...

How Big Data Helps Business Run Faster than Competitors

How Big Data Helps Business Run Faster than Competitors

Why Ukrainian companies are still cautiously launching innovations

A wave of interest in big data swept the world after one of the algorithms (AlexNet) completed the task 10.8% better than its competitor in the image recognition competition.

The Harvard Business Review published a note from one of the creators of the modern meaning of data science that this is the sexiest profession of the 21st century.

Big data and neural networks have been developing for more than half a century, but researchers and businesses have only managed to get results from them in recent years, when computing power began to suffice.

Since then, the effect of using Big Data has been constantly increasing both on the part of companies and on the part of specialists. In the wake of this request from the business, Kyivstar opened its Big Data School..

According to the mobile operator, 7,600 people applied for training in four waves of Big Data School, and every tenth person coped with the test task. 110 certified specialists entered the market, and 12 of the best got a job in the company.
However, the level of readiness to launch analytics on behalf of Ukrainian and foreign companies differs significantly.

Why aren’t many domestic enterprises investing in new technologies yet? How can big data remove the “expert prism” and cut operational costs? What transformation paths are Ukrainian companies choosing? And where are the “smart” refrigerators? The mentors of Big Data School 4.0 from Kyivstar Nika Tamayo Flores and Valentin Kropov spoke about this and many other things to Mind.

Nika Tamayo Flores has a background in project management for the implementation of data science and business intelligence in various companies. She studied at IE Business School (Spain) with a degree in Business Analytics and Big Data.

Valentin Kropov is an expert with 15 years of experience working with data. He is also the founder of the largest community of Big Data developers and analysts in Eastern Europe.

Treat data like money

According to the observations of Nika Tamayo Flores, companies often come to Big Data only after realizing that something is already wrong with their business: either there is no opportunity for growth; competitors are running out, but traditional methods cannot get out of the situation; or the cost of each new percentage of growth is too high, and new non-traditional approaches are needed.

In contrast to Ukraine, abroad, many of the basic points in companies have already been implemented. For example, data collection has been established.

We have a blue ocean for those who offer solutions. Often, Ukrainian business sees value in money, personnel, but not in information. If they collect data, they store it in Excel files, delete last year’s information if space is needed, do not monitor accuracy and often postpone investments in IT until better times.

Nika Tamayo Flores warns that without analytics, businesses are at risk of emotional decisions.

I woke up in a bad mood, something went wrong, a bad day. Real data could suggest that we need to act differently. But the company does not have analytics, because in order to invest in analytics, you need to be ready to use it. After all, the return will be gradual. That is why investments begin only when the pressure is applied. And when a lot has already been lost due to wrong, impulsive decisions.

The second reason for making bad business decisions is the “expert prism”. An employee can have 20 years of experience behind him and be a great specialist. But how many MRT images does a doctor manage to review during his career, and how many does a neural network have during training?

The difference is in the millions. The store owner may well know the habits of buyers from neighboring houses. But if a large elite residential complex is built nearby, will he be able to assess in real time how demand will change?

From “Excel” to Artificial Intelligence

“Today the most advanced businesses in working with Big Data are banks, retail and IT. They started collecting data earlier, so now they have more data and expertise,” Nika Tamayo Flores says.

Most companies are unlikely to be able to catch up with the leaders in three jumps. Between working in Excel and real Big Data analytics, there is a huge number of steps and changes in processes. Typically, businesses hire consultants in this area. Specialists conduct training for management and employees, who also need to realize this: accuracy in data is no less important than money.

According to the experts, an average Ukrainian company does not always have a data warehouse, but 1C for accounting and Excel for everything else can be installed.

Perhaps some departments use CRM. Data is duplicated, employees have no idea about the processes, owners – about what their company really lives on. And if a company decides to implement deeper analytics or reporting, it will take 90% of the time just to bring all the data together. There are only a few of those who already operate with complex analytical tools for studying key indicators.

What are the Benefits of Implementing Analytics?

The top managers of the companies thinking about implementing Big Data analytics have a reasonable question: “How do you calculate the effectiveness?”
Unfortunately, solution developers cannot give specific numbers. A lot of this depends on the customer’s business, how he will use the product, how well the data is collected and what decisions are made based on analytics.

