Using analytical tools for effective optical network planning

Using analytical tools for effective optical network planning

Most telecom operators face a lack of information when planning investments in optical network expansion and sales forecasting. This often causes huge expenses without expected revenue while strategic decisions are based on observations but not on available data.

Building a fiber optic network is an extremely expensive project for every telecom operator and therefore, optimal network rollout is very important in order to ensure the fastest return on investment. Business goals are always set to maximize the utilization of a newly built optical network, so network planning engineers have to rely on powerful analytical tools to get optimal results. This means that the optical network should first be developed in the areas where the largest amount of customer interest is expected and later in the areas with lower expected interest.

What problems are telecom operators facing when planning an expansion of fiber optic network?

Telecom operators are usually conducting so-called feasibility checks for every service inquiry received from an end-customer. These feasibility checks are, considering the cost of deploying a fiber optic link to the desired customer location,  a predominant cost in most cases, as well as the costs of active equipment, licenses, etc. On the other side, the revenues expected from this customer are projected. Taking all these expenses and expected revenues into the equation, a business case calculator gives an answer as to whether the feasibility check is positive or negative.

In the case where the feasibility check is negative, telecom operators usually must either decline the customer or offer a cheaper service with different technology, such as ADSL or VDSL, where feasibility checks are usually positive by default.

The problem with such an approach is that telecom operators often treat each customer inquiry separately, resulting in a huge number of negative feasibility check results.

A much better approach would be if the telecom operator could somehow cluster those customer inquiries, resulting in the splitting of the total network development cost to multiple customers. It would be a win-win scenario for the telecom operator and for the customer since the telecom would have a much bigger success rate in fulfilling the customer’s requests, and the customer would have cheaper and better service. It means a bigger penetration score, higher network utilization, and higher sales KPIs.

How to make an investment plan in telecom by using Qlik Sense

Qlik Sense is a self-service, visual analytics platform that offers powerful data visualization and research and collaboration capabilities. With built-in advanced analytics, visualizations, maps and reporting capabilities it can effectively address network planning problems, as well as the problem of customer inquiries forecasts, as shown in the following examples.

In addition, during the selection process, filters automatically respond to changes in all visualizations and are applied through all reports without writing additional queries. Also, when zoom out is done, all the sheets are reassembled and adjusted according to default settings.

Example 1: Geographical distribution of customer inquiries for fiber optic service

In this scenario, the Qlik Sense platform has been integrated with the CRM system of a telecom operator, and information about all customer inquiries for fiber optic service have been extracted. Since Qlik Sense maps have the capability to rapidly visualize thousands of objects on the multi-layered map, getting a useable report wasn’t a hard task to accomplish.

Qlik Sense: Geographical distribution of customer inquiries for fiber optic service
Geographical distribution of customer inquiries for fiber optic service

Bubbles on the figure above represent the concentration of customer inquiries for fiber optic service geo-positioned on the map, based on the street address of the customer. Bubbles change size and color proportionally to the number of inquiries from a certain geographical area. As the bubbles are bigger and redder, the concentration of customer inquiries in the area is higher. Bubbles or different marks such as lines, areas, pie charts, bar charts, heatmaps, etc. can be based on any measure, multiple layers at once. From this map, people responsible for making decisions about future fiber optic network expansion can visually see the true demands for fiber optic service in different geographical areas.

Example 2: Geographical distribution of negative feasibility checks

The previous example gives us a good overview of the geo-distribution of customer inquiries but tells us nothing about whether the telecom was able to fulfill those requests, and by what percentage. In the following example integration with a telecom’s workflow system and CRM system has been made, in order to visualize geographical locations where fiber optic service has been requested by customers, but the telecom operator declined the customers because feasibility checks turned out to be negative.

Qlik Sense: Geographical distribution of negative feasibility checks
Qlik Sense: Geographical distribution of negative feasibility checks

In the figure above, we see a selected geographical area clustered in uniform hexagons with a radius of 1 km (0,62 mi) and colored according to the concentration of negative feasibility checks falling within the hexagon. Again, red-colored hexagons represent a greater concentration of negative feasibility checks.

With such visualization, it becomes obvious which areas should be prioritized when making decisions about investments in fiber optic network expansion.

Example 3: Sales prediction for a newly built fiber optic network

In the previous two examples, it has been shown how the Qlik Sense platform can help engineers in the network planning department to make justified business decisions about investments in fiber optic network expansion.  In the following example, we show how the Qlik Sense platform in combination with R programming language can be used to predict the sales results after a network expansion has been made.

In the figure below we took the number of monthly customer inquiries for the period of the previous two years, and with the help of a statistical model written in the R programming language, a sales prediction has been made for the next 6 months. The blue line indicates the actual values for the observed period of time, and the red interval gives us a range of values ​​that will be responsive to the value of the predictor with 95% certainty.

Qlik Sense: Sales prediction for a newly built fiber optic network
Qlik Sense: Sales prediction for a newly built fiber optic network

By using similar statistical models, sales managers can set realistic revenue expectations for sales representatives, and make smart business decisions about further investments.

Good analytics get a high return on investment in a short period of time

The previous three examples represent just a subset of possible scenarios as to how data visualization can help your business grow. Building an optical communication network is always a huge investment for a telecom operator and even basic optimization can lead to huge savings in absolute amounts. Qlik Sense has proven to be an excellent analytics tool that can help network planning departments to see a wider picture, and to make smarter decisions.

Qlik Sense is more than just a tool, it offers embedded and custom-built analytics, guided analytics, self-service data visualization, data preparation, analysis, collaboration, data storytelling and reporting, and it pays for itself in a very short time. It not only pays for itself but also can potentially have a huge positive impact on the annual financial results of a telecom operator.

Need help in this field? Let's talk.