BIG DATA ANALYTICS - VanillaPlus
BIG DATAANALYTICSCan CSPs handle the complexity?Gold sponsorsSilver sponsor
IntroductionBig data has a broader meaning than the words suggest. Big data isless about database size and more about managing the hugecomplexities that exist in the networks and telco business.Communications service providers (CSPs) aredesperate to gain greater control over the big circusthat’s happening inside their tent. The global, overthe-top internet, broadband and big mobileextravaganza they support truly is the biggest showon earth.The risks are huge, of course, because the showgoes on whether the CSP makes any money or not.So the big data trend is almost a rescue mission tosave telecoms from: The onslaught of mobile broadband trafficgrowth The explosion of smartphones and apps thathave turned troubleshooting upside down The relentless margin pressure and revenueencroachment of rivals and over-the-top players.VanillaPlusIf the telephone operators of yesteryear knew theywere being groomed as ringmasters of such a wildcircus, they would have never accepted the job. Butthey did take the job and the rest is history. Nowcomes the challenge of profitably orchestrating allthe video, voice, data, elephant, tiger, clown andacrobatic acts that operate under the hugecommunications tent. The circus performersthemselves, of course – Google, Facebook, Netflix,Apple and many, many others – are often richly paidwhile the ringmasters face an uncertain future.Unfortunately, the network equipment providers areas new to the game of managing huge mobile datavolumes as the CSPs themselves. Sales of LTEnetworks are heating up for sure, but the essential‘How to Profitably Operate and Profit from an LTEThe author,Dan Baker, isresearch directorat TechnologyResearch Institute21
Network’ instruction manual is still just a work in progress. Thatmeans there’s plenty of opportunity for fresh companies to stepin and deliver what CSPs need.A flock of solutions vendors has emergedTo better understand this big data trend, TRI began a fullresearch investigation in mid-2013. I was frankly surprised athow many new players had come in since I wrote a report onthe analytics market seven years before. In the end, weinterviewed some five dozen carrier and vendor experts,publishing profiles on 41 active players.Big data is rooted in analytics, of course, but the field hasexpanded far beyond the churn reduction niche that drove themarket a decade ago. Big data basically is a layer of analyticsvalue that rides across all the traditional BSS/OSS transactionsystems. In fact, we figure the biggest solution market in bigdata today is in network and customer experience analytics.All kinds of solution vendors are involved today: deep packetinspection specialist firms like Allot Communications, systemsintegrators like HP, service assurance firms like NexusTelecom, and firms who grew up in the business assurancespace like cVidya.We think plenty more players are going to jump in and acquiresome of the startups – the major service assurance and billingvendors for example. Companies like Amdocs, TEOCO, JDSU,SAP and Comptel have already taken the acquisition leap.And as time goes on, it will become harder and harder todistinguish the new from the old solutions as analytics playersare acquired and the markets melt into one another.22Big data’s fourth VWhile big complexity is the greatest challenge, big data iscertainly about managing huge data volumes too. In manyways, telecoms with their massive networks practically inventedbig data. And plenty of telco use cases fit the so-called three Vsof big data: large Volume, Velocity (speed of analysis), andVariety (of technologies).But there’s a fourth V in the big data equation, and to miss thatfourth V – Value – is to miss a lot because big data drags withit a more open and entrepreneurial way for telecoms to workwith solution providers and create value.In fact, the confluence of cloud, commodity hardware, BYODand outsourcing is creating a new and highly organicmarketplace where value is king. The long term impact is clear:the results achieved become more important than how big asoftware supplier you are, what computing technology you use,and whether or not the IT department sprinkled fairy dust onyour contract bid.The key question becomes: how best to add value at low costand deliver an ROI? Analytics and big data provide the answerbecause these solutions can be inserted at relatively low costand with minimal impact on current operations. In most cases,they require little hand-holding by the IT department.In fact, the biggest customers of analytics are the businessunits themselves or standalone departments like networkoperations.Dining in ManhattanI live two hours’ drive from New York City. Manhattan Island isnot just the home of Wall Street: it's also a diner's delight. Thereare about 3,500 restaurants in Manhattan and every cuisine isVanillaPlus
