Google Analytics Custom Alerts You Must Always Use

It is extremely difficult to keep an eye on significant variations in your website traffic or any of your marketing campaign/traffic source manually and that too 24/7. Here is where Google Analytics custom alerts keep handy.

A lot has been written about GA customer alerts but unfortunately there is still no widespread adoption. Either marketers are not aware of what they are or they are just plain lazy when it comes to implementing them.

A custom alert from Google can help you in minimizing the impact of a bot attack.

A custom alert from Google can help you in quickly fixing a broken shopping cart.

A custom alert from Google can help you in quickly finding a search engine penalty.

A custom alert from Google can help you in quickly fixing completely broken website tracking.

A custom alert from Google can also help you in keeping your client or saving your job.

A long time ago, I was one of the many marketers who didn’t believe in using custom alerts though I was very well aware of its importance even then. Only excuse I can give you now is that I was just plain lazy.

So one day, I got a wake up call in the form of an email from one of my clients. He was unhappy/frustrated to learn that his shopping cart had been broken for the last 3 days and I didn’t notify him (also subtly telling me “that he now doubts whether I am even working on their account”).

Now the only problem was, I didn’t know myself that the cart was broken. But I was expected to know, as I am their analytics expert. I was expected to keep my fingers on the pulse but I couldn’t.

The bottomline is that nobody can keep track of each and every anomaly in data 24/7. You need a system through which you can automate monitoring of data anomalies. Fortunately Google Analytics provide such automation in the form of custom alerts. You just need to find some time and the will to set them up. It is not very hard either. 

Through Google Analytics custom alerts you can monitor significant variation in your website traffic and marketing campaigns. Whenever such variation occurs you get an email alert or text message from Google asking you to take immediate actions.

The custom alerts you create must be descriptive, so that just by reading them you know exactly what you need to do.

I have categorized custom alerts into two categories:

#1 Site wide custom alerts – alerts which are applied to all website traffic.

#2 View specific custom alerts – alerts which are applied to a particular segment of website traffic like: direct traffic, organic traffic, referral traffic etc.

Following are the custom alerts which you must always set up in your GA account:

Site wide Custom Alerts

#1 Average time on site decreases by more than 50% compared to previous day

This alert usually means either your website page load time has increased in the last 24 hours (because of some technical issues) or you got large volume of irrelevant traffic on your website.

There are two cases in which you usually get sudden large volume of irrelevant traffic on your website:

#1 You launched a new campaign which is sending traffic which has nothing to do with the product/service/content you promote.

#2 Your website has been attacked by a spam bot. Spam bots can send large volume of fake traffic on your website. Since fake traffic is not going to engage with your website, consequently the average time spent on the website may go down.

So if your new campaign is not working well for you, you need to optimize it.

If a spam bot is skewing your ‘average time spent’ metric then you need to block it. I have explained in great detail about how spam bot skew your website traffic and how you can block them in the article: Geek guide to removing referrer spam in Google Analytics.

Note: Make sure that you select the checkbox: “Send me an email when this alert triggers”. Otherwise setting up such custom alerts is pretty much useless.

#2 You are no longer tracking [Enter your Goal conversion]

So your custom alert could be something like: “You are no longer tracking Email Signups”. 

This alert is triggered when your specified goal value dropped by more than 90% compared to previous day. Your goal conversions can be signup, file download etc.

So for example, if you get 30 signups every day on an average and suddenly one day you got zero signups then there is a high probability that your Goal conversion tracking has broken and you need to investigate.

You will also get the conversion tracking broken alert through Google Analytics tracking notifications but you will see these notifications only when you log into your GA account and by that time it could be too late. Many marketers/web master don’t use GA every single day. So the best way to get such alert is through email and you get Google Custom Alerts through email.

#3 Your shopping cart has broken 

This alert is triggered when the website total sales dropped by more than 90% compared to previous day. If your website gets little to no sales on weekends or gets orders once in a while then this alert won’t work for you. However this alert works wonder for those ecommerce websites which get lot of orders even on weekends. 

