Home Tags Posts tagged with "machine learning"

machine learning

Artificial Intelligence (AI) is certainly a popular buzzword in the industry today and according to surveys by TM Forum, many have now “deployed” AI and are looking to scale. But deploying AI means different things and it depends on the strategy adopted by the organisation for implementing AI.

For some, a proof-of-concept (POC) has been successfully executed and a model has been created that uses machine learning (ML) algorithms. For example, in the context of telecom customer value management (CVM), this is often around churn or decay prediction, and the date and value of the next recharge.

For others, it means the technical capability has been put in place to build, evaluate and execute ML-driven models across multiple use cases – our own AI at Scale platform is a case in point. It can be used for building, evaluating and running in production both CVM and non-CVM models.

Typically, the AI platform will sit on top an existing big data lake or incorporate a data fusion layer to capture data from multiple sources (covering real-time and batch, structured and unstructured data).

A fully functioning AI platform provides a graphic user interface (GUI)- based workbench for data scientists to rapidly build models using ML algorithms. It also supports the full lifecycle of model creation and execution; data ingestion, data exploration, feature engineering, model development and evaluation, and deployment and execution in production.

But, along with the technology challenges associated with scaling for AI, there are larger implications for the “people” and “processes” that go along with it.  Often, our clients will point out that one of the biggest challenges they face is the hiring and retention of key talent in data science. Understandable, really, as data scientists are in a “hot” market. This is why companies will often turn to partners to assist.

In fact, this people-centric factor is not only limited to the data science team. In the context of CVM, marketers need to adapt. It is one thing to be designing and implementing campaigns that are based on business rules created by a marketer; the criteria for the segment, and the offer to make to that segment. It is quite a different thing to be prepared to leave this to a “black box” solution.  This is exactly what a ML model is. It is next to impossible with a ML-driven model to determine exactly why a particular offer is made to a particular customer.

Therefore, a scaling strategy is required, that allows confidence to be built. For example, allowing the model to apply to a proportion of the base while the traditional business rules approach is applied to the remainder. When positive results are seen, then the proportion of the base that receives offers based on the model can be increased.

An ancillary consequence of this is that the robustness of the methodology for measuring performance is critical. There must be confidence that when an improved result is seen, it is trusted. This is why effort is required to ensure the universal control group (UCG) is highly representative of the base. Performance is measured as the difference between the UCG and the universal target group.

The “process” side is also extremely critical. This covers the governance and practices in place to ensure data integrity is maintained, and that data is updated when expected. Critically, it also covers those processes associated with putting a new model into production. This DevOps side is particularly challenging for many organisations because there are new practices to be developed. A ML driven model is not constant. By its very nature it changes while in production, which is very different from what we see as “normal” software deployment. In fact, this is so different that the TM Forum has a Catalyst stream in progress for “AIOps” to develop frameworks for supporting the operations of AI. Engage with this Catalyst programme to keep abreast with the thinking as it develops. And perhaps become a contributor.

November 28, 2019 0 comment
2 Facebook Twitter Tumblr Pinterest

The digital world is growing for telecom operators. With advanced options and the opportunity to get shares in a rapidly advancing market, operators are realizing the impact of improved operational activities.

Understanding the Opportunity

This enhanced interest in the digital market does not come without any benefit or costs for operators. The market presents excellent revenue gain potential, but only for telecom operators that are willing to make the necessary investments. With customer demands also increasing, it is no secret that the expenditure on IT and the overall revenues made through operations are expected to increase for operators.

These findings have been included in the most recent ICT insights survey by the research firm Ovum. According to the survey, a majority of 70 percent operators are planning to increase their net IT expenditure over the course of the next year. Almost 30 percent of these operators are planning to increase these costs by either 6 percent or more. This increase in net total spend for the IT sector indicates the fact that telecom operators are beginning to realize the potential that the market holds.

This growing increase in the number of companies increasing their IT budget is expected to have a cascading impact on the revenue generation of the OSS/BSS domain globally. Ovum has also indicated that the annual revenues from the OSS/BSS market are expected to grow to $22.5 billion in 2022 from $17.5 billion in 2017. This CAGR of 5.8 percent is one of the best in the market and makes the telecom industry a hot property for revenue generation in the future. Current telecom operators with their established infrastructures have a great opportunity to latch on to these challenges. Read on to find out how.

