Using AppDynamics Business iQ to Uncover Airline Competitive Seat-Blocking

At AppDynamics, we frequently talk about how insights derived from application performance metrics can help achieve favorable business outcomes. It’s easy to imagine how this would be the case with a slow e-commerce application or full-blown outage, but there are many other use cases that don’t readily come to mind.

Here’s a hypothetical use case that illustrates the diversity of insights that AppDynamics Business iQ can provide. Imagine an airline has built a Business iQ dashboard showing the health of its reservation system. The normal look-to-book ratio—a travel industry metric showing the percentage of people who visit a travel site versus those who actually make a purchase—is about 15% in our hypothetical example.


A typical look-to-buy ratio.

The conversion funnel is part of several AppDynamics dashboards displayed in the airline’s global customer support center. In our use case, the “Get Availability” requests suddenly spike 50% to 125,093, while the conversion ratio drops by almost 400%—from 15% to 4%.


Flight risk: 4% is an abnormally low look-to-buy rate.

This low conversion ratio creates a business problem for the airline, which must hold back seats that other customers might want to purchase. At the same time, the high number of reserved seats triggers higher prices on the remaining unreserved seats—and in all likelihood causes customers to look elsewhere for cheaper tickets from other airlines.

By clicking “Get Availability” on the conversion funnel screen (above), we see a lot of mobile requests are coming from IP addresses that belong to public cloud data centers. These are illegitimate mobile users, as legitimate users would come from an IP address owned by mobile operators, not from public clouds. (To find out who owns a particular IP address, you can use a domain lookup website like ICANN’s WHOIS). Requests coming from IP addresses of public cloud data centers should be blocked from these APIs.

In our example, the airline then exports its table of mobile users to its IT security team, which implements proper filtering to block the rogue accounts (see “Export table as CSV” below). As a result, the mobile user requests quickly return to normal volume.

This near real-time business performance measurement greatly benefits the airline’s bottom line, enabling it to preserve millions of dollars in annual revenue. In addition, the sudden drop in the conversion ratio prompts the airline’s global IT operations center to probe into the company’s business transactions, a process vastly simplified—without any log-digging—by AppDynamics’ user-friendly UI. Business iQ enables the airline to solve the mystery of why some flights, which normally would be packed with passengers, are nearly empty.

The Benefits of Business iQ

This hypothetical case study shows how useful information is tucked away in enterprise systems, waiting to be discovered by an APM solution like AppDynamics. Of course, the problem of bogus tickets and low look-to-buy conversion rates existed long before before the deployment of APM. But in this case, the airline was unaware of the resulting revenue loss until it surfaced in the Business iQ dashboard, spurring its staff into action.

In a world of complex distributed systems and ruthless competitors, companies owe it to themselves to deploy a solution like AppDynamics Business iQ, which can identify potential problems like a sharp drop in conversion rates, as well as highlight the effect on the business and quickly diagnose root cause. The health of your business depends on it!

Good to Great: The eCommerce Metrics that Matter in 2016 [Webinar]

 

Question: when was the last time you stopped by the store to plan out your shopping? Now, when was the last time you opened the store’s website or mobile app to plan your shopping?

It’s more than likely your consumers are reliant on your software applications, rather than your storefront, to complete their buyer journey. This means Retail IT and e-commerce teams are continually faced with the task of identifying and applying the most useful metrics –  that align with business goals, focus on performance, and ultimately deliver an ideal experience to their customers.

According to research we conducted, over half of 1,000 smartphone users surveyed would be more inclined to physically visit a store if they provided an app experience with localized, personal offers when they were nearby. The study also found that a third of customers would avoid shopping with a retailer if they had a negative experience with its app or mobile app. Over 68% of consumers state that the performance of a mobile app impacts their perception of a retailer. Mobile apps play an increasingly important role in consumers’ lives with 70% of consumers surveyed making more than a quarter of purchases online, according to the study. In fact, 20% of those surveyed admitted to making purchases on their phone at work in front of their computer.

Today, retailers need more than just extensive information on their consumers and market. They need to be able to connect that data into insights in order to drive the seamless, omnichannel experience that consumers now consider the new standard of shopping. Today’s web and mobile applications offer a wealth of performance and business data but it could all be useless unless specific, actionable decisions can be made from it. It’s critical to have the most useful — rather than simply the most accessible — information available to make accurate and fast decisions.

Join AppDynamics and Forrester Research to learn about the opportunities available to retailers today when it comes to better aligning your enterprise goals to your customers’ needs. Special guest Sucharita Mulpuru from Forrester Inc. explains which retail metrics make the difference between being average and being exceptional, and how you can use them to enhance customer experience (and your bottom line). Register today for the live webinar, or view on your own time with the on-demand link. Looking forward to seeing you there!

