Not very long ago, in a European multi-billion turnover, multi-national dairy company (yes such companies do exist!) the executive team was struggling with a thorny challenge. They were confident that they could optimise a key supply chain to save costs and improve the quality of outputs, but they couldn’t get the data to prove this hunch.
However, the challenge was complex
It was all about the process of bringing milk products from farms into various sophisticated production processes which needed to be fine-tuned to respond to the composition of dairy products from multiple sources, each of which varied every day. The incoming batches needed to be combined to deliver optimum inputs, and the production processes in turn needed to be adjusted to deliver the best products. Plus, because the raw materials had a short shelf-life, delays, and mistakes were costly.
It was intensely frustrating. The exec team was sure that all of the information to make these production decisions was already available or could be easily captured, but it was on a mess of different systems, in different formats, and available at different times. For example, the tankers tested milk composition as they collected it and the GPS systems on the trucks tracked their journeys through various traffic conditions to the different delivery points. The production systems were individually automated so the volume of products in each could be tracked. And Market demand was being tracked in the supermarkets which were their very demanding customers. All of that information was already being used but the process was certainly too slow and was definitely not optimised. Money was being left on the table.
Throwing resources at the problem wouldn’t work
Attempts to capture the data ‘by hand’ with people analyzing it in spreadsheets were simply not robust. And As a result, they grew exponentially more expensive and complex before they could develop a comprehensive view. They needed a way to pick and choose live information from all the different data sources to build possible solutions using trial-and-error, which in turn could be tested to deliver the optimum process. All of these potential solutions needed to be automated so that they could be repeat-tested. Of course, it needed to be cheap and deliver results fast – because the business case was built on a hunch.
This is a common problem across many industries. Even when the goods moving in the supply chain don’t have a similarly short shelf life of dairy goods, the complexity in supply chain data sources makes it difficult to follow the intuition of experienced businesspeople to find and realise the value that they thought existed there.
It was clear that traditional IT approaches were not going to cut it
The IT team looked at capturing the data and feeding it into a data repository. They decided the solution would be i) too inflexible – they were sure that as soon as the business could see some of the data they would realise there was other data that would be needed and some would be redundant, and ii) too risky – integration projects usually overrun on time and cost and this was even more likely if the requirements would be developed as the solution evolved.
An unconventional approach was required
One relatively junior member of the IT team, recently recruited from a big four consultancy, had a suggestion. They had seen a tool called r4apps used in the rescue of a large integration project which was failing. It seemed to fit the requirement to access everything in real-time and solve a complicated problem using trial-and-error. And it had automation built in.
The results speak for themselves
In four months, the solution was in place and working. They could respond in real-time to the supply chain as the goods flowed from farm to factory and, because they could easily test different process scenarios to improve efficiencies, they showed a 35% increase in profits on the product lines in scope. To top it all there was a clear improvement in customer satisfaction.
It isn’t just the dairy industry who have complex supply chains that require real-time analysis across multiple existing data sources. You might need to optimise a dynamic production process or track emissions data to satisfy new ESG reporting requirements. Imagine if you could access everything and solve anything, in real-time – and do that quickly in a risk-free trial-and-error way. Well, you can. Talk to us to learn more.