Delivering Next Generation Services

To deliver next generation services, operators need to step out of the realm of what governs the industry currently and implement futuristic methods of governance. This movement towards the future is not just confined to keeping pace with technology trends and chalking out competitive and updated strategies. This step towards the future entails a lot more, such as the eagerness to update to the infrastructure of next generation technologies, such as 5G, network functions virtualization, and the Internet of Things among many others. The use of these technologies in creating smarter networks will dictate the future pace of development.

Simply put, since the telecom operator industry has an intensely competitive environment, operators need to use all up-to-date methods to stay relevant in the picture. The underlying idea behind this is to shift from the traditional system-centric approach and go towards a more end-to-end service-centric one. Operators need to revise their current strategies in a bid to incorporate smarter services that:

  • Are tested in the market for scalability and agility
  • Include a large industry ecosystem that is aimed at new customer markets and segments
  • Facilitate the growth of personal and real-time communication.

Understanding OSS/BSS and What Ails Them

OSS/BSS stand for operation support system and business support system respectively. These two systems work in tandem to support a wide range of telecommunication services and make life easy for telecom operators.

Operation support systems usually assist in the management of communication networks. These tools help in the coordination of resources for designing, building, maintaining, and operating communication networks.

Business support systems help organizations to reach out to their customers and create an offer for them. These offers are personalized in order to give special offers to each customer based on their preferences and usage patterns.

OSS/BSS systems currently have a long way to go before they can support this digital transformation. Most operators have a complex maze of legacy systems that are not compatible with the requirements of these technologies. To maintain a competitive edge, operators are now looking at and thinking of planning beyond the traditional KPIs of profit margins, revenue, etc. Instead, we now have criteria such as efficiency, customer experience, agility, quicker time to market, and seamless product delivery in the mix.

However, with the current extent and limitations of legacy systems, operators cannot meet these requirements. They’re falling short on numerous counts, including the duplication of information, which adversely impacts the time to market for several products.

Underlying Trends and the Journey so Far

Regardless of the implications, operators have come a long way in the implementation of OSS and BSS strategies. Some underlying trends are expected to phase out over the course of the next couple of years. It is initially expected that software as a service (SaaS) stacks will rise to the fore. Operators realize the impact of competition and are looking for cost-efficient stacks that can be rapidly deployed at any given time.

Moreover, it is expected that the revenue from analysis-based tools will significantly increase over time. This change will be driven by the maturation of AI and machine learning tools that enhance the analysis process over time.

The Changes in Play

The changes brought through digital transformation are already in play and can be seen across multiple industries. Here are three OSS/BSS examples in major industries:


In retail, addressing business activities like sales and inventory management are still a significant challenge for operators. This is how the BSS system can facilitate in this industry.

  • An efficient POS system. Point of Sale systems play an important role in the overall buying process. The addition of a CRM in these systems enables enterprises to configure the mode of payments and price lists, and to restrict individual channels of payment.
  • The digital solution plays an important part in executing end-to-end management of logistics at the backend. Finding the correct product mix with optimal inventory and store capacity is the final aim.
  • Retailers work with a broad ecosystem of partners, and a BSS solution helps them to manage this ecosystem efficiently. From managing invoices to setting delivery details, the system works best when creating and following schedules.

Enhancing the Usage Pattern of Customer Data

A BSS can be used across industries because of its ability to help organizations enhance the usage pattern of customer data. By augmenting services such as service bundling as well as switching and gift services, the BSS/OSS improves the selling process based on insights from customer data.


The following changes can be influenced by a BSS across the banking sector:

  • Simplifies the customer acquisition process for banking services.
  • Lead collection systems are facilitated.
  • Customer complaints are dealt with in a professional manner to mitigate the negative PR that might arise out of them.

Achieving the Transformation

The passage towards a modernized OSS/BSS system can come through multiple efforts.

