Influence of Corruption Resilience Patterns to Profits from E-Commerce for Companies:: AI&MLT for Big Data Analysis

In this paper, we describe the estimation of corruption patterns in a specific region and their influence on the potential results of e-commerce for companies. Estimating the level of corruption is extremely vital in the post-crisis (like post-pandemics) recovery times. When entering a new region, companies forecast costs and benefits. Analysts may calculate the profit from selling an item through the e-commerce channels as revenue minus costs. Nevertheless, there are no chances to introduce the corruption patterns into the common balance sheet or income statement. Thus, the analysts can receive biased results from the outputs of e-commerce, because they will not observe part of the costs. It is easier to count corruption for the companies that persist long in a particular region. However, it can be a challenge for a new company entering the regional market. The new company will see the costs and benefits of its potential competitors entering the region without seeing the underwater part of this “iceberg”. Thus, the planned e-commerce campaign is a subject for a biased estimation on any stage of planning. Every percentage deviation of estimated revenue and costs may be crucial in accepting the decision when dealing the e-commerce. The competition in these sectors is severe; therefore, companies cannot ignore kickbacks or similar costs when planning e-commerce campaigns. Regressions and classical statistics might not be very useful in estimating the covered patterns. This is why MLT and AI algorithms become vital in this field. This research suggests a methodology and ready modules for the best estimate of corruption patterns possible at this time. The researchers show the efficiency of their approach on the example of the BEEPS data and company financial reporting. Entering the new market means estimating the competitors not just in terms of products they sell, but in terms of their financial stability and bankruptcy probability as well. The companies can use different accounting practices when providing financial reporting to their owners (the real reporting) and to the third parties (it can be modified balance sheet). This research suggests methods to convert the modified financial reporting into real financial reporting through AI. The AI algorithm suggested can take the modified data on the company and restore its true financial conditions and positions on the market. This could be crucial information for the e-commerce campaigns in new regions.