User experience monitoring: using synthetics to prioritize end-users
Customer experience still matters in online spaces, user experience monitoring gives teams the edge they need to provide the quality users expect.
Whether you call it end-user experience monitoring or user experience monitoring, getting insight into how your customers struggle with your product is the first step to improving. If you’re not already monitoring user experience, we created this simple guide to help you get into the habit and start identifying Key Performance Indicators (KPIs) for your website or application’s user experience.
So, how do you monitor user experience?
First, you need data on your current user experience. It starts with three important dimensions:
- Availability: Can customers reach the website or app?
- Functionality: Is the website or app running the way it’s supposed to?
- Latency: How fast is the website or app running?
More than likely, the monitoring product you use provides you with data on availability and speed. Whether your product is functioning correctly is a more difficult question to answer, and it’s easy to over-complicate the data you’re receiving.
Using synthetic testing in user experience monitoring
Synthetic testing is an essential part of any monitoring system, but it’s absolutely necessary for testing user experience. You won’t be able to answer the question of whether your systems are functioning the way they should without synthetic testing. Both single request and multi-step synthetic testing create a better view of your product’s functionality than strictly monitoring the health of your devices.
We suggest using multi-step synthetic testing combined with single request tests for user experience monitoring. The single request checks will give you the basic availability data, while multi-step synthetics can provide teams with information about common user interactions such as completing a purchase or logging in.
Tip: Give your multi-step checks very specific tasks to complete to prevent over-complication.
Multi-step synthetic checks allow you to check various user experiences on your website or application, and often, a single test can exercise several parts of your product at once. However, it’s important not to over-complicate your checks so that they’re not returning alerts like, “step 15 out of 30 failed” which are often difficult to debug.
Instead, focus on very specific tasks which will make it easy for you to determine where the issue occurred, and what infrastructure might need maintenance.
For example, a multi-step check which adds an item to cart and completes a purchase on your website.
This type of multi-step check monitors several key parts of the user experience and also website functionality:
- Logging into a customer account
- Navigating the website
- Adding items to cart
- Paying for an item using a credit card
- Login is functioning properly
- Moving between pages on the website
- Items can be added to the cart
- The payment portal is accepting credit card payments
Due to the narrow focus, alerts coming from the check are going to be more actionable, and also easier to correlate. If you know the first step in the check failed, then you know there’s something wrong with the login portal, and you can immediately focus your attention on the related infrastructure.
Reach out to your QA team
Your QA team might have a wealth of pre-built multi-step checks that your team could use as a starting point for your monitoring. This can both save time, and create much stronger monitoring practices.
QA in production, as some call it, is a powerful tool for any IT team. By creating checks which align with the tests the QA team does before release, your team can continue to run those checks in production. QA in production allows your team to both easily diagnose issues, and get an idea of the health of the system overall.
Look for monitoring which provides outcome-based alerting
Outcome-based alerts are ideal in user experience monitoring because they provide information that systems engineers can act upon quickly. An alert like, “error submitting credit card information” provides information to your team about what infrastructure needs attention.
Highly actionable alerts not only reduce Mean Time to Diagnosis (MTTD) and Mean Time to Resolution (MTTR), but also create a wealth of data on the business impact incidents have on your organization. Ultimately, this can make triage and general incident management easier, and help your team manage your infrastructure more efficiently.
Bringing it all together to create better user experience monitoring
Gathering data on availability, functionality, and latency, your team will be able to identify the user experience KPIs for your website or application. Once you’ve established KPIs for your service, you’ll have a better idea of whether it’s available, will provide high-quality service, and if it’s easy to interact with. Synthetic testing will give your team all the building blocks they need to both maintain internal SLAs, and provide your customers with a better product.