In the telecommunications sector, the existing regulations allow us to know the flows of clients between the different companies, which together with the migrations of clients between the internal rates of an operator make up a complete graph of incoming and outgoing economic flows between own rates and competing operators.

In this article we are going to show you how to take advantage of daily analytical studies of the results of marketing campaigns, visualize the evolution of business volume by rates, project future results, work with simulations of the effects of a rate or how to weight it with the market sentiment.

When a customer performs a number portability, the originating company receives the notification of the change so that the portability can be carried out. This particularity enables the analysis of customer flows from the different rates of the originating company to the operators of the competition, and vice versa. Let's go with it!

Complex scenario

El marketing department of a TELCO boosts business results through successive campaigns with new rates, promotions and variations in existing rates. Attention, we are talking about the order of several dozens of rates active simultaneously.

Marketing and the market

If it is already complex to monitor each one in detail, analyzing the relationships between rates becomes an almost impossible task. Let's imagine the challenge if, in addition, we want to know it continuously, without waiting for elaborate monthly reports created in an artisanal way. Questions such as:

• What rates are attracting more customers from the competition?
• What rates are starting to suffer escapes?
• The new rate launched is meeting expectations?
• What is the trend, I need to do something now?
• Which is the market sentiment?

If from a marketing point of view the situation is complex, for the Finance department it is not much better. Also, the goals of different departments are not always aligned. It can be the paradox that a campaign is having an incredible reception, but is causing a financial hole. New questions are added:

• Is the apparently so successful new tariff cannibalizing own customers who had more beneficial rates for the company?
• Are the new campaigns really reporting Benefits?
• Are marketing initiatives going in the proper direction?
• What would be the economic projection in the next 6 months if the Churn was in the vicinity of 5% and the ARPU half out of the order of € 30?

Dynamic and global analysis of information

The basic information to answer these questions exists.
The information needed to answer business questions exists.
The main challenge is that the information is usually scattered.

Each department has a biased view of the company's situation, preventing agile decision-making based on global knowledge. It is necessary to identify the data sources, extract and organize all the information to build the basis on which to apply the dynamic and continuous analysis of the data. We will do it using, among others, graph analysis algorithms, time series, automatic learning y advanced graphical visualization tools. A whole techno festival!

The following figure shows the logical design of components of the continuous dynamic analysis system that allows responding to the needs described.

Logic diagram for dynamic and predictive analytics

Internal customer flow

It is very relevant to analyze the internal movement of customers among the multitude of rates offered, both at the rate level and rate families.
It is possible that a new rate designed to attract customers from a competitor has the unintended effect of attracting your own customers who had a higher ARPU rate. These types of situations can be serious and must be detected as soon as possible.

If information on internal rate migrations is available, alert systems can be configured to warn of possible unwanted dynamics.
At the same time, marketing or finance departments can precisely track the evolution of certain rates or rate segment, at the customer volume or budget level, all with a few clicks of the mouse. The information can be filtered by the desired time interval and displayed as a graph.

Customer flow graph between rates

In the example graph, we see the inflows and outflows for a selected Tariff B, with other internal tariffs, other operators, new customer registrations and cancellations. The color and thickness of the arcs indicates the positive or negative net balance and the magnitude.

This type of representations allow to visualize various levels of jumps
to show in a simple way all the flows associated with a tariff.
Spectacular!

Flows with the competition

We can also monitor if our clients are being attracted by rates from our competition, knowing in detail when, how many and where. We detect it automatically, generating alerts to decide if we want to propose new campaigns aimed at these clients.
For example, we can present it as input and output relationships in table form.

Rate migrations with other operators

In the example, the red balls represent negative net balances, indicating that there is a loss of customers from our rates to the competition. The size represents the magnitude of the flow. In this example, Rates A, B and C are very negative and require immediate action..

Market sentiment

To help interpret the results that are being obtained, we complement it with the dynamic analysis of information on Internet searches offered by Google Trends, or the dynamic analysis of social media information. This vision makes it easier to understand the impact and reception of new products and offers, as well as their relationship with other competitors.

Conclusions

Today's market is constantly changing, often in an unpredictable way. To survive it is essential to evolve from the classic executive reports, very colorful and aesthetic, but impractical and oriented to interpret the past, to new dynamic and living processes. Analytical processes that accompany us on a day-to-day basis, that allow us to immediately capture what is happening to continuously adjust our decision-making, drastically cutting reaction time.

This transformation is already taking place thanks to the technologies applied in the treatment of massive and continuous data flows, from different sources, with the capacity to process algorithms in a parallelized way and taken advantage of from latest generation graphical tools. If you want to know more details we encourage you to visit our success stories data visualization and analysis Welcome!