Big Data / Real-Time Data Harvesting

The fact is that yes, big data is just that: a way to store very large amounts of unstructured and structured data. What makes Big Data valuable to the business are what mCentric call

Focussed Repositories and Insight Applications that reach into the vast amounts of unstructured and structured data in a more efficient manner enabling the business to take decisions more accurately, timely and in a more granular form.

Focussed repositories are where information from all the unstructured data sources is synthesized in a coherent, continuously updated whole that provides structured visibility of key attributes relating to the focus area while retaining easy access to reach back into the unstructured data at any time incorporating new discoveries into a given focus area. Probably the most common example of a Focussed Repository is the Subscriber Repository.

Insight applications provide tangible value to the business by exploiting a Focussed Repository. These can be either off-the-shelf Applications from mCentric, or they can be built to order, based upon Client requirements.

Applications, Real Time Analytics and BI:.

mCentric goes a step further in Big Data treatment, beyond simply indexing, clustering, mining, they make it possible to take advantage of data at a more sophisticated level, practising what is called Data Harvesting.

mCentric knows that the applications that can be built are only as good as the data available to them, thus the true value to the business is provided by Automated and Continuous Learning while synthesizing data into customer behaviour, creating an always relevant profile.

The Insight Applications are thus empowered to provide agile, differentiated information to business managers and marketeers enabling better informed decision making, campaign design etc... The Applications can then concentrate on the job of predicting churn, retaining clients, segmenting or micro-segmenting the user base, providing accurate , up to the minute information to Law-Enforcement Agencies, Communication Ministries, Customer Call Centres, to name but a few.

Additionally Analytics has a fundamental role to play. Many rules and relationships are "built-in". However, it is also true to say that every mobile customer has different requirements. Every day we are seeing new trends and our analysts are finding new customer Habits that can be used as a basis for new rules, permitting new ways to find, for example, customer value networks and the corresponding group influencers, all of which can be dynamically introduced into the Repository.

mCentric harvests data by putting into play its Insight Applications, some of which are listed and explored briefly below.

  • Micro-Profiling
  • Law Enforcement Agency Support
  • Instant Analytics
  • Churn Prediction
  • Promotions
  • CEM 360º
  • xDR Navigator
  • KPI Dashboards

Go to Top

The following are real business examples which mCentric has rolled out in customers worldwide.

Profit Guru

mCentric's ProfitGuru, promotions engine goes where traditional IN/OCS and Billing can't go and thrives in operator infrastructure environments that are from a mix of vendors. ProfitGuru automates the end-to-end process of identifying, targeting, managing and analysing the promotions life cycle. ProfitGuru enables automatic targeting & qualification analysis and consumer participative messaging, helping an Operator maximise benefit both to the Operator itself and to the Consumers.

Continuous Micro-Profiling permits the creation of detailed, individualised profiles of each and every network subscriber. These profiles can be used in a variety of ways either by other Insight Applications or other operator applications for example any of the web self-care portals or Mobile Operator microsites, sales sites, etc.. Since the profiles are being updated continuously the down-stream systems have an opportunity to interact with the subscriber with more relevance and in a more spontaneous manner.

In summary it:

  • Enables a complete view of each and every customer
  • Improves customer engagement and click-throughs and increase revenues with personalised targeting and offers, in real-time.
  • Improves the quality of campaign targeting and hence improves ROI
  • Improves fidelity and retention due to being able to permanently provide relevant services and promotions
  • Provides promotions based upon device used to connect or device history
  • Incorporates sentiment and trend analysis

The Operator benefits from being able to:

  • Maximise the profit produced by their Consumers
  • Accelerate the uptake of new products and services
  • Ensure that Consumers remain loyal to the Brand and indeed serve as an important source of new Consumers
  • Have an up-to-date view of the network performance in business terms

ProfitGuru provides a complete dashboard, not only of parameters related to promotions but also, regarding the business impact of the operation of all the products and services offered over the network, empowering executives to take more effective control of business operations.

