Picture this nightmarish scenario:
A dev on your team just pushed a modification to your site… and broke your e-commerce integration.
People can’t buy.
You leave for lunch, maybe even work a few hours after that before you realize what happened.
Meanwhile, BIG money is being lost.
Panic ensues among the developer team: “OMG, how could that happen to us? We do unit tests, heck, we even have some integration tests— they all passed!”
You have a monitoring tool telling you that your servers are up. PagerDuty didn’t call you: the site isn’t down, so your typical monitoring systems weren’t triggered.
You fix everything quickly, but the damage’s already done. Disastrous business day.
After gobbling down a few chill pills, it’s time for a post-mortem.
“How can we avoid that in the future? How can we know when there is an issue with our online money-making machine?”
That’s where Google Analytics alerts come to the rescue.
How? By monitoring the symptoms of such breakdowns: drop in conversions and sales, for example.
I’ve seen real-life horror stories just like the one above. People losing thousands of dollars before even noticing there was something wrong! Today I want to make sure it doesn’t happen to you. I'm going to do so by listing essential Google analytics alerts and showing you how to set them up.
I won’t spend too much time introducing the feature as the name itself is pretty self-explanatory. In a nutshell, GA custom alerts are notifications sent to you, by email or SMS, when selected metric’s thresholds are triggered. I’ll explain how to set them up in GA further down.
As in my prior example, I know that some devs might argue that they already do unit and integration tests, that it should be enough to know everything’s working fine. Sadly, this isn’t true. You should absolutely do them, as much as possible. No question. But they won’t assure you nothing will break, ever.
And the day something does break, analytics alerts might be the last barrier before disaster. At least, you’ll know way quicker something isn't right.
The best way to know there’s an issue with a significant revenue source is to monitor as closely as possible to the money.
→ Read the full post here