Telcos find themselves in a highly competitive market at the moment where customer loyalty is everything. Traditional services like broadband, internet connectivity, GPS tracking and streaming are all disrupted by the innovations in latest technologies like AI. Telecom data analytics helps companies to shape strategy and adapt to the harsh market conditions.
We worked with Neterra Telecommunications as a data analytics consultancy to create company-wide analytics in Power BI. The CFO of Neterra, Mariyan Rangelov used our Power BI reports to apply for and win the CFO of the Year 2024 award from Ernst and Young!
In this article we will explain the areas of analysis that our project covered, the challenges we faced, technologies we used and impact of this analysis. We are sure that these will be similar for other telecommunication companies too.
A large proportion of revenue in the telecom industry is recurring in nature. Therefore, analysing monthly recurring revenue is important to predict financial performance and ensure stability of the company.
Our MRR analysis for Neterra Telecommunication helped them predict future financial performance of the company and plan investment activities accordingly.
B2C customer analysis in telecom industry focuses more around customer segments. Growing and shrinking segments are analysed to inform marketing strategy.
Crucially, we analysed revenue per product and customer group for Neterra. This helped the management to inform cross-selling strategies and market additional products to existing customers.
B2B customer analysis is also highly important because for telcos many suppliers are also their customers. We helped Neterra to analyse their trading balance between every B2B counterparty to help them identify counterparties with whom the trading balance could be improved.
Analysing revenue per location is important because telecommunication companies compete to become a supplier of choice for locations that they cover. This analysis helped Neterra to identify their strongest locations and inform expansion strategies to locations nearby.
Telecommunication companies often have a large portfolio of services. Some of them are great at bringing new customers in but those are not necessarily services that make the most money to the company. It is therefore important to analyse how these different services tie together to achieve a strong financial performance.
Analysing telecom data is no easy task and it is important to be aware of the challenges that you might face. We will also describe our experience in overcoming these challenges based on our case study with Neterra.
Large data volume in the telecommunication industry offers a great opportunity because this makes the analysis more conclusive. If a large volume of data supports the identified patterns, decision makers can act with more confidence.
On the other hand, analysing large data volumes is a challenge on a technical level. The graphs produced as a result of analysis may take a long time to load leading to bad user experience and long development cycles. We solved this problem in our project with Neterra Telecommunications using a variety of measures:
Even though telcos often have an ERP for their transactional and customer data, a lot of data is still siloed. For example, many telecom companies have a need to analyse marketing data together with the data from ERP.
If you find that your data is siloed in many different data sources, aggregating it in a single place is an important step towards comprehensive analytics
You can usually extract the data automatically from your data sources by using data engineering services. Your data sources likely have APIs which means it is possible to write code that sends API requests, receives data and inserts it into a database like SQL Server.
In real-world telecom data analytics projects data often requires substantial cleaning and transformation. This is perfectly normal and does not necessarily mean that your company isn’t ready for analytics.
Common data issues involve missing values in certain columns, inconsistent naming conventions, conflicting data types, etc.
We usually transform the data in either SQL or Power Query. These tools allow us to specify the transformation steps applied to your data every time the data is refreshed in your analytics reports.
Even if you know what analysis you want to create, you need to choose the right technologies to enable you to create it. You will find our recommendations for technologies to use below:
You would typically use these technologies for creating management reports that highlight current and historical performance. You would use them to create charts and tables that analyse your data and present the results in a visual form.
Power BI is our personal favorite being the leading software in the Business Intelligence market. It supports the analysis of large data volumes (millions of rows) which is so important for telecommunication analytics. It also integrates well with other Microsoft tools like Sharepoint and OneDrive so if your telco is using many other Microsoft tools, Power BI might be a natural choice.
Tableau is the main competitor of Power BI and it is also capable of analysing large data volumes. It is a bit more common for businesses outside of the Microsoft ecosystem and has superior data visualization capabilities. Tableau is slightly more expansive than Power BI though so make sure you are prepared for the licensing costs.
Feel free to read our guide for a detailed comparison of Tableau vs Power BI.
We definitely do not recommend using Excel for your descriptive data analytics since it is only capable of loading up to 1 million rows per tab. This is often insufficient for the data volumes in the telecommunication industry and leads to heavy files that are slow to run.
Some simple predictive analytics techniques such as linear regression are possible to apply in Power BI or Tableau. However, larger scale predictive analytics in the telecom industry usually requires machine learning algorithms to complete the project.
Python is probably the most common tool of choice for predictive analytics which is great for large data volumes. Several Python packages such as Scikit-learn and Numpy that can be used to create advanced machine learning models predicting your financial performance or customer churn.
If you prefer a lower-code tool to create your machine learning models, you might want to consider SAS. It allows you to create the same models but offers a drag and drop functionality to create them and you can write code to customize those.
Apart from data analytics technologies, you might also need some supporting data infrastructure. You would typically use a data warehouse technology to store large amounts of data and aggregate data from multiple sources.
Azure SQL Server is a great choice for businesses that primarily use Microsoft technologies. Azure SQL Server databases integrate easily with Microsoft Fabric which enables companies to share data sets easily between teams and access AI support chat bot called co-pilot.
Google Big Query is an alternative to Azure SQL Server which also supports storing large volumes of data. It might be a more suitable solution for pulling the data from other google data sources such as Google Ads and Analytics.
Our team would love to help you create tailored telecom data analytics solutions for your business! We have delivered 1000+ analytics projects to businesses around the world and accumulated significant experience in the telecom industry. Simply contact us to start the conversation!
