Getting more out of data
Data is becoming increasingly important for developing new products and services. SMEs can produce more efficiently and address customers more individually.
Moritz von Soden couldn’t believe his eyes at first: within a few minutes, data was streaming to his cell phone via Bluetooth from the new threaded spindle equipped with mini sensors that his development team presented. They provide information about the loads and vibrations of the spindle. “It’s like an X-ray view into the thread,” says the head of Bornemann Gewindetechnik, delighted with the figures spilling out of his smart thread: “You can even see which lubricant is being used.”
Bornemann customers, primarily manufacturers of lifting platforms or material handling equipment, can now monitor the condition of their products around the clock. This allows damage to be detected at an early stage, maintenance intervals to be optimized and breakdowns to be avoided. For the Lower Saxony-based thread manufacturer itself, this innovation opens up a completely new field of business: “By analyzing all the measured data, we gain knowledge that we can offer our customers in addition to our spindles in the future,” says von Soden. “The development of a consulting division is conceivable.”
More and more small and medium-sized companies are using technologies such as business intelligence (BI), big data, analytics and artificial intelligence (AI) to develop products and services and open up new business areas. While BI is primarily used to process and present historical and current data, analytics is geared towards the future. Technologies such as data mining or machine learning are used to determine the probability of future results. This makes well-founded forecasts possible. “Developments can be anticipated in order to initiate any necessary changes for a successful result at an early stage,” explains Christoph Oelrich, head of HR consultancy Steerer.
AI has also arrived in the purchasing, production and logistics departments of medium-sized companies. Whereas in the past, nonwoven production systems were set manually according to the trial-and-error principle, today data is collected on the condition of the machines and the environment, current nonwoven qualities are measured optically and neural networks are trained with this information. An algorithm simulates setting parameters and shows which ones produce the required quality at the lowest cost. In another company, a camera system recognizes whether machines are assembling screws and dowels correctly. If not, staff are alerted within seconds. Product quality increases and the number of complaints decreases.
Another example: for several weeks now, a cosmetics manufacturer in Hesse has been recording in detail how potential customers behave on its website and in social networks, what information a woman clicks on, what questions she asks and what colors she prefers. “With the anonymized but individualized data, the company can get a precise picture of her, recognize the reasons for her behavior and recommend suitable products and content in real time,” explains Rupert Steffner, founder of the start-up Wunder.ai and expert in BI and AI in retail, telecommunications and financial companies. “We also try to map emotional factors and decision-making processes in the human brain in our system as far as possible.”
Business analysts, data engineers and data scientists are in demand on the job market. Prospective computer scientists, physicists and mathematicians are already being recruited on campus. “Some companies cooperate with universities to get the best talent early on,” says headhunter Oelrich. Graduates are lured with annual salaries of 60,000 euros and more, with a doctorate over 70,000 euros are possible: “A data scientist with five years’ experience earns around 90,000 euros on average.” And the trend is rising. “For many high potentials, however, the decisive factor is that the business model is exciting and forward-looking and whether they can work with the latest tools.”
Large companies usually win the race for the brightest minds, while smaller ones can rarely keep up in terms of salary and career opportunities. “What’s more, they often don’t have a manager with practical experience of implementing digital projects,” complains Oelrich. That’s why many SMEs turn to freelancers with daily rates of between 800 and 1,200 euros when needed. They also use the services of software companies that offer programs and consulting for AI, visual analytics, data science or the Internet of Things – including for complete business processes based on the use of data.
Growing interest from SMEs
The global market leader is SAS. The German branch of the US company is registering growing interest from SMEs in solutions that were previously reserved for the big players. “Corona has made many companies painfully aware of their shortcomings when it comes to digitalization,” says Annette Green, responsible for Germany, Switzerland and Austria at SAS. Many decision-makers are not fully aware of their own company’s weaknesses. The cause is often inadequate data management. In a crisis, companies without methodical, conceptual, organizational and technical procedures for using data as a resource are more likely to falter than companies that already rely on big data, analytics or AI: “The latter have a resilience that will protect them in the future in the event of unexpected events and enable them to develop new products and services.”
With the help of analytics and AI, a better customer approach and individually tailored offers are possible. “These are based on all the information available about the customer and the services they have used to date and create trust and satisfaction even in times of limited personal contact,” summarizes Green. This opens up new revenue models. “The technological solutions used should be automated, flexible and, above all, scalable.” This is also helpful when new analytics models such as machine learning are used, “because this requires more computing power over a certain period of time, which can then be added or reduced again as required.”
Bornemann boss von Soden is convinced that smart threads will be a core product for his company in the future. He is already collecting and analyzing data for online marketing: “With success.”