The Enablers

These three key aspects of the digital transformation have to be at the core in order to enable the transformation towards modern OSS/BSS systems:

  • Enhanced customer experience
  • Operational agility
  • Focus on platform play

Models to Drive Change

To drive change in the customer experience, operators need to be focused on:

  • A cloud based server for micro-services.
  • Real-time communication, or the provision of real-time analytics.
  • A layered architecture that hinders all forms of data duplication.

Download the Mahindra Comviva E-Book to learn more about the digital transformation for telecom operators, and how you can be a part of it.

October 15, 2018 0 comment
3 Facebook Twitter Tumblr Pinterest

If you look back at the history of technology you find that what we know is not often what will be.  What we forecast is likely though not often the result.  Only one can guess how many trillions of messages of all types will be sent and shared by 2020.  Forecasters say close to 3 trillion messages will be sent annually but if you look around the room and around the globe you only see the tip of the iceberg in user growth and use cases for messaging.  When I grew up we had one rotary telephone and one TV.  Now everyone has a TV in their hand and messaging types beyond your wildest imagination bombard you like hurricane.  Sometimes SMS, chat apps, videos, email, and realtime meetings are all occurring at the same time.  Looking forward even a little and beyond our traditional thinking, if digital transformation takes place in one industry can we not expect it to have a ripple or tsunami effect in many others. For example, take a look at Mahindar Comviva’s white paper on Digital Transformation and you will see an interesting view of the changing world of messaging.  In another sector, IoT-internet of things devices will also be sending messages to me and everyone else on their chat lists to let people know what is going on.  Are you beginning to see the compound or exponential growth in the world of messaging?  Looking at this from another angle that of a provider or operator there has been the long “race to zero” where once high fees were charged for SMS, long distance and other services are now free.  Now fee-based services are bundled together so the customer sees value in the overall benefit rather than any one feature or app.  New fee-based features may arise though the business models may vary whether subscription or advertising derived revenues can level the playing field or even give rise to sustained revenues.  Linkages with other business all aiming for the customer will be the new “middleman” in the mix.  That is, companies who link with other companies to message you about things you need or want.  Some are obvious like Amazon buying Whole Foods to bring all their other business to you whether you walk in the door, home delivery or even by drone in the future. At the same time there are tens of thousands of startups and others that are building unique and valuable solutions to solving specific problems with innovative technologies such as artificial intelligence, machine learning and wireless tech. I like to think that there is a new wave coming quickly of technology embedded literally in everything we buy including food, medicines, wearable and anything else that we do that will improve our lives in many ways we haven’t even thought about yet.  For example, we really don’t know much about what we eat and its impact on our health.  Yes, we know there are a lot of bad things but we don’t know conclusively if we eat one thing for twenty years that it will give us cancer or that it will help us live a lot longer.  There are so many medical issues that via IoMT-internet of medical things that may be able to improve our lives, give us more energy and less live-threatening diseases later.  I like the idea that my IoMT device will give me a warning or even a nudge when I eat something I shouldn’t or say that my weight-loss goals could be achieved if I did more of one kind of exercise than another.  You can now see the world of messaging goes far beyond us chatting with each other which opens up even more possibilities than I have time to discuss today.

September 11, 2018 0 comment
0 Facebook Twitter Tumblr Pinterest

Perhaps the biggest challenge while penning (yet another) article on customer experience and value management (CXM) is not repeating oneself. After all, how does the audience stand to gain from reading ad nauseam about how defining a sound CXM strategy is a good idea? Or, for that matter, how to best measure the success of such strategies?

The point is that one can only emphasize half a million times about how CXM is not a boxed solution. Here’s what ought to be highlighted instead; projects undertaken in one part of the world cannot and more appropriately should not be replicated as is in another. Each part is different and even within one, customer preferences vary greatly compared to those in the same region.


The ASEAN Puzzle

The Association of Southeast Asia Nations (ASEAN) is an interesting example in this context. For one, the region is a mix of contradictions. Take, for instance, customer segmentation. The region not only houses an array of cultures, but multiple income levels across varying geographies as well. In other words, the customer base in this region may well be considered a pyramid-like structure. Customers belonging to the middle-class and international corporate segment are demanding advanced services. These are perched atop a large base of low to-middle income consumers in both urban and rural areas. This segment largely leans towards prepaid services. Overall, the region displays medium to low average revenues per user (ARPU). Meanwhile, some markets are overly saturated, with penetration rates greater than 100 per cent.

