Taking Blockchain Beyond Bitcoin and Payments

To the extent that they know anything about blockchain distributed ledgers, people associate it with bitcoin, banking or payment systems in general. However, as I mentioned in an earlier research note, blockchains have a range of potential use cases. Indeed, blockchain distributed ledgers can look like just another technology in search of a mission. However, that’s because there are many ways of putting the technology to practical use that complement and enhance established patterns of doing business. For example, Walmart recently announced it will be using blockchains to establish authentication and traceability in its food supply chain; a French financial services company started a project to facilitate compliance with know-your-customer rules; and there is an anticounterfeiting service that can be used for authenticating diamonds and luxury goods. Technology that conforms to how an organization operates and provides immediate, clear benefits usually is adopted broadly and quickly.

One potential use that intrigues me that hasn’t received much attention is the role that blockchains can have as a machine-to-machine universal data connector for business-to-business (B2B) transactions. This fits nicely in an evolving computing environment that is increasingly shaped by hyperconnected and loosely coupled systems. Let me unpack those ideas one at a time.

“Hyperconnectivity” is a fuzzy term describing a state where something is instantly and easily connected to everything else. The term was coined a few years ago to describe the increasing use of multiple channels (such as email, instant messaging and the telephone) for communication. In a broad sense, hyperconnectivity can be applied to the internet’s potential to offer universal machine-to-machine connectivity. Yet while it’s easy to describe, achieving this aim isn’t simple.

Achieving universal connectivity is an important advance because until relatively recently, it was a major challenge to connect the data contained in one system to another without human intervention. An example is passing customer information contained in a company’s CRM system to its ERP system. In a business-to-business scenario, another case is one company generating a purchase order and having the information read by the seller’s systems without anyone on the seller’s side needing to enter the data. In both cases, automating system-to-system data transfers speeds up processes and substantially reduces the chance of errors occurring if data is re-entered manually.

Especially in B2B settings, the challenge has been system compatibility and security. An early approach to address this issue was proprietary electronic data interchange (EDI) networks, which address a need for secure machine-to-machine connections to exchange documents. For example, in a B2B setting, Company A sends a purchase order through an EDI network, Company B processes the purchase order and sends Company A an invoice. Company A sends an electronic reply to Company B with a functional acknowledgment. However, the disadvantage of EDI systems is that a company can only connect with other companies on that network and the data must be structured to conform to the network’s protocols. Setting up and maintaining connections to an EDI network as well as the charges imposed for communicating with other companies can be costly. While this cost has declined and it’s easier to maintain connections to multiple networks, the expense and hassle still limit the number of networks a company will join – many companies belong to just one. This isn’t hyperconnectivity.

Over time the process of moving data between applications and organizations has become less burdensome but by no means simple. Electric plugs, which connect devices to the electric grid, provide an analogy for how blockchain distributed ledgers can be useful. Electric plugs were devised to provide greater flexibility and ease of use relative to hard wiring a device to a power source. At first, there were dozens of plug configurations in a given geographical market, but this gave way to standardization. Even so, today there are 13 basic alternating current plug designs in use, each a standard within its own region. Recently, though, the universal serial bus (USB) plug has emerged as the worldwide system for providing current to an expanding set of products. Although USB-to-AC adapters are the most common means of powering phones and other devices, USB outlets are beginning to appear in hotel rooms and on airplanes to address the power needs of travelers from all over the world. It offers a universal standard.

Similarly, at first data integration was accomplished by “hard wiring” one computing system to another by laboriously defining the data types, data definitions and interchange process. It was time-consuming, and the setup was brittle: Any change that altered some aspect of the data and data handling in one system could break the integration. Over time, methods for data integration using middleware and messaging became more standardized. Some applications are promoted as having “prebuilt connectors” to the most popular software with which users are likely to want to integrate. More recently, integration platform-as-a-service software has emerged as a means of connecting a range of applications, especially those needing cloud-to-cloud and cloud-to-on-premises connections. As a next evolutionary step, blockchain-enabled distributed ledgers can serve as the universal data connector in a hyperconnected world. And easy connectivity is necessary to support a market that increasingly wants loosely coupled systems.

In the design of computers and systems, “loosely coupled” systems are those where each component has little or no knowledge of the definitions of other separate components but can function without such knowledge. Loosely coupled systems are designed to require less effort to integrate components and less effort to maintain those integrations because changes in one system does not necessarily require any change in the other or in the connection. Such systems give users greater flexibility in selecting the software that they use as well as in deciding what they want to add or change and when. Loose coupling can lower barriers to entry in software markets in some cases because it erodes the value of an integrated suite, but established vendors can also benefit by having greater flexibility in adding to or updating their offerings.

Distributed ledgers can be used to establish loosely coupled hyperconnectivity in B2B commercial communication using a publish-and-subscribe structure. Today, commerce networks derive some of their value by maintaining a data transport service that has well-defined parameters and security. This enables subscribers to that network to reliably connect their computing systems to it in a way that automates the process of sending and receive data to others on the network in a pure machine-to-machine fashion. The network operator supplies the security of the network. Participants on the network may be vetted to the extent that they are visible and accountable to the network operator.

In the future, distributed ledgers may be used to replace the network component of proprietary commerce networks. Companies would have the option to join multiple permissioned distributed ledgers that are free or available at a nominal charge. Free services are likely to be provided by existing EDI vendors that will collect fees for services other than data transport or companies that use it as a loss leader or to serve as another marketing tool to attract business.

In this world companies would use their multiple blockchains to publish requests for proposal (RFPs), tender offers, publish price lists and “listen” for bids, among other actions. Having loosely coupled core systems would enable the creation of commerce networks that facilitate interactions between buyers and sellers by establishing automated, secure and device-independent connections.

Distributed ledgers have the advantage of built-in security. Permissioned ledgers can have an advantage if managed by actors (such as self-regulated commercial bodies or governments) that have the trust of the participants. Permissioned blockchains provide highly verifiable data sets because the consensus process creates a digital signature visible to all parties. The blockchain structure provides a permanent audit trail since no records can be deleted without collusion on a massive scale. In addition, distributed mirrored databases substantially reduce the ability of anyone to tamper with data since each instance would have to be altered in an identical fashion almost simultaneously. Public key encryption can be used to control access to communications depending on circumstances. Companies might make a communication universally consumable or limit access to a specific group or recipient.

A quarter century ago the World Wide Web had seemingly limitless potential to change commercial and personal life. In retrospect it wasn’t a technology in search of a purpose, even if there were many instances of failed applications of the technology. No doubt there will be failed misapplications of blockchain distributed ledgers and even failed commerce networks using the technology. However, I think it’s almost certain that when the technology kinks are ironed out with establishing them, one of the earliest uses of blockchain distributed ledgers will be as a machine-to-machine universal data connector for business-to-business transactions.

Follow me on Twitter @rdkugelVR

and connect with me on LinkedIn.

NICE Robotic Automation Improves Interaction Experience

Robotics is nothing new to some aspects of manufacturing and the IT industry, but it is relatively new in the customer experience (CX) market. The term often conjures up images of little gray machines taking over tasks previously handled by humans – machines making cars, programmed vacuum cleaners and the like. In the CX space, however, we are not talking about machines but about software that can automate routine tasks. For the time being, I don’t believe robots will take over the contact center and replace human agents. Indeed our recent research into next-generation contact centers in the cloud strongly suggests the opposite. It shows that the telephone is still the top channel of communication and that almost two-thirds (62%) of organizations expect call volumes to rise over the next 24 months. Thus agents will continue to handle large volumes of interactions, which may become more complex.

This complexity, plus demanding consumer expectations, requires organizations to handle interactions efficiently and capably or risk losing customers and/or business opportunities. This opens up the opportunity for organizations to take advantage of robotic process automation. NICE, a longtime contact center systems vendor, has offered real-time process automation since 2001, and it recently launched a new product in this market. It now has three products in this space – desktop analytics, desktop automation and its latest, robotic process automation. NICE Desktop Analytics captures information about what agents, or other designated users, do on their desktop, including systems they access, information they look up, data they enter, information they give callers, and systems they update after finishing calls. The analytics enables organizations to track the four basic components of a call – identifying the caller, identifying the caller’s issue, providing a response and completing any required after call work. The analytics component thus can identify best practices for interaction handling and agent performance, and recommend changes to processes or coaching and training.

NICE Real-Time Decisioning helps organizations improve interaction-handling processes. It can, for example, take an account code or calling number and automatically present the customer’s data to the person handling the interaction. This set can include demographic data, financial data, marketing, sales or service data, and an interaction history. It can then take data entered by the person handling the interaction to look up other relevant information; for example, if the caller is asking about a product this information can be automatically popped onto the desktop. The system includes rules and algorithms that suggest what the person should do next to ensure the best outcome of the interaction. The system can also complete some basic after-call work such as updating other systems. This is where robotic automation comes into play.

Either by using analytics or by observation, users can identify processes to be automated. NICE Robotic Automation includes tools for users to map the process, and then develop “robots” – software that includes algorithms and is driven by rules and data – that can automate the process; for example, the software can populate name and address changes across multiple systems, complete claims forms, initiate customer onboarding or send personalized messages. The tools use point-and-click techniques, so robots can be developed by business users with minimal assistance from IT. A robot can be initiated because it detects a trigger (data received) or is kicked off by what NICE calls the  “robot controller,” a person designated to manage the operation of robots It then runs in a virtual environment until the process is complete, when it either picks up another task or is terminated. The system is highly scalable: The number of robots can be scaled up or down to match the number of tasks to be carried out. The product also includes multiple security options that program robots to comply with specified regulations.

At one point in my career I worked for a partner at a management consultancy who was famed for saying, “Software makes bad processes go wrong more quickly.” This is often true, and NICE’s process automation is about achieving the exact opposite – creating smart processes that run more efficiently and delivery better outcomes for customers, agents and businesses. In relation to the four components of call handling mentioned above, it can immediate identify the customer, capture the issue, guide the agent how best to resolve the issue and then reduce or eliminate after-call work. In doing this it can also reduce data entry errors, make agents’ jobs easier, improve the customer experience, help ensure that more interactions are completed successfully, and achieve something all contact center managers I know have at the top of their to-do lists – reduce average call-handling times.

vr_ngce_research_01_impetus_for_improving_engagement_updatedOur research into next-generation customer engagement shows that these capabilities align with the top objectives organizations are focused on as they try to improve customer engagement: improve the customer experience (76%), customer service (70%) and business processes (54%), become more competitive (46%) and reduce operating costs (43%).  NICE’s process automation products have the potential to impact all of these, so I recommend that companies assess how it can help in their efforts to become more efficient and effective. Does it mean robots will take over the contact center? I think not, but it can make processes run faster and smoother and free up employees to focus more on the customer.


Richard Snow

VP & Research Director Customer Engagement

Follow Me on Twitter  and Connect with me on LinkedIn