The economic effect of Big Data can be varied. Analytics can help reduce operational costs and decision time, reduce manual hours and staff, or shift them to more complex tasks.

Valentin Kropov – Vice President, Head of Retail-vertical at SoftServe illustrates the increase in business value in two areas for e-commerce development: product and consumer analysis.

The first concern is search.. Buyers are gradually moving online, where their choice is almost unlimited.

But the seller has a problem: how to make available, let’s say, 10,000 units of goods with the minimum resources.

Algorithms come to the rescue that are able to recognize the image and select options for the user according to his specific requests.

For example, if a customer knows what kind of dress she needs, she can upload an image and find all similar options in the store. Or enter a description in words.

The next area of ​​application is the recognition of the user himself. After the presidential election in the United States, the media started talking about personalized work with users depending on their psychotype.

But if in the case of the presidential elections the approach raised ethical questions, then in business use, subject to the consent of the user, this isn’t a problem.

As an example, Valentin Kropov cites a solution based on OCEAN technology – a system for creating a portrait of a user. Each letter of the abbreviation is a specific psychological characteristic. Algorithms analyze human behavior on the internet and build his psychological portrait. Then, the delivery of goods on the site is changed so that the user receives information in a form convenient for his personal perception.

There are buyers who find it easier to make decisions emotionally. For example, if they see clothes, they value the impression of it, a good picture.

Other people prefer detailed characteristics and comparison with alternatives. Such site visitors should show the table and recommend different options. To make shopping personalized, the buyer, in turn, only needs to agree to the processing of personal data. This analysis can boost sales by 3%. For retail, this is a tangible indicator.

According to Valentin Kropov, the main challenge in this area is the transition from online to the physical world. This is where the next battlefield will unfold, since many things are still not automated offline.

For example, “smart” industrial refrigerators have appeared in Ukrainian stores. They can not only measure the temperature and determine when the product is running out, but also signal if something other than the brand’s product is put into the refrigerator.

With the help of algorithms, the space of outlets also “grows smarter.” Cameras in stores have learned to recognize when a shelf has run out of goods or someone has been standing in one place for a suspiciously long time.

The solution will tell you that you need to order a new batch or send an employee to the buyer. The system recognizes when customers enter the store, who they are, what they are interested in,which parts of the store are visited and which are not.

Based on this data, the owners can form an assortment and launch promotions. This is what was said above: the algorithm protects the business from the “expert prism” and provides the tools for making decisions based on the real needs of the audience in real time and in any territory.

Two Ways: Completing the “Rooms” or Constructing a New Building

But to get the effect, you need investments in IT. Often, Ukrainian business moves in a vicious circle: I would like to show good economic results to shareholders this year, and investments in IT are less than ideal.

To break this circle, some companies decide to start doing what will bring economic results in the short term: they choose priority tasks that will show results within 2-3 months. If management sees the initial effect, then it begins to invest in more expensive and long-term developments.

There is another option. It is longer and more resource intensive. This is not the imposition of technologies on current business processes, but wrapping technologies around business processes, optimizing the redundant and outdated. Algorithms can qualitatively change the way a company operates.

Big Data Trends

The main trend is understanding the use of data. Companies that already have data sets are beginning to figure out what this information is and what can be done with it.

In addition, new types of data are emerging that are transmitted through IoT devices. For example, soil moisture sensors in agriculture, air pollution indicators, etc.

Consumer understanding is another important area of ​​work with analytics. Even Facebook still asks questions about how the user feels and thinks, why he does certain things. Business strategies that are based on data are also based on understanding why the consumer makes a particular decision.

From there, the next trend is data management. This includes security, data access, and the transition from file systems to a single structured platform.

Nika Tamayo Flores emphasizes: all this is important only when the business has a development strategy. Ukrainian enterprises often do not have it, they just want to grow. This is not enough. You need not only want to grow, but clearly understand what the company wants to become in the near future.

Today, many companies are starting from scratch with big data and moving forward. In the future, they will have an advantage: those who do not take risks now will lose to the competition in the coming years.

There are already 2-3 more successful data players in almost every market today. And their competitors are aware of it. They face a choice: either invest more money in unconventional methods, to stay on the sidelines, or to start working with Big Data, attracting partners or hiring their own Big Data analysts.