It looks like you are using Internet Explorer, which unfortunately is not supported. Please use a modern browser like Chrome, Firefox, Safari or Edge.

Mika Tiilikainen: “Recommendation engine helps people to do their work better”

Published in Analytics, People

Written by

Samuli Visuri
Managing Director, Analytics and AI

Article

January 21, 2022 · 6 min read time

Nitor continues to invest in digitalisation and advanced analytics solutions. In September 2021, Principal Consultant Mika Tiilikainen joined the company. He brings heavyweight leadership experience in agile and successful projects in modernising analytics solutions at Amer Sports and Finnair. Mika anticipates that sustainability is becoming one of the major drivers for modernising analytics.

Mika, you have 25 years of experience in various roles in the industry. You have now started as a consultant for the first time in your career. What excites you about your job right now?

I’m excited that the breakthrough with data and analytics takes place during my career. I have seen how laptops, mobile phones, and the Internet have changed how we work. I think data science and AI, enabled by big data and the increased computational capacity, can provide even a bigger step in the industrial evolution path.

To put it simply: the former industrial steps improved efficiency and made processes faster. Still, if so desired, they could have been implemented with some traditional methods, too, although that would probably have required huge resources.

In comparison, modern and advanced analytics can solve problems that had been impossible to solve with the earlier technologies and data available. We can now both see and understand things we could not see and understand earlier by any means! This helps businesses solve recurring, multifaceted, and complex problems previously unsolvable.

For instance, people have tried to optimise those problems using a lower level or partial suboptimisation and utilising experience of the subject matter. The sophisticated guessing has been the last resort. The results may have been pretty good, but the modern analytics now provides better visibility to logic and directions where the completely optimised solution could reside.

It is all about achieving the last mile in the value-add within the business. If the business is big in absolute terms, the value-add can also be significant in absolute terms. Alternatively, through competitiveness, it can even be a make or break. It is very motivating to work with these kinds of questions in my current role.

Getting started with the modernisation of analytics and the successful leadership of initial projects are my core competencies. As a consultant, I can now support many companies with challenging tasks. I can also help them develop solutions further once the first steps have been taken.

What are the typical target areas where modern analytics can be applied?

In general, it is beneficial to try applying data science in all the functions where significant decisions are made and that affect business results.

Some great examples include improvements in customer experience, efficiency, or productivity and taking those to the highest level. By moving from the annual product releases to daily releases and receiving rich feedback in real-time, productivity can be multiplied by ten or even a hundred.

Another example can be a recommendation engine provided for every employee making significant business decisions. The most potential lies in those cases where people are still a crucial part of complex optimisation in the decision-making process.

The recommendation engine is not replacing people but providing support to let them do their job even better. A highly valuable benefit is that there will be more time for strategic and tactical development work, rather than staying busy with non-optimal and even pressurised operative work.

Background material used for business decision-making is often incomplete and can have ‘black boxes’, where some areas may lack information completely. With advanced analytics, gaps in the big picture may be filled, and the resolution improved. For instance, there might be gaps in the logistics chain leading to sophisticated guessing in the decision making instead of a data-driven approach. At the same time, the business value of these decisions can be remarkable.

AI can find entirely novel insights that might not have come to people’s minds at all. The transition from spreadsheets to automated dashboards, or from hindsight to predictions, can all improve a company’s performance. The next steps are optimisation and recommendations for the next-best-actions, a.k.a actionable analytics, to get the full competitive advantage from analytics. This is the ultimate goal, toward which everyone should proceed as fast as possible.

A part of the equation is to remove people from the optimisation phase, as they cannot optimise multifaceted and complex problems to the maximum level. It is better to leave this task to the computers.

How to ensure that investments in data science will be successful?

A cornerstone in agile development is to fail fast and learn whenever a selected approach is not feasible. Failures are a natural part of any development, and they will occur on the road to success. When the core targets are selected carefully, failures tend not to cause significant trouble or ‘stop the train’, and the development can continue relentlessly toward the fine-tuned and successful solution.

The dynamics of success is built upon a series of correct and cumulative selections. For example, when selecting the optimal combination of tools from a set of the 30 tools at your disposal, the number of possible tool combos is about one billion. Very importantly, for choosing these tools, the vast experience we have from the successes of our earlier projects demonstrates we can find the best possible combo for each case.

Therefore, both the big picture and the details must be well understood; integrity is needed in all critical points, as well as a humble attitude and activeness with change management. Finally, customers must always be at the centre of everything, just to mention a few examples. It is very beneficial to have the passion for work that we have in this work.

What topical themes would be good to note if you plan the modernisation of analytics right now?

The Covid pandemic has accelerated digitalisation, which is now a must for competitiveness.

A significant theme is optimisation for sustainability and consumption, consisting of carbon and material footprints and the carbon handprint, including compensations. Those will put such intensive pressure on everything that the business operations need to be optimised to the maximum level. This facilitates achieving the climate targets within the tight schedule given while maintaining a healthy business.

Sustainability alone might be the driver for digitalisation in many places. A lot of work needs to be done, although many companies have already started with the modernisation of analytics. Some functions can already have state-of-the-art solutions for some core business areas. However, there are a number of other areas where not much modernisation has taken place yet.

Digitalisation can provide significant improvement in productivity. Therefore, early investments can provide substantial competitive advantage. Whenever data and analytics are used, security, GDPR and ethics need to be paid attention to. 

It is easy to build prototypes for modern analytics, but the professional implementation into production has been a major pain point for a long time. Technologies and solutions are also evolving quickly, so partnering with our professionals is highly recommended to ensure successful and cost-efficient implementations.

Written by

Samuli Visuri
Managing Director, Analytics and AI