How to use data for product development — Basic types of analysis relevant to the business. | by Alex A. Szczurek | Mar, 2023

Tasks performed by product managers and ux’ers largely focus on data analysis, as it is a key element in each of the four stages of the product life cycle. What can such analysis involve? Trends in the market, competitive environment, target group needs, or customer satisfaction levels. Proper use of data also facilitates sales optimization, marketing, and promotional activities. How to use corporate databases and develop products using them?

Business analysis and its challenges Organizations worldwide have been using more or less sophisticated analytical methods for years to scale their operations and develop a competitive advantage with lasting effect. Effective data analysis also allows them to find answers to questions about events that have occurred in the past, facilitate current business decisions, and provide an opportunity to accurately identify or predict future trends and tendencies, which in turn enables a faster response to market needs. Despite the long tradition of using analytical methods in business, many developed methods, as well as a wide range of tools designed for conducting analytical activities, the extremely dynamic development of Big Data, artificial intelligence, and cognitive systems create a series of new challenges. It changes the previous approach to analytical methods, creates the necessity of using completely new tools, and forces the organization to go through a multi-dimensional transformation process.

The analysis of the data collected by the organization brings many benefits to the business, and ignoring them is a huge mistake. It is based on the data that a range of analyses important for business is carried out, answering various questions. Among other things, these analyses are classified into five basic groups:

Descriptive analytics

A descriptive analysis that is based on historical data and leads to the identification of certain patterns and quantitative relationships — answers the question, “what happened.” The results of the analysis are left to the discretion of interpretation and utilization of the acquired knowledge. A typical example of descriptive analysis are corporate reports, which are a historical review of its operations — sales results for the last quarter or the level of budget implementation last year.

Diagnostic analytics

Analysis aimed at explaining the causes of a particular phenomenon. Like descriptive analytics, it is based on historical data and looks for factors among them that influenced the realization of such a scenario and not another. It answers the question, “why did it happen,” why sales were lower than expected in the last quarter, or why the marketing budget was significantly exceeded last year.

Predictive analytics

Despite the fact that predictive analytics uses historical data, it tries with greater or lesser probability to answer the question, “what will happen in the future” and what will be the consequences. Such analysis finds broad application and is used not only for forecasting but also for risk modeling, predicting customer behavior, verifying hypotheses, etc.

Prescriptive analytics

Prescriptive analysis, in turn, answers the question, “what to do.” It takes into account possible scenarios and eventualities that result from predictive analysis, and then supports the choice of the most favorable solution for the organization. What is the application of prescriptive analytics in business? It is primarily used in broadly understood optimization processes (including investment or customer portfolio), risk management processes, or offering. Therefore, they work wherever the automation of complex decision-making processes is necessary due to the operation of a large amount of data.

Cognitive Analytics

The development of artificial intelligence and cognitive systems has resulted in the emergence of a completely new category of analysis — cognitive analytics. What characterizes it? It utilizes artificial intelligence tools, machine learning, and constantly developing natural language processing algorithms (NLP — Natural Language Processing) to maximize the automation of the analytical process and, consequently, the decision-making process in an organization. This process can be almost entirely automated or programmed in such a way that some decisions are made with the involvement of humans. Cognitive systems are used today, among others, in the medical and banking industries, where they contribute to the effective identification of financial fraud.

While just a few years ago, cognitive analytics was not a separate category, today its presence should not surprise anyone. This is naturally related to the increasing importance of data, organizations recognizing the value that lies dormant in them, and the enormous benefits of automating decision-making processes. The use of cognitive analytics allows for huge time savings and rapid decision-making. Furthermore, delegating analytical tasks to task-oriented AI (Artificial Intelligence) systems allows for the elimination of errors and a significant improvement in efficiency. It is also important to note that the development of deep learning technology enables such systems to continuously improve and adapt to changes in existing patterns.

