The customer

The client company, a company operating in the energy market, has been developing a marketing campaign aimed at acquiring new customers for more than 2 years through the Google Ads and Facebook Ads platforms.

The problem

The marketing campaigns launched by the aforementioned company suffer from the lack of an analysis of the results, so it is difficult to evaluate their effectiveness.

The solution

The solution provided has a dual objective, namely to provide the client company with detailed reports on the impact of the major reference metrics in the Social Marketing field in relation to the campaigns already carried out and to equip the company with powerful forecasting tools in view of the new campaigns to be start in order to correctly target their promotional activities.

Planning and Implementation

The project implemented by Koros Consulting required access as an Analyst to the Google and Facebook Ads accounts of the client company, so as to be able to carry out a data mining activity followed by a data cleaning activity using tools such as Office and Power BI.

At a later stage, Power BI reports were created in the form of dashboards concerning the analysis of the aforementioned data relating to the time interval from 30/01/2017 to 25/10/2019.

The last phase involved the goal of designing machine learning algorithms using libraries developed ad-hoc in Python language to be implemented on a web interface using the Flask microframework for the backend and the Vuejs framework for the frontend.

Google Ads Analysis

In order to evaluate the implementation of the customer assistance service, Koros Consulting has analyzed the data relating to the telephone call log resulting from the results of the Google Ads campaigns in the period from 06/10/2017 to 04/10/2019 . In addition, the main Social Media Marketing metrics were analyzed to plan future marketing campaigns and establish the most suitable content to increase engagement, including: impressions, coverage, CTR, cost per conversion. The goal was to pay attention not only to the content, but also the time and date of publication on Google, in order to become aware of the variables that affect user feedback.

Facebook Insights Analysis

Finally, following a data mining and data cleaning activity aimed at evaluating the effectiveness of the posts published on Facebook, an analysis was carried out, through the Power BI software, of the data from the posts published on Facebook in the interval temporal from 13/10/2018 to 07/10/219, managing to obtain important information about times, days of the week, months and target people in which the posts were most effective in terms of impressions, engagement rate, coverage, results, clicks and other metrics as well as the results achieved by each type of post published, as well as possible future trends.

Facebook Ads Analysis

After carrying out an analysis of the posts in general, Koros Consulting focused on a specific analysis of the sponsored posts. Tracing the activities of the previous point, he was able to obtain detailed information on an hourly, daily and monthly basis of the main performance metrics, including: coverage, impressions, results, amount spent, cost per result and conversion rate. In this way it was possible to provide the client company with useful considerations to better target the next marketing strategies via Facebook.

Support application

The final phase of the project involved the creation of an application to support business decisions in the marketing field. After having developed and trained a machine learning algorithm, we proceeded to its web implementation through the Flask framework.