When most people think of operational analytics, they think of monitoring of internal processes and systems to ensure smooth performance and to identify potential issues before they become problems. What often isn’t considered is how big data and operational analytics can enhance customer relationships and improve knowledge of customers’ propensity to purchase, turning what internal operations that often are considered cost centers into profit centers.
Data Informed spoke with Charles Rich, Vice President of Product Management at Nastel Technologies, about how IT Operational Analytics can deliver actionable business insights, speed problem resolution, and enhance the customer experience.
Data Informed: Businesses often apply IT operational analytics (ITOA) looking inward, to ensure optimized IT operations. How can this same approach and information be applied outward, to improve the customer experience?
Charles Rich: Customers are non-employees on the other side of the firewall that utilize your applications to transact business with your company. They might purchase a product, reserve a seat on an airplane, sell a stock, or execute other transactions specific to your business.
They have no visibility into what happens after they hit the Enter key. But they will quickly judge your company’s responsiveness and the quality of their relationship with you based on how fast their screen updates after that Enter key is pressed. This is why it is so important to monitor, measure, and analyze the end-user experience. Your company’s reputation and revenues may depend on it.
Using ITOA, we also can reduce event volume by analyzing KPIs [key performance indicators] or metrics about events rather than just simplistic thresholds, which is time intensive and not very effective. Fewer events allows better prioritization.
With this approach to ITOA, businesses can ensure their customers have positive experiences and when there is a problem, rapidly resolve it. This approach improves the customer experience.
What other kinds of business insights can ITOA provide?
Rich: Businesses also can measure business milestones accomplished, such as purchases or claims processed, and correlate that with the end-user experience as part of a bigger picture business-performance view.
Today’s big data analytics can provide valuable input into customer behavior and buying patterns. The business can learn a lot by analyzing user interactions with their application using ITOA. Changes in the visualization layer of applications can have a drastic impact – positive or negative – on sales. Marketers have known for a very long time that slight modifications to subject lines, landing pages, and forms can deliver substantial improvements. The same holds true for the UIs [user interfaces] of business applications.
Capturing how users interact with applications is essential for determining what a good user experience is. This means capturing errors – where the user hesitates as they are unsure about what to do, how fast the applications perform, and whether user actions lead to orders. However, if you aren’t acquiring user-experience data, you can’t do the analysis. Marketers do multi-variate testing to determine empirically what works better. It can be a brute-force approach, but it is one that delivers results. Using the performance of big data systems, we can perform live tests with business applications, stream user interaction data in for analysis in order to decide what works best for optimal user experience, and then make changes in near real time.
How can user transaction data be used as a barometer for IT performance?
Rich: One way to look at transactions is that they are the “actions” or business outcomes that the business wants to occur. These can be purchases, sales, approvals, reports, acknowledgements, credits, debits, and many other actions. The “trans” part signifies that they happen between or across two or more parties, often the customer and the business running the application. IT implements these actions, technically as transactions within applications.
Using this definition, transaction tracking could be considered the most important thing that ITOA analyzes. Most importantly, we want to capture the “flow” of transactions. While, technically, a transaction could be considered a logical unit of work, an “atomic” event, such as the operations and messages between a “begin” and a “commit,” from a business perspective a transaction such as capturing an order has many steps within it. This includes the messages and exchanges between components as well as individual messages, invocations, requests, and calls.
Tracking these components and their relationships across infrastructure is essential in determining where fault is and where latency can be reduced. This can be a challenging effort in complex topologies or in situations where the transaction crosses firewalls and enterprises, sometimes as long-running or asynchronous. But it’s well worth the effort.
Tracking transactions can be extremely helpful in situations where there aren’t obvious errors or alarms. This often happens in scenarios where we are required to monitor compliance or in situations where we are trying to find lost orders. Tracking transactions provides a great barometer or metric as well as a diagnostic tool for determining the health of the relationship between the business and the customer in at least as far as IT can measure it. And when that relationship is not right, transaction tracking – especially when correlated with response time, error rates, and end-user behavior – provides sufficient detail and specificity to make rapid problem resolution possible.
How does the real-time/streaming nature of ITOA data make it more valuable for understanding the customer experience?
Rich: Data-in-motion or real-time streaming data has a unique aspect to it: time. There may be insights hidden in the data that have a brief shelf life. That means the analysis has to happen fast enough to still be relevant. This analysis can consist of measurements of elapsed time for a web transaction, waiting for a response from the JVM [Java Virtual Machine], DB response time, other unexpected results from queries, or even customer behaviors that offer an opportunity for a better relationship.
In an eCommerce scenario, if the business can react to this data in near real time, action can be taken to avoid losing a customer. In brick-and-mortar environments, there often is a tremendous amount of user interaction data, such as product preferences and geolocation, that are captured by POS [Point of Sale] systems. In many cases, this data is moved in batch to the headquarters of an enterprise to mine for insights into customer behavior that would later be applied to advertising and promotional activity. It is quite advantageous to analyze this data in real time while the customer is still at the POS and engaged with your business. The real-time analytics capabilities of ITOA can make this possible.
This approach to ITOA data appears to transform IT from a cost center into a department that can contribute to a company’s profitability. How much potential do you see for IT to be a cost saver and revenue generator in the big data era?
Rich: If the modern definition of IT is used, which from a DevOps perspective includes both development and operations, then are we looking at the department that has the most contact with customers. Customers over the Web interact with the applications that IT is monitoring and analyzing. Therefore, using ITOA, IT can be the group with the most influence on customers and, thus, on the business. From this perspective, IT can have a significant impact on revenue and could even be considered as a profit center.
One approach to making this happen in an enterprise is to link together business milestones and transaction tracking. Business milestones are aggregates of discovered transactions that, taken together, define a specific business objective. It isn’t always sufficient that the transactions are completed on time. They may need to be done in a specific order to be considered compliant. These are metrics that the business measures. We could consider tracking business milestones as a form of business process monitoring. An example of a milestone could something in financial services called “trade execution” or “trade enrichment.” Linking these together provides instant awareness of whether a business goal (milestone) has been accomplished as well as which transactions are used to accomplish this goal. This gives IT the “knobs and dials” to do something to increase the achievement of business goals. It also builds a bridge between the business and technology in order to accomplish their joint goals together. This approach is something that has great potential to elevate IT to an essential part of revenue generation and customer retention.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise.