Decoding the Economics of Telecom Big Data and Analytics

The digital economy is increasingly becoming a major contributor of gross domestic product (GDP) growth on a global basis. This digital economy is underpinned by the fourth industrial revolution, a collective force that fuses physical, digital, and biological worlds.  According to Huawei, industry digitalization (e.g., manufacturing, public service, transportation, and consultation services) stands to create US$23 trillion by 2025. Total production in the digital economy will be a function of conventional factors of production (land, labor, and capital), but also technology and data. The latter stands to be a key foundation behind many products and services that are fast becoming indispensable to modern life. At present, less than 2% of data that gets generated has been stored. Within that, less than 10% has been analyzed and applied. In other words, 98% of the means of production when it comes to data is not being used. That is significant in terms of productivity and innovation gains.

In telecoms, there is agreement that with 5G, the direction of travel is towards huge amounts of data and insights. With data becoming a crucial production element, smart storage and processing of data are key strands that provide both challenges and opportunities. For example, according to Huawei, it is expected that there will be an over tenfold increase in Call Data Records (CDRs) from 4G to 5G. According to Ericsson, an exponential increase in data and network Key Performance Indicators (KPIs) increases network operations cost by 100-130%. But big data also creates opportunities and it is key to capitalizing on the digital economy. It is the fuel that will be fed to commercial decision making and new value creation. A first step for Communication Service Providers (CSPs) to capture some of that value is to build a data infrastructure, buttressed by a data governance function. This infrastructure should include data collection, storage, processing, and an understanding of what are the specifics of data economics.

In general, everything in the realm of big data and analytics is headed to vast quantities. Companies like Amazon, Apple, and Google leverage vast data repositories to move forward and reach decisions rapidly for their global operations. Similarly, CSPs—and the industry at large—realize by now that large quantities of data can transform the nature of how they conduct business. Data after all, is now a kind of capital. It is on par with financial and human capital in creating new digital products and services. Economists at UC Berkeley have tallied up the total global production information and calculate that new data is growing at 66% per year. This demonstrates that data (and metadata) is the new wealth; its value increases when a specific data sets are linked to other data sets. The least productive part for a data point is to remain naked and alone. A data point uncopied, unshared, or unlinked with other data points does not yield economic gains. It is therefore important to understand the economic features of data. They are as follows:

• Non-rivalrous ‘goods’: Tangible products are rival in nature. Rival means a zero-sum game where one product prevails. There are marginal costs associated with rival goods. On the other hand, big data falls into what is considered as intangible goods or services. They are non-rival. For example, more than one department within a CSP can use a single dataset simultaneously. With non-rival ‘goods’, there are no marginal costs to deploying an extra copy of it. Also, total cost or investment in in a data set drops precipitously as it is shared by many users instead of one.
• Experience good: An experience good is one for which the buyer does not know its value before the purchase decision. In telecoms, much like other industries, the value of data is apparent after use, and not before. This is a significant feature that both vendors and CSPs need to consider as they seek new growth. Growingly, CSPs’ business is the delivery of information and lifelike interactive experiences. CSPs need to look at their business as more than simply the building and selling of consumer telephony services. Much like an experience good, with data and analytics, the industry needs to make the right decisions today, to obtain value tomorrow.
• Fungibility: An aspect worth highlighting is that data, unlike other commodities, is not fungible. In an economic sense, the notion of being non-fungible means that one data set is not substitutable for another. In other words, one unit of data is fully distinguishable from another. This is a worthy feature to keep in mind for the industry, particularly when considering that the type and scale of data needed to orchestrate new value creation in reaches new heights of complexity. Those that will win in the digital economy are CSPs with the expertise that molds that complex data into new forms, novel products, and innovative services.

Increasingly, CSPs’ consumer base expects personalizes experiences. So collecting quality data is more important than ever. With data becoming the currency of digital lives, and the fuel that runs the digital economy, CSPs must ensure the privacy and security of consumer and customer information. This is particularly important in a 5G ecosystem, where increased digitization creates more interfaces and processes and, by extension, complexity. Furthermore, to increase data productivity, CSPs will need to interact with suppliers (e.g., Accedian, Cloudera, Enea Openwave, Guavus) and trade entities to put a cost on data sets so that estimates can be taken on required data quantity, quality, and security from point of origin to point of use.

In a digital economy, data trading stands to create global opportunities. The objective for CSPs should be to seize the opportunities that improved worldwide data trading and sharing confers, particularly in process digitalization and decision making. But protection of all stakeholders involved—data, privacy, and security—remains a key aspect. Therefore, increasing data security, increasing revenue output per data input (data productivity), and increasing data liquidity enables stakeholders to securely monetize data. Data liquidity refers to CSPs’ ability to get data from the source (e.g., apps, sensors) to multiple places where it might be used as efficiently as possible from time, cost and effort perspectives. Oracle, for example, offers an infrastructure for user-centric, application-centric, and data-centric functions that address data security, productivity, and liquidity. Oracle guides CSPs to institute a data-centric model that achieves a global data view, so an inter-domain association spanning both structured and unstructured data. Unstructured data plays a key role when considering that 80% of the world’s data is unstructured and 50% resides outside of the data center.

A data-centric model is at an embryonic stage in telecoms. But when adopted widely, it addresses challenges associated with manual data recording, which can typically take 25% of quality engineer time. CSPs should aim to understand current digital and big data capabilities and identify key leverage points, places in their operations where a small change in terms of data and connected business insights can drive a large shift in behavior. Next, CSPs should leverage data platforms and analytics solutions that are available today by pursuing early pilot projects with Minimum Viable Products (MVPs). Commercial benefits accrue with later iterations once product maturity is in place and suppliers fully operationalize MVPs. Telefónica, Rakuten, and Singtel, for example, already leverage data to become a data-driven company. The aim is to establish new processes apt for the digital economy, new ways of working, and cross-functional correlation between user experience, service quality, infrastructure resource quality, and network performance.

Lastly, change management is critical and constitutes the bulk of CSPs’ effort in their journey to capitalize on the digital economy.  CSPs should seek to bolster data literacy competencies with AI experts, big data capabilities and service experts.  CSPs and the industry at large must invest in human capital that embraces big data and AI to monitor performance, align decisions, and act with confidence. Orange, for example, claims that by the end of its Engage2025 strategic plan, most of its employees will have undergone a data and AI awareness training. Furthermore, to effectively use data and analytics is not only a design challenge, but also a governance challenge. CSPs, with guidance from solution suppliers, should implement data governance techniques. A data governance function, long promoted by ZTE, promotes universally applicable techniques and methodologies. This is important given that new data formats from disparate sources, new analytics functions from a fragmented supplier base, and new AI algorithms will emerge in the ecosystem.