represented: Indonesian, Ethiopian, Lebanese, Vegetarian, youname it.Now the contrast between the Fortune 500 firms on Wall Streetand all these restaurant entrepreneurs is striking and I thinkthere's a parallel here to what's going on in telecoms software.Being a big company didn’t prevent Wall Street’s LehmanBrothers from going bankrupt so are CSPs headed on thesame perilous path that many bankers faced in 2008?Well, the warning signs are certainly there. One authoritativeone is a 2012 report by PricewaterhouseCoopers (PwC) entitled‘We Need to Talk about Capex’ which singles out telecoms forpoor management of its capex. Putting PwC’s conclusions intomy own words: network investments are out of control in theCSP world. Nobody can accurately tell whether a particularcapex expenditure pays off or loses money, so senior telecomsexecutives think it’s better to err on the side of buying moreequipment. Sounds like a glorious plan to make JohnChambers and company rich.This capex issue is fundamental to the CSP business. It cannotbe just dusted under the table. Overly restrictive policies andclosely held relationships with a few large suppliers are clearlynot getting the job done. Marketing and network operations aredesperate to take action because everywhere they look thereare savvy OTT competitors who seem to enjoy devouringtelecoms revenue and traffic streams for lunch.So these entrepreneur-like analytics firms are delivering exactlywhat CSPs need to shake up the solutions market and helpsolve such long-standing problems in capex and elsewhere.Now, like the restaurant scene on Manhattan: this bigVanillaPlusdata/analytics business is also fiercely competitive and a veryrisky business: only the fittest vendors can survive long term.The exciting trend here is that telecoms software is nowbecoming a results-driven marketplace.And if you’re a buyer of analytics, this is a great place to bebecause you have choices you never had before. This bravenew software world will also be massively disruptive for thoseincumbent suppliers who figure they can sit on their nest eggs.Market drivers and challengesThe biggest driver of big data is the new-found freedom toallow results to decide which solution providers win and whichones lose, Lots of opportunities emerge when there’s a lid puton over-zealous purchasing controls and internal politics.Big data opportunities cut across the entire BSS/OSS stack ofapplications and at every step in the service delivery process.There's also no question that commodity hardware, SaaS,storage and open frameworks like Hadoop and MapReducehave lowered the barriers of entry.Another welcome change is faster speed to analysis. With bigdata, you can employ the freshest data and not have to wait 30days to pull results off the warehouse.Yet another big driver is the ability to access data at a finer grainof detail – and to see the outliers of the data set as well.Analytics used to be a sampling game but today, scanning thedetails is essential for detecting fraud, monitoring mobile abuseand measuring true customer profitability.For weeks, a tier one US CSP experienced a big unexplainedspike in traffic every business day from 9am to 5pm in one of its23
largest cities. Employing a big data solution, they isolated thetrouble to a single taxi company who was abusing an all-youcan-eat plan intended only for credit card authorisations. Thetaxi company’s drivers were using the circuit to pump freevideos to customers sitting in the backseats of taxis throughoutthe city.By living in the outliers – the data to the far left or right of thebell curve – the taxi fraudsters escaped notice. Sampling couldnot detect them. And that’s precisely the value of big data – tomake visible countless details of the business where so manyfraud control and optimisation opportunities live.The big data market is not all rosy for suppliers. First andforemost, the large flock of analytics firms out there putspressure on price, and that's a very good trend for serviceproviders.Big data vendors are also taking lessons from the previousgeneration of analytics software which quietly went fallow in thelast decade. Continuous and sustainable analytics programmesare the key and here, firms are getting creative by marryingsoftware with in-shop marketing consulting, managed services,and risk/reward compensation programmes.Yet there were troubles with this approach: 1) high cost; 2) thelong time it takes to create value; 3) the risk of a programmefailing to deliver sufficient value; and 4) the many knownpockets (or reservoirs) of value became too expensive topursue because of limitations around approved suppliers,authorised database and computing technologies, data sourcesin the data warehouse, and IT staffing to support new systems.But in the big data era – the “After” diagram, opportunitiesopen up. Rather than relying on expensive, monolithictransformation programmes alone, the operator can also fundmany smaller programmes to pursue pockets of value acrossthe business. The margins are there to support analytics firmsteaming with small groups of industry experts to tackle nicheoperating problems.Like oil fracking techniques, analytics programmes developedfor one programme can be redirected to horizontally explorepockets of value beyond that of the original project mission.Thus, greater reuse of the analytics asset is achieved.Rather than be restricted by IT’s narrow support windows andauthorised suppliers, the business units and individualdepartments work directly with big data suppliers, often usingthe vendor’s infrastructure or a SaaS cloud.A cheaper and better way to extract valueBig data is all about exploring and exploiting a CSP’s existingtransaction data to discover and extract value. I think we candraw an analogy between telecoms data exploration and oilexploration by energy firms.In the previous IT era – the Before diagram, expensive systemsintegration programmes were the order of the day. The ideawas to transform the CSP enterprise through large-scaleprogrammes architected and managed by the CIO/CTO offices.24Data sources expand beyond the data sets maintained in thedata warehouse. There’s an expanded use of deep packetinspection (DPI), radio access network (RAN) data and manyother third party resources including location data, marketingdata and social media.And just as oil exploration benefited from advancedseismograph technology to locate oil wells underground, thelower cost of analytics programmes allows more testing andVanillaPlus
trials. Thus, low cost data exploration projects are funded andthe ratio of oil strikes to dry wells improves.The high demand for network analyticsThe key dilemma for telecoms today is not service delivery.Skype can do that in voice. Apple does that with iMessage.CSPs and cable operators face a much bigger task: deliver theservice with sufficient quality or download speed to earn apremium. Network analytics and optimisation is thereforeessential to keep telecoms a sustainable and profitablebusiness.Now just because CSPs own terabytes of data doesn’t meanthey know how to manage them well.Guavus found a way to analyse big data before it’s transportedto a warehouse. Now that’s a clever idea. Ontology, HP, andSplunk, meanwhile are using Google-style search engines toskirt around APIs and deliver applications that don’t requireexpensive systems integration – they don’t even requireaccessing a database.Analytics also brings great value to service assurance. Forinstance, comparing the best performing cells against the worstperforming cells is a great way for mobile operators to isolateproblems or sort out which of the 60 devices in the network iscausing an issue. When it comes to optimising IP networks, offline analytics can actually deliver greater value than waiting foralarms to trigger in the NOC. So analytics becomes the newservice assurance paradigm and will steadily replace previousgeneration software, at least in the non-real-time segment.In fact, with the rise of 4G, LTE and smartphones, the telecomsnetwork becomes more and more a data network, and with therate of growth in wireless traffic and devices, engineers need allVanillaPlusthe help they can get from outside suppliers to understandcapacity issues as firms like Subex, SAS and others arestepping up to provide.Certainly one of the biggest stories is the race to analyse datafrom the radio access network, where TEOCO and Amdocsrecently made big investments, and Tektronix Communicationsand InfoVista are also key contenders.Consider this: probably 80% of the QoS issues in the mobilenetwork live in the RAN. This is why RAN data is so crucial. Inthe years ahead, radio spectrum that’s become scarcer andscarcer is likely to become even more precious as regulatorsforce operators to share their spectrum in real-time.Finally, the geo-location capabilities of the RAN are invaluable toreal-time data monetisation, basically sharing intelligence withpartners in retail and other industries, a practice that ChinaMobile is pioneering with the analytics help from AsiaInfoLinkage.Still another area of development is coupling analytics withdeep research. Firms like Ericsson are doing that to understandthe behaviour of advanced telecoms networks, which franklybehave much like complex organic systems like the weather.The benefit? Greater customer satisfaction, faster troubleresolution, and lower call centre costs.Market analyticsOne of the largest segments in telecoms analytics is marketinganalytics, which is the effort to deliver real-time offers or ongoing campaigns to entice subscribers to spend more money,renew their new contracts, upgrade their services, and buycontent and other services on demand.25
Several models are now vying for attention as best approaches.Predictive Segment Analysis basically groups subscribers bydemographics or behaviour characteristics to receive aparticular marketing treatment. This approach has stood thetest of time plus key vendors like SAS and IBM support it.Meanwhile, companies like SAP and Alteryx are improving themodel by automating model creation and democratising theanalysis to business users, not just data scientists.Another approach is Social Network Analysis looks at thenetwork of people a subscriber is connected to via phone anddetermines who the key influencers are in the group and putsspecial emphasis on treating those people as a way of draggingin their loyal followers.Contextual marketing is yet another methodology on the rise.The idea behind contextual marketing is to make offers toindividual users based on the context of that individual’ssituation. Contextual marketing got its start primarily in theprepaid sector of developing markets, where user