Traditionally, focus is simply upon usage and the last month's performance. Consumer behaviour is clearly far more complex than that and is influenced by many more factors. ProfitGuru enables the operator to create a complete picture of a Consumer's behaviour, and hence, better understand automatically their needs and tendencies.

What revenue generating, terminating traffic did the subscriber generate? What did they do this time last year? Many of our habits and behavioural patterns have strong seasonal tendencies. Hence it is important to be able to consider long term trends when making decisions. ProfitGuru permits evaluation of behaviour over any length period. Thus such considerations as what did this Consumer do at this time last year? or what have they done for the last 12 months? may be simply and automatically included in the process.

In summary, ProfitGuru gives the Operator control over the stimulation of revenue growth simplifying the definition and automating the execution of promotions to the Consumers, and removing all the generally used, reiterative, time-consuming manual processes.

Go to Top

Real Time Analytics

At a superficial level, the Analytics Application can provide ad-hoc query and reporting functionality to "Explore" the universe of Big Data. In more depth, these are the tools that will be used by analysts to discover trends, relationships and potential anomalies in, for example, network performance, subscriber behaviour, charging, geographical service or activity dispersion, and a large number of etceteras.

In order to accelerate such analysis, mCentric has developed a combination of methodology and technology. Based upon our experience and upon analysis of data from operators on three continents, we have developed a technique for synthesising a Statistically Representative Mini DataBase (SRMDB). The Analytics Application can work both against the whole data universe, or against the SRMDB. Initial analysis is thus performed against the SRMDB to facilitate the fastest possible turnaround time for the "Discovery, Postulate, Query, Analyse, Discover" cycle. Once the analyst has "found what they were looking for" or feels that the they have the desired segmentation, the tool then permits switching to using the entire universe to retrieve the proverbial "needle (read customers) in the hay stack".

This approach not only reduces discovery times but, also lightens the load on the service leading to more available compute cycles and better performance all around.

Go to Top

CEM 360º: Measure, Communicate, Act

mCentric’s Customer Experience Measurement, Communication and Action system is the first end-to-end solution that truly delivers real time, hyper-precise QoE/QoS information to the operator whilst empowering customers to improve their experience on the mobile operator’s network.

A comprehensive and transformative approach to customer experience management can unlock value across operator networks and product offerings, as well as positively transforming customer relationships.

Taking QoS Measurement Further

mCentric incorporates these approaches into its global solution, taking and synthesising information from any available sources such as OSS/BSS and network probes.

Customer Centric Measurement. mCentric’s technology converts each device on the network into a mobile QoS/QoE measurement station whilst ensuring the Customer is always in total control of what information is gathered.With its Intelligent Network Probe, mCentric delivers an inline network probe that provides fundamental insight into not only how the network is performing, but also its relationship with other networks such as the internet.

The Customer is empowered to communicate, providing spontaneous opinions regarding the QoS Received, similarly to many Best in Class internet services such as Skype or Google

The Customer is enabled to take Action. All traditional actions, such as more effective investment in infrastructure, take time to have any discernable effect upon Customers. Here too, mCentric has gone further to enable Customers to take immediate action to improve the received QoS. Guaranteed voice connections. TurboCharged Data connections. Take the decision to charge or provide to stimulate loyalty.

Go to Top

xDR Navigator

xDR Navigator is a powerful Insight Application which allows Call Center Reps to have all relevant subscriber activity at their fingertips in a matter of a few seconds. Agents resolve complaints in record time and subscribers avoid long queues waiting for a representative. Agents may be confident they are going to get the answers quickly and deliver accurate information to the enquirer. The subscriber comes away satisified with a quick resolution to doubts.

Admins map raw data coming off the IN to user-friendly commercial language in real-time so that CSRs have the latest descripions of the issues which subscribers will be asking about. Built right into the system is a way to identify new types of raw data and flag it as a necessary item to be mapped for the appropriate originating system.