How IT can redeem itself from critical outages

In my previous post, I discussed how AppDynamics Application Analytics can rapidly troubleshoot gradually degrading apps. In this blog post, I’ll discuss how IT can redeem themselves from unfortunate outages leveraging ITOA or application analytics.

With so many factors impacting application performance today, it is not surprising that applications frequently are either slow or simply stall entirely. Sometimes it’s the server, sometimes it’s the database, and some days it’s the mobile network carrier to blame. With the complexity of today’s environments, business transactions (such as logging in, adding to a cart, or checking out) have many opportunities to fail. These transactions often fail long before they reach the backend operational databases where all the “committed” transactions are stored. So, as far as the traditional business intelligence (BI) solutions — that provide insights only into committed transaction data — are concerned, these in-flight, yet failed transactions never existed.  

Another trend that makes the equation all the more complex is IT is delivering more and more modular applications at faster and faster rates making the entire application ecosystem very dynamic and highly error-prone. Applications are bound to be slow or stall; It’s how IT redeems itself from such critical outages with their business stakeholders that matter most.

AppDynamics Application Analytics provides real-time visibility to deliver insights into both aggregated or rolled-up metrics, as well as details into individual customer interactions. The breadth of visibility covers the frontend to backend technology stacks — from clients (mobile or browser apps) to the Java or .NET or other language apps, to databases, to servers and infrastructure components. The depth covers code-level visibility into live in-flight transaction data captured as the application processes it, and is not restricted only to those transactions committed to the operational database. 

Let’s say that John is trying to purchase a shirt on his mobile device while commuting back from work. His checkout transaction times out as his mobile device frequently switches cellular towers in the fast moving train. John finally gives up after repeated attempts. Now, this transaction failed on the network carrier and never hit the application inside the data center servicing such requests. So, the application infrastructure has no knowledge of John’s attempt at purchasing the shirt. In another case, Susan was trying to buy a purse from her browser but her order got stuck in a queue due to errors in the messaging queue. Susan’s transaction also never hit the operational database.

In such situations, IT can leverage AppDynamics Application Analytics to not only identify these individuals who had poor experiences, but also to capture what they were trying to accomplish (purchase items, in this case).  

Now rather than treating these as lost revenues and giving up on them, IT can redeem itself by extracting key information into a spreadsheet with a click of a button. This information includes account information of the purchaser (name, email, etc.) as well the exact items being purchased. The marketing team can then use this valuable information and run some win back campaigns that offer 10% discounts to win the heart of the customer. The more sophisticated retailers might go a step further by pre-creating a shopping cart that has all the items pre-loaded, with a discount applied, and ready for one-click checkout from the email itself. Now that’s what I’d call going above and beyond!

IT, typically looked at as a cost prevention insurance method, can actually help drive revenue given the right tools!

Want to check it out for yourself? Download a FREE trial today!

How Omni-Channel is Your E-Commerce Application Monitoring?

Screen Shot 2014-10-30 at 4.36.31 PM
Like it or not, an ever-expanding holiday season is here, and the words moratorium and code freeze are being heard around the halls of many retailers who are gearing up for their Super Bowl of shopping. Just this time last year I was responsible for middleware performance of a leading online and mobile payment provider, and I remember the excitement, and stress, of heading into this time of year. With mobile (mCommerce) and ecommerce shopping trends skyrocketing, expectations and demand for the business and technology we worked so hard to build were extremely high.

This year alone eMarketer predicts Business to Consumer (B2C) e-commerce sales will increase over 20%, to upwards of $1.5 Trillion Dollars, and by 2017, that number jump to $2.3 Trillion. Mobile trends are predicted to show an equally impressive growth; mCommerce in the US alone is estimated to grow $52 to $86.8 Billion from 2014 to 2016 (eMarketer). Moreover, with the growth comes the challenge, as IT organizations are pushed to understand, manage, and support heterogeneous customer access points and technology stacks in parallel.

Screen Shot 2014-10-30 at 4.36.41 PM
With the competitive nature of e-commerce, managing and optimizing end-user experience has never been more important. Studies co-authored by Bing and Google – http://radar.oreilly.com/2009/06/bing-and-google-agree-slow-pag.html– prove a direct correlation between page speed and revenue, determining that even a 500 ms delay can have a significant impact over most any Key Performance Indicator (KPI).