loyalty isoften very low and the CSP has less intelligence on who theuser is. It has certainly been enabled by big data because itwould be very hard to manage offers to individuals without lowcost processing power.A related hot area is data monetisation – basically deliveringreal-time or brand intelligence to third party businesses, oftenbased on mobile location and actions triggered by somebehaviour or preference of the mobile user.Look at the retail industry and you begin to appreciate howvaluable mobile data is. Retailers can use bar code data to findout what a customer bought in the past, but they lack themeans to send a timely offer when the customer is travelingnear a store, or even when she’s in the store, to direct thecustomer to a particular product or merchandise section. The26personal mobile device is the answer.We figure data monetisation is the next generation equivalent ofYellow Pages for operators. Remember that as recently as1999, the Yellow Pages directory business represented 10% ofrevenue for large U.S. LECs (local exchange carriers) likeBellSouth, and it was a hugely profitable business. Yet datamonetisation may be a far more compelling offer to enterprisesthan Yellow Pages ever was because of the immediacy of theinformaiton. Plus data monetisation is already proving itself inChina where China Mobile monetises relationships with storeowners in malls and websites, offering discount coupons forrestaurant meals via the mobile phone.Greater efficiency and lowercosts in customer careSales and customer care is a hot analytics market and that’s abit of a pleasant surprise.Here, analytics can drive business through real-time analysis ofcustomers. Back end databases, such as those supplied byNeustar, can significantly improve the order uptake ofcustomers and prospects calling into a call centre, visiting awebsite, or using IVR.Gaining a quick diagnosis of problems is another critical areawhere big data is helping. If a CSP has 75 active smartphonesin use in a region, imagine if a customer calls in to complainabout QoS and the care representative has no way of knowingwhether or not there’s a problem with a particular handset.But comScore’s analytics system runs a defective handsetanalysis on a regular basis so when the customer calls in, thecare representative knows immediately whether the handset isthe likely issue. This can save a few minutes in average handlingVanillaPlus
time which translates to hundreds of thousands of dollars indecreased care costs.Meanwhile, the light user of mobile may be highly profitablegiven the low costs he incurs.Business assurance and enterprise searchThe search capabilities of Google, Baidu and Bing are poweringthe modern world, but today that same power comes in ashrink-wrapped version, namely the enterprise search enginesfrom companies like Splunk, HP and Ontology.Business assurance is an analytics sector concerned withlowering costs and plugging revenue leaks due to operationalerrors, fraud and detected ill-conceived partnerships withinterconnect and content partners that are either not profitableor not generating business value.One area with huge potential in business assurance is margin orprofitability assurance. At a macro level, the finance departmentcan measure costs going out and revenues coming in but thereal challenge is getting a view of profitability at a very granularlevel.At face value, a mobile user paying a premium price with lots ofdata usage sounds like a VIP, but on closer examination, youare paying out huge interconnect fees to support that customer.What enterprise search basically allows a CSP to do is skirtaround APIs and other data access issues and perform eitherad hoc analytics queries or build a full-blown application usinglogs and other machine data.ConclusionI’ve spent time scanning a software market that’s exceedinglybroad and deep – but that’s the nature of the big data beast. Itsrise is an extraordinary opportunity for CSPs to bring innovationback into the business through hundreds of low cost, quick ROIprogrammes delivered by solution vendors, large and small.About Technology Research InstituteTechnology Research Institute (TRI) is a boutique market research firm that has been trackingtelecoms BSS/OSS developments since 1994. In 1996, TRI published the first-ever syndicatedresearch reports on fixed and mobile billing systems. In recent years, TRI has focused onbusiness assurance and analytics. In 2013, TRI published a sweeping 40-vendor, 515 pagereport on the market for ‘Telecoms Analytics and Big Data Solutions’. Dan Baker, TRI’sresearch director, is a regular contributor to 7
