How can data be used to boost performance & business growth?
April 17, 2018
When companies are looking for the big answers to how they grow, improve productivity or increase profits they often look outside to consultants for help. Yet, if you’re looking for answers with strong evidence, provable returns and clarity then the alternative is to look inside the company – at your data.
As businesses have become software and internet enabled they are creating data at a staggering rate. 90% of the world’s entire data was created in the last two years, and this exponential growth is continuing. In companies, your enterprise software, sensors, handheld devices, trackers and phone systems all create and store vast volumes of data, all of which can be used to understand where profit is being lost, where opportunity lies and how to grow.
The first challenge is seeing your data. Few companies have a comprehensive oversight of their core data and then use this to measure against KPIs. Even those companies that do look at their data often do it in silos with production, sales, finance and logistics all looking at their own data without realising the impact across the business. The software now exists to dashboard your data and share it across the company to improve understanding of the business.
The second challenge is acting on it, as looking at data is both pointless and stressful without action. Some actions will become obvious from the data; however, others will need greater understanding and this is where algorithms come in.
Algorithms have been around since the Ancient Greeks but their use has accelerated as three key ingredients have come together – the vast proliferation of data; the internet and faster broadband speeds; and greater computer processing power.
An algorithm, which in simple form looks like a flowchart, is an automated process that precisely defines a sequence of operations. From Google’s search engine page ranking, to how Netflix recommends films, to how your satnav works, algorithms are powering improvements across businesses and helping them get ahead of their competition.
For example, a UK company with a large field service team acquired a complimentary business in a multi-million-pound takeover. The acquirer was looking to find efficiency gains from the acquisition through rationalising the number of its combined depots. Both companies had data on every job they undertook, the exact location of the depot and the client site, the time taken to travel, the number of staff required and more. Then by adding in external data such as the most optimal routes and fuel costs, we proposed the ideal model for the company’s depot locations. This greenfield scenario saved approximately £2.2 million in annual running costs. However, this included a depot on Oxford Street in London, as the algorithm showed it was the easiest location to serve central London clients. Once exclusions based on lease lengths, rents and related costs were included, the final scenario, which was implemented, reduced the number of depots by just over 40%, produced £1.55 million in annual savings, jobs would have been serviced in the same time or quicker in 96% of cases, and only 5% of staff would have faced a longer commute to work with more having a shorter commute.
On their own, algorithms provide a one-time answer to a problem, providing both insight and action that can lead to substantial company improvements and all by using your own data. However, for continuous improvement, companies are now looking to machine learning.
Rather than be the stuff of nightmares with robots taking over the world, machine learning is simply self-improving algorithms. Using training data and live data the algorithm learns and improves itself to continually optimise the process and increase the accuracy of predictions.
Machine learning is used in situations where the capacity of a human being to effectively undertake a task, both in terms of speed and accuracy, is challenging. If a machine learning process can do it more quickly and more accurately then the benefits are clear. If, for example, manual processes or simple software predict say only 55% of potential fraud cases and machine learning can improve this to 78%, and more rapidly, the resulting savings, customer satisfaction and on-going improvement will have a huge impact on the company.
Across business, machine learning is now playing a vital role.
– At BMW, and other leading manufacturers, they use data from sensors and machine learning to predict when equipment will fail. This enables maintenance in advance of failure thus offsetting the huge costs of production shutdowns.
– At Regit, they’ve used machine learning to predict which of their 2.5 million users are going to change car and when. This helped personalise their service and increased call centre revenues by 27% and a 35% reduction in operational costs.
– Ecommerce companies are increasingly using data from customer behaviour on their websites and machine learning to automate product searches, upsell and cross-sell buttons to deliver products most relevant to the individual customer and thus increase sales and conversions. Those with a large product catalogue can also use machine learning to optimise price points and product-price elasticity to maximise margins.
– In financial companies or in finance functions, machine learning is used on internal and external data to understand which customers are most likely to pay late or are at risk of default.
– In medicine, machine learning is working alongside clinicians to spot tumours, skin lesions and other signs of illness with greater accuracy than their human equivalents.
– Recently, substantially improve the prediction of the timing and location of vehicle returns at a leading car rental company. As the company had to pre-book transporters they could more confidently book them based on the improved predictions. This led to a reduction in costs of over £1 million and better vehicle optimisation.
Around the world successful businesses are using data to build new revenue streams, improve operations, grow profitability and improve productivity, all by simply using the information they generate every day.
If you’re looking to grow your business with confidence, then often the answers lie inside.
If you would like to discuss how data can be used to boost performance and business growth in more detail, please contact Michael Gibson who is the Director of at [email protected]