Naturally, then, operators have their work cut out for them. A difficult task, no doubt, especially since these players have, by and large, stuck to traditional methods. The classic ARPU model is still the benchmark, as opposed to focusing on improving overall share in the customer’s wallet. Likewise, they choose to

stand apart from the competition by offering a vast service bouquet, priced very competitively.

The catch here is that these methods (while sound in their own right) aren’t really relevant anymore. Customers today are no longer impressed by a differentiated service bouquet-this is now considered a given. Optimum quality of services is a prerequisite, as is value for money. Factors such as data speeds, service quality and minimum response time are what make customers tick. As a brief side-note, it would be worthwhile mentioning a recent customer survey carried out by Bain & Company. The survey covered prepaid customers in the five largest economies in the region- Indonesia, Thailand, the Philippines, Malaysia and Singapore. The results suggested that this customer segment wanted strong voice connections with no dropped calls, fast video downloads and hassle-free social media access. In other words, usage experience was the most important factor influencing how they viewed the operator, above value for money, plan selection, service and corporate image.

Operators, please sit up and take note. Merely slashing prices will no longer ensure you a seat at the leader’s table. Ensuring a differentiated and optimum customer experience will.

What Ails Operators in the Region

Why operators in the region ought to focus on CXM is simple. Customer expectations have increased manifold, especially as an operator’s network expands to include high-sensory applications such as music and video. In this context, operators have to ensure that these experience-led applications function effectively on their networks. In fact, CXM itself has increased in complexity and functionality. This, in short, makes it more important to deploy tools and capabilities to effectively manage customer interactions and touchpoints.

So, the challenges facing operators in the region broadly include:

  • Low Efficiency: Cost efficiency and service quality is what makes all the difference.
  • Improved Productivity: These players require decluttering of their service bouquet.
  • Varied Focus: Operators by and large tend to focus on post-paid subscribers, even when the market itself is predominantly prepaid. This is because the former mean higher ARPUs and retention levels. This, however, may not prove effective in the long term.

Here’s where advanced analytics and machine learning step in. CXM-driven service modelling and offering personalized services in real-time is the ideal way forward in ASEAN. The idea is not to add more services but to zero down (and

focus) on what the customer really wants. The ARPU may not register a meteoric increase, but the potential to enhance revenue is significant.

Operators would do well to remember one crucial point. It isn’t about how much data is available or how frequently it is refreshed and processed. It is about garnering actionable insights from that data plie. Making sense of that data and leveraging that for customized and real-time marketing to drive customer experience. The bottom-line is for operators to keep in mind that the scope of CXM itself is very broad. The idea is to drill it down to specific domains, sub-domains and use cases, one that best provides the most optimal return on investment.

Measuring the Success of CXM Initiatives

Overall, while the measurement mechanisms vary and are usually not very standardized, Net Promotor Score (NPS) is still a critical metric. This is most commonly used by operators to assess the outcome of CXM projects. Simply put, it is externally referenceable, customer-sourced, and can be applied at the brand, product and business process levels.

Of course, all organisations have a series of metrics derived from systems and business processes to track performance. These are typically correlated with NPS to identify drivers for improving the same.


Net, net, the telecom game is no longer about how much a customer spends on an operator’s network. It’s about how long they stay on your network. Deploying a CXM strategy will certainly help an operator keep their ears to the ground. After all, isn’t knowing what the customer wants the bottom-line?

August 30, 2018 0 comment
0 Facebook Twitter Tumblr Pinterest

Digital LATAM Ripe for CVM Evolution

The volume of data generated every day in Latin America is enormous. Most of the countries in the region are in the middle of a digital revolution due to high mobile penetration, expanding IoT ecosystem, rising adoption of cloud, social media, and e-government initiatives. This data revolution is allowing banks and telecom companies to leverage Big Data and Analytics (BDA) capabilities for creating value for their customers, which in turn, promise to bolster their revenues while bringing down costs.