Screen Shot 2014-10-30 at 4.36.50 PM

AppDynamics is deeply involved in helping online, and offline, retailers guarantee service delivery during these critical times. While mobile is certainly the wave of the future, improving end-user experience does not start or stop, at the browser and mobile device. The brick and mortar point of sale (POS) remains an essential element to many retailers omni-channel strategy.

Screen Shot 2014-10-30 at 4.37.03 PM
Whether it is a custom solution or built on top of a 3rd party platform like Retalix, AppDynamics provides complete monitoring visibility, fault isolation, and diagnostics end-to-end across the entire POS system. Below is an example of a Retalix environment where AppDynamics is providing real-time, end-to-end application performance monitoring coverage, including: in-store registers, self-checkout, systems, price scanners, backend services, integrations with external systems, and backend databases.

Screen Shot 2014-10-30 at 4.37.12 PM
Using AppDynamics, we’ve enabled IT to monitor, map, and score 100% of the Retalix Point of Sale (POS) transactions in real-time as well, as well as getting out of the box visibility into the exact business functions being executed. If and when there is a problem, the operations team is equipped to take immediate action by:

  • Receiving a notification identifying what performance degradation is occurring, and which stores and business services are impacted; e.g., a response-time problem at critical severity affects all mid-west stores.
  • Identifying the business function impact: understanding the problem affects the item scan feature in the self-checkout aisles
  • Isolating precisely what components are degraded, by visualizing the end-to-end application topology; knowing the failure lies within the store or regional servers, a network link, shared backend services or databases.
  • Quickly diagnosing the issue down to a source line of code or SQL query.
  • Screen Shot 2014-10-30 at 4.37.22 PM
    Screen Shot 2014-10-30 at 4.37.31 PM

    Consumer trends are driving the business rapidly to expand commerce over multiple channels, and as highlighted above our goal is always to enable IT to follow suit with an equally comprehensive monitoring plan. Failure to do so, despite best intentions, leads to what we call the Frank-Monitor problem; a series of loosely or de-coupled point solutions that fail to integrate into an effective end-to-end capability.

    So when you are considering how best to define, initiate, or develop, your monitoring strategy, make sure to ask yourself just how omni are you, actually?

    Because despite the best efforts to stitch together disparate solutions into a cohesive understanding of system performance, it almost always ends up a science project!

    Want to give AppDynamics a try for your e-commerce platform? Download a FREE trial today!

    4 Insider Tips to Help E-Commerce IT Departments Get Ready For The Holiday Season

    Parag Sanghavi helped contribute to this post

    When consumers are thinking about which gifts to buy for their friends and family, retail businesses are crossing their fingers hoping their e-commerce stores will survive the holiday rush. As per Internet Retailer, e-commerce holiday sales are set for double-digit growth this year. Internet Retailer also projects there will be 70% growth in 2014 from mobile commerce revenue that is projected to grow to $34 billion. Branding Brand reported that the tipping point came in August 2014 when 51% of traffic to leading retailers came from smartphones and tablets. It is likely that Black Friday 2014 is going to see the majority of online traffic from mobile devices, for the first time in history. Thus, performance of your mobile applications is very important to your business this holiday season.

    Mobify provided insights for Monetate’s report published in Internet Retailer mobile 500 guide stated that conversion rate from the mobile channel is very low. As per Kissmetric’s survey, 73% mobile users reported that they encountered a website that was too slow to load. While 38% reported that they encountered a website that wasn’t available. In spite of good traffic, is your mobile channel driving customers away due to bad customer experience?

    To avoid runtime performance issues, clearly retailers want to know how to best prepare their m-commerce and e-commerce stores for this scaling traffic. In fact, as part of your preparations, most of you must be busy with load testing and thrashing e-commerce applications to 4 to 5 times the anticipated load for the upcoming year.

    Despite load testing and planning ahead of time, unforeseen and unplanned application performance issues will always surface on the day of peak traffic to your e-commerce site. This will significantly impact your customer experience, and more importantly, your online revenue. Thus, your holiday preparations are not complete without devising a well-thought strategy for managing these real-time performance issues on the day of peak load.

    Based on helping many retailers for their holiday preparations last few years, we would like to provide following insider tips. We believe that one more important thing you can still do for getting ready for this holiday season is gifting your Ops team with the AppDynamics Application Intelligence Platform.

    Tip 1: Gain end-to-end visibility into every transaction across mobile, web, and e-commerce platform layers:

    A common problem for many e-commerce sites is performance problems caused by third party services and applications such as payment gateways. Performance problems with these services can be just as harmful as problems with your storefront application because your customers will still blame you for any problem. Thus, it is utmost important for you to gain visibility into the end-to-end flow of every transaction – all the way from mobile applications, web end to back end transactions.