Does the development of artificial intelligence and cognitive systems mean the end of traditional analytics? Such concerns are entirely unnecessary because, according to Gartner, IBM, and SAS analysts, cognitive analytics is a natural complement to the so-called analytical continuum, i.e., the analytical process that begins with descriptive analysis and ends with predictive analysis. It is the culmination of this process and a way to take business analysis to an entirely new level.

Product Development Planning

Using Data The classic product life cycle assumes the existence of four phases:

  • market introduction
  • sales growth
  • market maturity
  • decline, and subsequently, withdrawal of the product from the market.

Each of these stages presents a series of challenges for product and marketing teams, whose close cooperation is one of the key factors in success. The first step, introducing a new product to the market, is particularly critical because it generates significant costs and risk — substantial investments associated with creating a new product and bringing it to market do not always find coverage in generated revenues. Moreover, this phase requires making many important decisions — including decisions regarding promotion, pricing strategy, or distribution — which largely determine the transition to the second phase, i.e., sales growth. Despite the enormous challenges associated with the initial phases of product development, it cannot be assumed that managing the product in the market maturity phase does not pose difficulties. The huge competition, changing trends, or the development of new distribution channels are probably the biggest challenges that product managers face every day.

One of the most effective ways to manage and develop a product, especially in the age of Big Data, is to use insights from analyzing the company’s databases in the decision-making process. Treating them as an important asset of the organization is undoubtedly the first step towards implementing a data-driven management strategy. How to use data for product development? Contrary to appearances, there are really a lot of possibilities. Advanced data analysis, both owned by the organization and those currently ignored by it, allows for the development of a range of tangible benefits.

The first and most important issue is the analysis of sales and cost data, as they are the basic indicator of product evaluation. A more detailed analysis of basic cost and revenue data can also lead to increased operational efficiency and reduced risk levels. The use of modern Business Intelligence technologies and tools, on the one hand, takes analytical actions to a completely new level — allows data to be combined from different sources and thereby more effectively monetize corporate databases, and on the other hand, brings additional value in the form of the possibility of analyzing them in real-time. Such a possibility for product managers is invaluable because it allows them to make decisions much faster and respond to changing trends or customer expectations. And it is precisely customers, alongside results, who should be the next area of interest for analysts. Regardless of the type and number of activities carried out, analyzing information about customers allows identifying certain business relationships, assessing the effectiveness of promotional campaigns, implemented pricing strategies, loyalty levels, etc. Knowledge of customer purchasing habits, representing individual segments, as well as a precise understanding of their needs, allows not only improving the offer and the product itself, but also optimizing inventory levels, reorganizing the supply chain, and taking marketing actions that will maximize sales.

As research shows, for over 60% of customers, satisfaction with service is more important than the price of the product or service, yet only a few percent of organizations gather information allowing for comprehensive analysis in this area. To optimize the level of customer service, it becomes necessary to reach for modern analytical tools that allow for gathering not only data from the transactional system but also from social media, the corporate website, or even the telephone customer service office.


Simultaneously focusing on increasing sales and improving customer engagement and satisfaction is the most effective way to build long-term relationships. Monitoring customer behavior and satisfaction with the offer is somewhat easier in the case of e-commerce, as the customer’s purchase path on the Internet can be tracked extremely precisely — from entering the online store’s website to finalizing the transaction. It is also possible to identify viewed products and the moment of abandoning the shopping cart.

Collecting data in this case allows not only to optimize the offer but also to personalize customer service and use of recommendation systems that dynamically select products to suit the customer’s preferences and present them at any point in the purchase path. Personalized recommendations are an effective way to increase sales by several to even dozens of percent. Nevertheless, automation of marketing, i.e., directing various marketing messages to online store users in real-time based on an analysis of their behavior, allows for even more satisfying results. Without collecting data, realizing such benefits would not be possible.

What do you think?

Leave a Reply

Your email address will not be published. Required fields are marked *

GIPHY App Key not set. Please check settings