Telecom LATAM: Going digital

According to Analysys Mason, the telecom industry is going to grow to $140 billion by the year 2020. Mobile data service will overtake voice revenues by 2020, as it will be fueled by Latin America’s increasing internet and smartphone penetration. Similarly, LTE is making a big impact in LATAM, and it is expected that by 2020 most mobile SIMS will be on LTE. In Brazil, Chile, Peru, Venezuela, Argentina the rise of data consumption (on LTE) and fixed broadband services will drive service revenues. However, all markets are not on a similar growth path, with some growing at a slower pace than the others.

Segmenting the Markets

Basis data collected from GSMA Intelligence, the markets can be segmented into the following –

Growth Markets

By the year 2020, Latin America is expected to add 60 million new subscribers. Mexico (31%), Brazil (22%), Colombia (12%) and Peru (9%) will grow the fastest in the region. Growth catalysts include Latin America’s economic recovery, as well as the increasing uptake of mobile data services in the region..

Stagnant Markets

According to GSMA forecasts, growth will slow down or stagnate in Argentina, Guatemala, Nicaragua, Uruguay, Chile and Venezuela. The economy and saturated markets will contribute to slowing down of growth in the region.

Competitive Market

Of all markets, the Brazilian telecom market shows the lowest spread in Market share and ARPUs, which characterizes it as a highly competitive market. On the other end of the spectrum, we have market monopoly in countries like Mexico, Uruguay, Colombia, and Venezuela, which are dominated by a single operator.


In most developing markets telecom companies are struggling to maintain profitability. The market in Latin America is no different. The high cost of customer acquisitions, rising price wars, and increasing customer churn are some of the factors contributing to low ARPUs across the board. The years 2011 to 2017 were marked by new customer acquisitions, with double digit growth in several markets. However, in 2018, it is increasingly becoming clear that the industry’s long standing orientation for customer acquisitions have not worked, which is evident from low ARPUs and churn in the markets. The time has come for operators to strike a fine balance between acquiring new customers and catering to their existing customers in order to increase their profitability.

The advantage of CVM is that it is an end to end holistic solution that can be fitted to customer acquisition, retention or even up-selling and cross-selling in growth markets as well as markets showing the characteristics of having attained maturity.

Growth Markets

Brazil, Mexico, Colombia are adding new customers every-day. Telecom ARPUs in these markets are low because of the subsidies given to new customers, as well as retailer commissions. According to Strategy & analysis cutting off the payback period at eight months will yield greater than 40% saving on customer acquisition costs. In these markets, CVM solutions will help to evaluate the price elasticity of certain customers with regard to subsidies and credit which will help in reducing the payback period for new customers

Saturated Markets

In many of these countries, the market is saturated. Market penetration, which is estimated in % unique subscribers, is over 90% in Argentina, Chile, and Uruguay. By 2020, we expect more countries, like Mexico, Peru, and Colombia, to grow their base of unique subscribers. Once markets becomes saturated, it will become necessary for telecom operators to shift their focus towards retaining customers through dedicated, one to one campaigns, based on a holistic understanding of the customer’s past behavior.

Stagnant Markets

In stagnant markets, it is no longer profitable to increase market share. Instead, telecom companies must look to increasing profitability from their existing customer base. In general, the operators will have to strike a fine balance between aligning channel activities, optimizing customer contact and improve product portfolios.

What can Analytics do as of now in a predominantly pre-paid market?

Big data and analytics capabilities are helping operators to devise customized services, pricing plans, marketing campaigns, promotional strategies, with less effort and more success.

Since the telecom market in LATAM is largely a prepaid market, low credit can potentially lead to millions of dropped calls or terminated data sessions, which can easily lead to loss in operator revenues, or even worse, when the customer churns to the operator’s competitor.

An end to end CVM solution would leverage big data and analytics to mine actionable insights from customer data, and using that information to offer relevant and contextual services through the customer’s preferred channel.

This would mean capturing customer events on a real time basis, running analytics based on customer profile, and then orchestrating services like marketing automation, call completion services, e-payments, and financed telecom services that will help the subscriber to enjoy uninterrupted services on their mobile devices.

The use of AI and Machine learning will only help in optimizing the services further by optimizing services as well as delivery across the subscriber’s preferred channels.