Go to Top

Law Enforcement Agencies

Within the xDRNavigator framework, mCentric presents its LEA solution. Law Enforcement Agencies (LEA) World-Wide require an ever greater amount of information from Operators, and frequently they want it "now" with strict deadlines and penalties set for non-compliance. Given the volumes implicit in storing all the information regarding network users activity, Big Data provides an appropriate way to store it.

However, once again we are faced with the problem of how to make all this information useful; how to provide/expose it to the LEA, simply and effectively.

  • NAT'd IP to MSISDNCorrelation

    - ID subscribers using a particular public IP during a particular time period

  • Domain Watch
  • - Which MSISDNs have visited a particular domain (honey pot) by date range

  • MSISDN Trace

    - Which Domains has a particular MSISDN visited by date range

  • MSISDN Watch
  • - Generate notification of all activity for an MSISDN in real time

    - Forward copy of all traffic to LEA (includes IAP provisioning)

  • Voice/SMS
  • - Activity history for a particular MSISDN for given date range

Go to Top

Micro Profiling

Continuous Micro-Profiling permits the creation of detailed, individualised profiles of each and every network subscriber. These profiles can be used in a variety of ways either by other Insight Applications or other operator applications for example any of the web self-care portals or Mobile Operator microsites, sales sites, etc.. Since the profiles are being updated continuously the down-stream systems have an opportunity to interact with the subscriber with more relevance and in a more spontaneous manner.

In summary it:

  • Enables a complete view of each and every customer
  • Improves customer engagement and click-throughs and increases revenues with personalised targeting and offers, in real-time
  • Improves the quality of campaign targeting and hence improves ROI
  • Improves fidelity and retention due to being able to permanently provide relevant services and promotions
  • Provides promotions based upon the device used to connect or device history
  • Incorporates sentiment and trend analysis
Go to Top

Churn Prediction

mCentric's Insight Application for the identification and benchmarking of churn prediction algorithms leverages Big Data's massive parallel processing capabilities to efficiently execute historical analysis to enable the discovery and evaluation of the relevance and impact of Contextual and Subscriber behavioural attributes on Churn.

Every market is different, however our experience leads us to believe that at least the following attributes should be evaluated: Spend Evolution, MO/MT & Onnet/Offnet Relationship Evolution, Offer, Price, Customer Relationship and Experienced QoS.

This Insight Application permits, based upon a definition of "what identifies a churned Subscriber", a dynamic historical evaluation of the evolution of the selected attributes. Two periods are identified: the first from which to extract the set of target churned subscribers and a second to define a period over which the evaluation will be performed. The IA then analyses the attributes and reports upon the most likely drivers for churn discovered, with their respective weights. It can also provide automatic guidance with respect to what measures should be taken when, i.e.: what attributes are on the critical path and when they must be stimulated to avoid churn.

Furthermore, this information can be correlated with contextual information such as frequented locations or time-of-use correlated to network issues.

All information is presented dynamically in a variety of graphical formats permitting instant feedback and thence the tuning of the algorithms. The analyst can select to add any of the attributes that are available in the repository for incorporation into the historical and real-time analysis.

Once the analyst feels they have found clear indicators of a tendency to churn, these are marked and the tool permits the benchmarking of the algorithm against selected period(s) to validate the quality of the prediction.

The Algorithms are self-tuning in that Continuous Real-Time Analysis can adjust the weight/score of the Churn drivers based on "recent relevance" events; (browsing competitors sites to discover competing offers, high-end smart phone subscriber regularly receiving data QoS inferior to his expectation, frequented locations + time-of-use correlated to network issues, etc.)

Go to Top

KPI Dashboards

In addition to the Insight applications, that are discretely targeted at specific business functions, Key Performance Indicators Dash-board accessible from a browser, provides executives with up to the minute information regarding the performance of their network in business terms, in a simple, didactic format. All results can be exported to Excel or PDF for distribution as required.

Many elements allow the dynamic application of filters to modify the appearance and content of the data presented, e.g.:

  • Period (day, week, month)
  • Location (Region, State, town, ...)