Company summaryKey differentionSAP is using its big data, cloud and mobile expertise to bothsell to and partner with CSPs. The company offers a fullenterprise analytics suite bolstered by its flagship SAPBusinessObjects and Crystal Reports assets. Recently, SAPhas added advanced drag-and-drop data visualisation with aproduct called SAP Lumira and rapid predictive analyticscapabilities through its KXEN acquisition.Predictive Analytics based on KXEN technology and combinedwith it’s own innovation in trusted data discovery, is key toSAP’s bid to gain a time-to-model and time-to-insightadvantage over competitors such as SAS and IBM SPSS.Big data analytics credentialsSAP’s CSP go-to-market strategy is built on three key pillars:Customer engagement – Here, better and more consistentmarketing and selling across contact channels is achieved withthe hybris solution, which, in concert with SAP’s CRM, billingand analytics suite also enables CSPs to more efficientlymanage campaigns and send more relevant, customer centricoffers. SAP Mobile Commerce delivers mobile wallet andmicro-banking solutions as well, especially for CSPs thatoperate in countries where the handset is the primary internetaccess point.Competitive pressuresSAP enables business analysts and marketing specialists torapidly create and maintain – on their own – hundreds tothousands of predictive analytics models that micro-segmentsubscribers. This task normally requires enormous manual workby statisticians or data scientists which are in short supply.Likewise, SAP quickly figures out which campaign offers areworking and which are not. Combining its speed on thepredictive side with the company’s larger IT infrastructure andcloud strengths and its real-time, big data, and in-memorydatabase capabilities, SAP is in a good position to grow intelecoms analytics.Operational efficiency – Building on its ERP platform, whichhelps over 4,500 CSPs globally to run their businesses better,SAP offers its entire Business Suite in the cloud, providingflexible deployment options for CSPs. In addition, recentacquisitions such as SuccessFactors in human capitalmanagement and Ariba for cloud-based procurement helpCSPs maintain costs and quickly innovate in these areas.New revenue generation – SAP partners with leading globalCSPs to help them generate revenue from new, non-traditional,services. Services such as enterprise-class cloud services,mobile commerce, data monetisation, machine-to-machine andmanaged mobility services. CSPs can monetise its traditionaldata and services with SAP Mobile as managed service asevidenced by existing partnerships with major CSPs worldwide.28VanillaPlus
Big data's fourth V While big complexity is the greatest challenge, big data is certainly about managing huge data volumes too. In many ways, telecoms with their massive networks practically invented big data. And plenty of telco use cases fit the so-called three Vs of big data: large Volume, Velocity (speed of analysis), and Variety (of .
tdwi.org 5 Introduction 1 See the TDWI Best Practices Report Next Generation Data Warehouse Platforms (Q4 2009), available on tdwi.org. Introduction to Big Data Analytics Big data analytics is where advanced analytic techniques operate on big data sets. Hence, big data analytics is really about two things—big data and analytics—plus how the two have teamed up to
big data analytics" To discuss the in-depth analysis of hardware and software platforms for big data analytics The study only focused on the hardware and software platform for big data analytics. The review is centered on the impact of parameters such as scalability, data sizes, resources availability on big data analytics. However, the
India has the second largest unmet demand for AI and Big Data/Analytics, driven primarily by large service providers, GCCs and the start-up ecosystem NCR Others Hyderabad Pune Mumbai Bangalore Chennai Top Skills Talent Big Data/ Analytics 5,800 AI 1,200 Top Skills Talent Big Data/ Analytics 19,100 AI 7.400 Top Skills Talent Big Data/ Analytics .
Q) Define Big Data Analytics. What are the various types of analytics? Big Data Analytics is the process of examining big data to uncover patterns, unearth trends, and find unknown correlations and other useful information to make faster and better decisions. Few Top Analytics tools are: MS Excel, SAS, IBM SPSS Modeler, R analytics,
example, Netflix uses Big Data Analytics to prescribe favourite song/movie based on customer‟s interests, behaviour, day and time analysis. 3. Python For Big Data Analytics 3.1 . Advantages. of . Python for Big Data Analytics Python. is. the most popular language amongst Data Scientists for Data Analytics not only because of its ease in
The Rise of Big Data Options 25 Beyond Hadoop 27 With Choice Come Decisions 28 ftoc 23 October 2012; 12:36:54 v. . Gauging Success 35 Chapter 5 Big Data Sources.37 Hunting for Data 38 Setting the Goal 39 Big Data Sources Growing 40 Diving Deeper into Big Data Sources 42 A Wealth of Public Information 43 Getting Started with Big Data .
Retail. Big data use cases 4-8. Healthcare . Big data use cases 9-12. Oil and gas. Big data use cases 13-15. Telecommunications . Big data use cases 16-18. Financial services. Big data use cases 19-22. 3 Top Big Data Analytics use cases. Manufacturing Manufacturing. The digital revolution has transformed the manufacturing industry. Manufacturers
Ulster Archaeological Society at the Divis and Black Mountain site. 5 Illustration 1: Divis Mountain viewed from the south west 1.2 Aims In order to enhance the archaeological record of this site, the aims of this survey were to produce an accurate plan drawing of the monument and carry out a photographic survey. This information was compiled into a report and submitted to the Environment and .