LATAM Banking: Trending towards Inclusion

The year 2016 saw the exodus of a number of international banks from Latin America due to weak earnings, continuing losses, and stringent banking regulations.

This provided regional banks with a great opportunity to expand their business by increasing their focus on their customers. Key challenges before banks include acquiring and retaining “profitable customers” while keeping costs under check. However, all markets are not the same. Some are maturing fast, while some are still in the initial stages of growth. Basis data collected from Findex 2018, we’ve segmented the markets on the basis of their growth state and potential.

Matured Markets

These markets have grown the quickest in the region, with banking services reaching to over 60% of the population. In some of the countries the banking penetration has doubled in the period 2011 to 2016. The countries that belong in this group include Brazil, Uruguay, Chile, Venezuela and Bolivia.

Digital Adopters

Besides high bank penetration, these markets are characterized by higher than average mobile phone penetration, data consumption. Indecently, these countries have higher uptake of digital payments compared to other countries in the region. The countries in this group include Chile, Brazil, Venezuela and Uruguay. Digital laggards will include Mexico, Colombia, and Peru.

Growth Markets

Many of the countries have low banking penetration because of the lack of access to branches, or the lack of funds. Mexico, Colombia, Argentina, Guatemala are some of the countries that can be put into this bracket. However, these countries have a high mobile penetration, which provides an opportunity to leverage digital channels for providing banking to all.


CVM solutions are very flexible as they are applicable in a growing market as well as an emerging market. In a growth market CVM solutions are allowing banks to improve the payback period of their new customers. Similarly, in advanced markets, CVM solutions allow banks to explore newer source of revenues and increase their return on investments.

Growth Markets

In the growth markets of Argentina, Mexico, and Colombia the rise of Digital Only Banks promises to bring banking services to the financially excluded. Digital only banks will be in a better position to tackle major roadblocks to banking like lack of access, funds and trust, and thus helping in providing banking for all. CVM solutions will help to cater to these prospects in a cost effective and profitable manner by providing contextual, one to one services, like savings, loans, credits, transparently, without friction.

Matured Markets

In the advanced markets of Chile, Brazil, Venezuela, Uruguay, CVM can help banks to increase customer stickiness and increase the customer lifetime value. Here, banks can leverage customer data generated across chat services, facebook, SMS, email for creating a more engaged customer base. The benefits would be the following –

  • CVM solutions will help incumbent banks to deliver an Omni-channel, personalized and responsive digital experience to their customers.

  • It will allow banks grow value from their existing customer base by influencing monetary as well as relationship metrics.

  • CVM will provide banks visibility into customer intent and leveraging this information to influence customer’s behavior in line with business goals.

  • It will help in creating stickiness and customer value through upselling, cross selling new products and services like loans, and insurance.

A good example here would be a customer using his card or mobile wallet for online or offline payments. As soon as the customer swipes his card or pays for online goods, the banks will be able to initiate a real time offer or promotion that will help to incentivize usage of the card or digital wallet.

Similarly, in competitive markets, like Venezuela, Chile, CVM will help to differentiate through innovative services and products like loans, tailored savings plan, loyalty point management. In Chile, a leading bank is allowing its customers to redeem points at international online stores. In Venezuela, a leading bank is providing end to end loan applications on their mobile app, which is aligned to their customer’s contextual needs and requirements.

What can Analytics do in this market?

Big data and analytics can help banks in LATAM to measure near term gains for individual customer, as well as estimate future opportunities in growth markets. In matured markets, it helps in increasing wallet share by identifying relevant cross sell and upsell opportunities. Targeted marketing based on a better understanding of the customer through analytics will help to migrate the customer from less profitable relationships to more profitable ones.


For banks and telecom companies looking to differentiate themselves in an increasingly competitive market, with low ROIs, leveraging CVM has become essential to winning and retaining customers. In the coming days, one can expect Banks and Telcos using AI/Machine learning powered CVM solutions to get that one bit closer to their customers, with contextual and personalized solutions that are meaningful and that relates to the what the customer needs.

August 30, 2018 0 comment
0 Facebook Twitter Tumblr Pinterest