    AppDynamics will give you visibility into every component of your production environment including visibility into third-party applications, web services, and insight into your end-user performance for your browser applications.

    Web_End_User_A-960x0

    Screen_Shot_2014-07-17_at_3.42.47_PM-960x0 (1)

    AppDynamics will provide code-level visibility for every transaction running in any of the following e-commerce platforms:

  • Hybris
  • Oracle ATG Web Commerce
  • IBM Web Commerce
  • Magento
  • Any custom e-commerce platform built on J2EE platform
  • Any custom e-commerce platform built on .Net platform

  • Overstock.com, one of AppDynamics’ customers (ranks 31st in Internet Retailer’s Top 500 guide), has several million customers. “We do over a billion dollars in revenue a year,” said Neil Hartner, Senior Architect at Overstock.com. “Our customers expect fast response times, and we work hard to ensure that we’re the fastest retail site on the Internet.” Hartner worked with the development staff to determine the architecture and improve the performance of the website, ensuring it was as fast as possible. He said, “I myself am a developer by background, and I find that the AppDynamics solution has everything I look for. I can not only look at bigger picture of deployment but I can also look at call stacks to see exactly what code is being executed, I can see SQL calls so I can see queries that are being run and how long they’re taking to execute, and I can see web services that are being called.”

    Tip 2: Track health of Mobile API gateways closely

    Typically, user-experienced mobile performance problems are just the tip of the iceberg caused by issues in the backend application. Last year we found that many retailers encountered issues with their mobile applications as they didn’t monitor health of mobile API gateway very closely. Mobile API gateways are one of the applications thrashed very heavily during holiday peak. You need tools to track health of mobile API gateways in real time.

    AppDynamics platform provides the best in class support for most commonly used mobile API gateways including Mule, WSO2, and Tibco. Using AppDynamics platform you can monitor real time performance of your mobile API gateways.

    Tip 3: Fix production problems before they affect end-users by deploying self-healing tools

    The financial stakes are high on Cyber Monday in the US and during the entire holiday shopping season. The estimated brand damage could average $3.4 million for one hour of downtime as per report published by EMC based on survey conducted by Poneman Institute in 2013. You need monitoring tools that not only alerts as soon as it spots that a transaction is deviating from its normal path, but also self-heals in an event of an error. This will provide you ample time to proactively debug an issue even before it starts impacting your user experience for your customers and hurting your bottom line.

    AppDynamics provides capabilities for baselining and setting health rules so the system can identify unusual behavior before problems become problems. You can identify slow backend transactions by looking at transactions snapshots showing user requests. Similarly, you can identify if end-user response time is beyond normal range by looking at browser snapshots or identify issues associated with higher network response time by looking at network request snapshots. Dynamic baselines learn your environment behavior and appropriate response times, ensuring your Ops team isn’t pummeled with alert-storms or false-positives.

    AppDynamics platform also provides functionality like automatic JVM restart for self-healing. This provides you sufficient time for diagnosing problems without impacting end user performance.

    Screen Shot 2014-10-21 at 3.28.17 PM

    Pizza Pizza is the largest pizza chain in Ontario, Canada, with more than 600 restaurants and over 3,000 employees. Amar Narain, Director of IT at Pizza Pizza, was responsible for making sure the applications were always performing well, especially during peak hours. He said:

    “With the AppDynamics platform we can catch problems before they cause an outage. And that means we’re saving money, and keeping our customers happy. That’s what matters.”

    Tip 4: Track Omni-channel revenue impact in real time

    When the e-commerce website is your storefront, every business transaction impacts your bottom line. For running targeted marketing campaigns, you need tools for tracking revenue from various channels in real time. Slowed transactions or errors impact your business and lead to lost revenue. Therefore, you need tools to help you connect dots and help quantify the impact of performance issue to your revenue.

    AppDynamics lets you configure business metrics for capturing data from instrumented applications and return values to report on the performance of the business.

    Screen Shot 2014-10-21 at 3.28.27 PM

    “I’d recommend the AppDynamics platform to anyone running a public website where you need to know what’s happening,” continued Hartner from Overstock.com. “You want to know in real time how things are performing—and when there are problems, you need to quickly figure out where the problems are.”

    Powered with AppDynamics Application Intelligence Platform, your Ops team will be ready for any application emergency on the day of peak load. This, in turn, will ensure that your customers will get the best online shopping experience during the holiday season.

    Don’t wait to be reactive to performance issues this holiday season, start a FREE trial of AppDynamics!

    We sincerely hope you have a successful holiday season this year!