The Convergence of Transactions and Analytics

Write this down: Converged transactional and analytical data platforms stand for digital business.

None of us has been alive long enough to remember a time without the printing press, but such a time did once exist. The printing press revolutionized the world of books, journalism, scientific research, and countless other areas by automating a process that was once tedious and time consuming.

Interestingly, data management solutions are experiencing a similar transformation. An exciting new architectural advancement—converged transactional and analytical data platforms—merges online transactional processing (OLTP) and online analytical processing (OLAP) onto one platform. For the first time, enterprise IT professionals can conveniently perform advanced analytics on live transactional data to generate real-time insights and recommendations on a single data management platform. And if you can do all of this on a single copy of data without data duplication, this will help you achieve real-time insights while reducing costs and overhead associated with copying data.

In a typical enterprise, database architects manage the transactional, operational, and analytical workloads across separate databases, data warehouses, and data lakes. This setup creates data silos that make it difficult to support real-time insights and analytics without moving data from the transactional systems to operational systems to the analytical systems.

Porting this data across multiple systems is slow and cumbersome, and it prevents enterprises from being competitive in the fast-moving digital world. Businesses can’t provide real-time price quotes to the sales team, monitor customers’ on-site assets, or recommend a product based on past shopping preferences. Likewise, partners in a global supply chain might not have the latest available inventory or shipment arrival dates.

New Database for New Digital Era

For companies committed to being digital and competing in the digital era, converged transactional and analytical data platforms move them forward on their journey.

This emerging database technology is so new that industry experts are still defining the term. At Forrester, analysts call this a translytical data platform and describe it as having the ability to “deliver faster access to business data to support various workloads and use cases.” They released the first Forrester Wave report on this space and solution providers this past November (“The Forrester Wave: Translytical Data Platforms, Q4 2017” by Noel Yuhanna and Mike Gualtieri, Forrester, November 2017).

Gartner refers to certain database architectures as Hybrid Transactional/Analytical Processing (HTAP) and has been using the term for a couple of years. Gartner states that “HTAP architectures remove the latency between when a piece of data is generated and when it is ready for analytics. HTAP becomes a key enabler of digital business and its demand for real-time data. The implications can be challenging to understand” (Gartner, “Delivering Digital Business Value Using Practical Hybrid Transactional/Analytical Processing,” Adam Ronthal and Roxane Edjlali, April 27, 2017).

For enterprise architecture pros, this is an exciting next step for data management advancement, and they are leaning on these new platforms to help define their enterprise strategies. Here are the digital business projects that these databases can move forward:

  • Real-time apps, such as stock trading, fraud detection, counterterrorism, patient health monitoring, and earthquake monitoring
  • Internet-of-Things analytics and operational data that improve risk management, such as asset monitoring on windmill farms, oil and gas fields, air compressors, and manufacturing plants
  • Connected data apps, such as a trusted view of critical business data or real-time analytics for public safety and city planning
  • Machine learning, such as the continuous learning about customers, business processes, and operations

A 60% Improvement in Operational Efficiencies

Looking over the past few decades, database providers have pushed the boundaries of faster processing and better performance with every release, and businesses have kept pace with the improvements, always ready for more. Current business needs include machine learning, IoT, and predictive analytics so they can revamp their operations and keep up with customer expectations.

While these technologies may seem future-oriented, they are delivering very real, practical results for early adopters. A reseller’s sales team, for example, could be competing head-to-head with rivals to close a deal before the quarter’s end. Machine learning can determine the deal’s probability to close, leading the team to adjust the negotiation points before making a final offer that will up their odds of winning the sale. In another example, the CFO can apply predictive analytics to pending orders to forecast more accurate profit and loss expectations.

At Vectus Industries, a manufacturer and supplier of safe storage water tanks, pipes, and fitting systems, analyzing operational data on a single, in-memory translytical platform is giving them critical business insight. The company leaders are using that insight to make data-driven decisions to expand into national and international markets.

Vectus employs 1,300 people in 13 plants and 30 locations across India, and it is digitizing its business as quickly as possible. Recently, the company added IoT sensors to customer water tanks and is monitoring them from a mobile app. The field personnel rely on this data to help service equipment and meet customer needs.

Once Vectus moved its data to a single translytical in-memory platform, visibility increased 50% across the entire company, allowing managers to know what is happening regarding sales and inventory in real time. “Our operational efficiency increased by 60%, and we also achieved 50% faster access and improved visibility into our real-time data,” said Manish Sinha, head of IT at Vectus.

The invoicing process also went from frustrating to painless. Previously, Vectus would create multiple invoices for a single dispatch. Now, it provides customers with a single, easy-to-read invoice that includes all entries from the dispatch. In the future, artificial intelligence (AI) could automate matching invoices to orders without manual intervention.   

Convergent IS, based in Calgary, Alberta, has also moved to a translytical database platform. The consulting firm made the leap because it wanted accessible, actionable data. “The idea that you can get yesterday’s data faster didn’t excite me that much, but if you can get tomorrow’s forecast faster, if you can get access to more in-depth questions more flexibly, then those are the sorts of things that get more interesting,” said Shaun Syvertsen, managing partner at Convergent IS.

Convergent IS has created a single source of truth that is accessible to all departments. Business leaders can access and query the data to help them make business decisions with better business outcomes. “It’s truly on-demand, real-time processing and reporting with new metrics not previously available at the point of action,” Syvertsen said.

The Digital Cliff Shouldn’t Be Scary

Converged transactional and analytical platforms fit the label of “less is more.” From one platform and single copy of data, enterprise architects can analyze operational data, apply AI to customer data, or use machine learning on IoT data from sensors to predict and respond to the health of remote assets. These platforms deliver the real-time capabilities that are the hallmark of digital business.

A key question for enterprise architect professionals is, “When is real time necessary?” Not every project demands real time, but the list is growing. Reports and business insights based on days-old, stale data, or even relatively fresh, yesterday’s data, are completely appropriate for some situations. Looking forward, though, large numbers of digital projects will demand real time.

Converged transactional and analytical data platforms bring other business benefits, specifically more power for reduced costs. Combining analysis and operational data on a single system reduces the number of database platforms needed, simplifying the data model and streamlining the landscape, and potentially decreasing overall RDBMS costs by 40% to 45%. On the other side of the coin, when translytical database platforms are combined with an in-memory technology, performance spikes significantly. The unique combination of performance, real-time, and dual-mode capabilities is entering the market at exactly the right moment.

The Steady Rise of Digital Investments

Similar to the first time that pen and ink were united, translytical database platforms will trigger a series of database technology improvements, and they will spin up opportunities that deliver business value. With the entry of translytical database platforms into the market, enterprise architect professionals can move forward on the army of digital projects proposed by executive management and line-of-business leaders.

The desire to predict customer purchases, improve the accuracy of financial forecasts, respond more quickly to—or even prevent—machinery failures at remote sites, and track and improve supply chain deliveries is not going away.

In fact, the demands are only getting started. In the 2017 cognitive technologies survey from Deloitte, the analysts lumped machine learning, AI, robotic process automation, and others under the term “cognitive technologies.” A huge majority, 87%, of the executives surveyed, expect that these technologies will play a significant role in improving their products and services, and more than one-third of the organizations have invested $5 million or more in AI and cognitive technologies. They are investing primarily in product development and customer service.

These early adopters and other digital businesses dismiss outlooks of “business as usual.” Change is brewing, and you can ink in converged transactional and analytical data platforms as the catalyst for responding to those changes and staying ahead of them in the digital era.


Subscribe to Big Data Quarterly E-Edition