#4 Your website bounce rate increased by more than 70% compared to previous day

This alert also means either your website page load time has increased in the last 24 hours or you got a large volume of irrelevant traffic on your website. This alert is a better indicator of a bot attack on your website as spam bots affect the bounce rate much more than ‘average time’ metric.

Note: You may need to adjust the 70% bounce rate percentage according to your website average bounce rate. 

#5 Your Website tracking has completely broken

This alert is triggered when the website total sessions dropped by more than 90% compared to previous day.

#6 Your website traffic has increased by more than 50% / 100% compared to previous day

When you get such an alert you need to investigate the cause of traffic spike immediately. Sometimes a traffic spike could also be a result of bot attack i.e. a spam bot sent large volume of fake traffic to your website.

Again you need to adjust the percentage (50%, 100%) according to your website average traffic fluctuation. If you run a low traffic website, your site traffic can easily go above 100% any day because of some paid or free promotion. On the other hand if you run a high traffic website, even 50% jump in traffic will be quite significant. 

The thing to remember here is to select a percentage which is way above (ideally double) your daily average traffic fluctuation. 

View Specific Custom Alert

There are 10 google analytics view/profile which every marketer should create and use. They are:

1. Direct View – this view includes only direct traffic from your website.
2. Main view – this view is the unfiltered view of your website.
3. Mobile View – this view includes only mobile traffic.
4. Organic View – this view includes organic search traffic.
5. PPC view – includes only Paid search traffic.
6. Referral View – includes only referral traffic.
7. Search View – includes only search engine traffic (both organic and paid)
8. Social View – includes only social media traffic.
9. Testing view – this view is just another copy of the main view and is used for only testing GA configurations.
10. Target Market View – this view includes only the traffic from your target market location.

I have explained creating each of these views in great detail in the article: 10 Google Analytics Views that you must always use

Once you have set up these views then you should create following customer alerts for each view:

#1 [View Name] Traffic dropped by more than 90% compared to previous day#2 [View Name] Traffic dropped by more than 50% compared to previous week#3 [View Name] Traffic dropped by more than 50% compared to previous month

#4 [View Name] Sales dropped by more than 90% compared to previous day#5 [View Name] Sales dropped by more than 50% compared to previous week#6 [View Name] Sales dropped by more than 50% compared to previous month

 #7 [View Name] Traffic increased by more than 100% compared to previous day#8 [View Name] Traffic increased by more than 50% compared to previous week#9 [View Name] Traffic increased by more than 50% compared to previous month 

#10 [View Name] Sales increased by more than 50% compared to previous day#11 [View Name] Bounce rate increased by more than 70% compared to previous day

So for example, in case of organic view, your custom alerts name would be:

• Organic Traffic dropped by more than 90% compared to previous day
• Organic Traffic dropped by more than 50% compared to previous week
• Organic Traffic dropped by more than 50% compared to previous month and so on.

In case of direct view, your custom alerts name would be:

• Direct Traffic dropped by more than 90% compared to previous day
• Direct Traffic dropped by more than 50% compared to previous week
• Direct Traffic dropped by more than 50% compared to previous month and so on.

Similarly, in case of referral view, your custom alerts name would be:

• Referral Traffic dropped by more than 90% compared to previous day
• Referral Traffic dropped by more than 50% compared to previous week
• Referral Traffic dropped by more than 50% compared to previous month and so on.

Granted they are so many custom alerts but then that is the only way to continuously monitor any anomaly in data regardless of the traffic source.

So for example, if your organic traffic has declined by more than 50% in the last week, then there is a possibility that your website has been hit by a search engine penalty and you need to investigate.

If your direct traffic or referral traffic has increased by more than 100% in the last one day then either your content has gone viral or you have been attacked by a spam bot. 
Feel free to adjust the alerts frequency and percentages to meet your individual needs.
If you liked this article you may also like: 
4 Steps To Get More Out of Twitter Analytics

8 Must Follow Google Analytics Tips To Understand The Data Better

Using Analytical Analysis To Help Improve Conversions

To view the original article Click Here