Web Reliability

50. Continuous QA and Maintenance

Mitchell Kimbrough
Written November 30, 2021 by
Mitchell Kimbrough
Founder & CEO

Reliable execution helps ensure reliable and continuous customer flow. Continuous flow and continuous monitoring of that flow signals excellence at the monitoring node of execution. The customer is our principal concern. So monitoring their interaction, feedback, and flow through the web system is essential. Continual improvement is only possible where the observation and conversation are continuous.


You’ll be happy to know that there are a variety of tools constructed to help monitor customer interaction with your web property. The market for these tools continues to grow as their value becomes more widely known. Hotjar and Mouseflow are just two of the many tools that allow you to review the customers' journey through your online sales funnel. These tools record your customer's mouse activity and in addition to tracking engagement, they create an opportunity for you to see areas of friction and see where depletions of purpose occur, through the use of session reviews as well as heat maps.

Of course, a classic version of this form of interaction monitoring has been happening for decades in the digital space. UX designers routinely conduct user testing, putting versions of new user interface systems in front of sample customers, and then observing their interactions carefully. Revisions are made based on successes, and new versions tested. This method of iterative design and continuous improvement results in a final version that may be released with confidence. As with many other topics we’ve covered, there’s a deep well of knowledge that already exists in the marketplace. And if you’re interested in pursuing this type of pre-release monitoring, you'll find an ample supply of specialists and experts to help.


That old tried-and-true method of eliciting and tabulating customer feedback is now on the web, where it is more efficient and effective. Customers can be invited to easily and quickly fill out a poll or survey to capture their input regarding your site. There is a broad range of systems and services available to support this way of obtaining feedback. Capturing, monitoring, and effectively utilizing this data to continually improve your site may not be as exciting or urgent as the other tasks that need your attention every day, but its importance should never be underestimated.


Machine-level metrics are available in any commercial web system and can show you where exactly in the web app your customers drop out. In addition to this sort of passive mapping, these systems can also be utilized proactively. An example of this is a feature our team built into an e-commerce site that sent follow-up reminder emails to potential customers who had abandoned their shopping carts. The system watched for latency in the customer's flow through the buying process, and then when the customer interaction slowed or stopped, it responded with encouragement to help usher them through the rest of the process. This type of marketing activity that tracks cart abandonment has of course now become its own niche industry.

Analytic tools and machine learning systems abound which support the effort to comprehend customer behavior as they flow through a system. The realm of big data on the web is one that by its nature has come to rely on this concept of monitoring customer flow, and vast amounts of information are being collected on some sites.

But you don’t need to have a huge site for this concept to provide value for you. Even for a small site with a limited budget, monitoring a customer's progress by reviewing database statistics, log files and the like can bear very useful insights in and of themselves.

Monitoring at the execution layer is not meant to merely be passive, but interactive. It continuously generates the material which informs and directs the team in its strategic iteration and improvement efforts. At any scale, this concept can serve to continually enhance customer flow over time.