Top Six Attributes of an Ideal End User Monitoring Solution
What’s an IT organization to do to prevent launch day mayhem?
Oct. 5, 2011 09:45 AM
As someone whose job it is to look after user and customer experience management products, I have an affinity for websites that are available, performing and provide a great customer experience. So imagine my disappointment when I heard the website of a well-known retailer crashed during a major product launch. According to a company spokesperson, the resulting "mayhem" was unprecedented.
When any organization puts in the tremendous amount of work required to successfully launch a new, in-demand product, launch day is when all that hard work finally comes to fruition. If that launch is via your website, the day may be a big party for everyone but the IT team. That's because user experience is paramount to the success of a major launch, but, for most IT organizations, user experience is a blind spot - a metric not illuminated by traditional monitoring tools. If customers or users have a poor experience, there is a strong possibility you've lost them for good - and that's a metric the business can track... in cold hard cash.
What's an IT organization to do to prevent launch day mayhem? The short answer is to monitor user experience on your website in real-time. By keeping tabs on how users are experiencing service delivery and identifying any delays or processing errors they are encountering, you can quickly spot problems and deploy the right resources to resolve those problems before they impact the business. If you're on a search for the ideal solution, here are things you want to look for:
1. Manage problems and incidents with real-time and historical data
The ideal solution provides an at-a-glance view of website health and calls out any performance problems that may be impacting user or customer experience. This perspective empowers you to see performance problems and service level agreement violations. It also may show you which data centers are associated with the performance problem, and isolates the geography where users are experiencing poor performance. Poor performing pages, images and transactions associated with the problem should be highlighted, and a tier overview can help you quickly identify which teams should be involved in troubleshooting the problem.
2. Definitively isolate performance issues to the client, network or IT infrastructure
While identifying a problem is the first step, it's what you do next that really matters. The ideal solution should give you the information you need to determine if the root cause of a problem is due to an external or internal source, and whether that performance issue is causing you to violate an SLA. You should be able to see how many real users are impacted with now and prior, response time, and what threshold has been violated.
3. Respond to problems before end users call the help desk or abandon your website
The solution should provide alerts when monitored times exceed established limits. Examples include end-to-end transaction times measured at the user's browser, and processing times measured at the web server. Such alerts serve as an early warning system that enables a proactive approach to performance issues. This approach enhances SLA and OLA management by basing the agreements on true end-user experience indicators as opposed to synthetic metrics or incident resolution times.
4. See what real users are doing while on your website and show you what they saw
With the ability to replay an individual session, including the screen content, you see exactly what the customer did at each step of every transaction, viewing the same screens the customer viewed. Playback facilitates problem verification and root cause analysis. It also eliminates the need to manually query users about problems - a time-consuming and error-prone task.
5. Measure business impact of performance issues in dollars and cents
The business measures success in dollars and cents and, therefore, can directly correlate poor customer experience to a decline in revenue. The ideal solution should provide dashboard capabilities that allow you to keep current on exactly how your online business is performing - in real time, anytime. This level of visibility is critical to ensuring your website performance is not hindering conversion rates and revenues. To quantify business impact, you should be able to watch traffic volume from various locations, order revenue, and abandonment rates..
6. See problems by geography, browser type and other common user attributes
Visibility into detailed user attributes permits highly targeted correlation analysis, as well as consolidated performance views across multiple dimensions. It also permits highly granular drill-downs to facilitate root cause analysis. Consequently, an effective monitoring solution should capture robust attribute information, including:
- Site visitor attributes such as user ID, geographical area, ISP, connection speed, and browser type
- Transaction attributes such as the transaction type (i.e., checkout), errors, completed steps, and step conversion rates
- Hit attributes such as date and time of day
- HTTP response codes such as 4xx response codes indicating client-side errors, and 5xx response codes indicating server-side errors
- Segmented response times such as client time (in the browser), network time (between the browser and web server), and processing time (time spent in the web server, app, database, and any other involved systems before the first part of the response is received)
While the big launch for this particular retailer has turned out to be a website blunder, history doesn't have to repeat itself. If your organization relies on its website to sell goods and services, to provide customer service, or to make a splash in the market, you need visibility into customer experience. User and customer experience is a metric that matters to the business, and, therefore, is a metric IT must monitor and manage... on a major launch day, and every other day of the week.