Production monitoring

Why uptime monitoring can miss broken customer journeys

The monitoring dashboard is green. Availability over the past month is almost 100%. The server responds, the page loads and everything appears to be fine.

Yet no order has been completed since Friday’s release. The payment gateway is returning an error. Or a template update removed the button from the basket, leaving customers with nowhere to go. Without checking the whole journey, you may learn about the problem only from a complaint or a fall in orders.

In the meantime, you have been paying for advertising that sends people to a website where they cannot buy anything.

A common problem: the server responds, but the customer cannot complete their goal

Conventional monitoring checks whether the server responds. It calls a URL, receives a response and records that everything is fine. This is a legitimate check—but it answers only whether the machine is running, not whether the business process works.

The server can respond without difficulty while a purchase remains impossible to complete:

What these cases have in common is that an availability check alone may not detect the functional failures. The difference between “the server responds” and “the customer completes the goal” has direct business significance.

A green result from regression tests before release is not enough on its own. The tests may have run in an environment with different configuration, a payment gateway in test mode and controlled data. A problem may appear in production only after release or following a change made by a partner.

How to monitor the outcome for the customer

Instead of asking the server whether it responds, we follow the entire journey as a customer would.

At regular intervals, a browser follows an agreed journey directly in production—for example, signing in, finding a product, adding it to the basket or completing a safely prepared payment step. If the scenario stops, the timing of the alert depends on the interval and confirmation runs.

We prepare scenarios to reduce the risk of affecting production: they use separate accounts, the payment gateway’s test mode where available and controlled data clean-up. Before starting, the team must verify what the particular production integration permits safely.

When the scenario fails, the alert arrives through a channel the team actually monitors—by email or through Slack or Teams. We configure it to avoid creating noise that the team begins to ignore after a week.

What monitoring a key journey gives you

You can respond sooner. An alert within the agreed interval reduces reliance on a customer complaint or a delayed look at revenue.

You receive more useful evidence. Depending on the tool and configuration, a failure may include a screenshot, video or network trace. This evidence can shorten the search for the cause, although it will not always replace reproducing the issue.

Third-party systems are monitored too. The payment gateway, delivery services and third-party sign-in. You feel a partner’s outage just as you would your own, so you want to know about it just as quickly even when you are not responsible for it.

You can see how response times change over time. Gradual slowdowns after deployments can be investigated before they become an outage.

Next step

Tell us which journeys must not stop working—typically purchasing, sign-in and payment. Contact us and we will agree what to monitor and how often.

Related topics

You might also be interested in

Detect an outage before a customer reports it

We write scripts that validate purchase, login and payment in production and plug them into your pipeline or scheduler - you run them.