dept of Computing Science and Mathematics, University of Stirling
"Working with Interface has allowed us to create the beginnings of a really useful, innovative algorithm which will allow our business to run more smoothly and ensure our ordering and packing processes can keep up with the demand of our speed of growth in customer numbers. We knew we needed a really bespoke solution to our problem and that input from an academic collaboration was going to be necessary. Interface helped us link up with the right team at the University of Stirling, as well as providing funding towards the project. We have thoroughly enjoyed working with our academic partners at Stirling - the team are so enthusiastic, knowledgeable and resourceful. We have built a really strong working relationship and have just been accepted for our follow-on innovation voucher to continue working together to develop our project further and we are looking forward to continuing our collaboration." Fiona Houston, Coast and Glen
The Scottish EDGE fund 2014: WINNER.
Scottish National Chamber Awards 2014: Business of the Year - FINALIST
Highland Business Awards 2014: Outstanding Performing Business (less than 25 employees) FINALIST
Business Leader of the Year - Magnus Houston. FINALIST
Highland Food & Drink Awards 2014: New Product Award - FISHBOX - Highly Commended
Please note that Interface administers the Innovation Voucher Scheme on behalf of the Scottish Funding Council. All funding applications are reviewed on a case by case basis by the Scottish Funding Council, guidelines can be found here.