The Ultimate Guide to Diagnosing Traffic Drops in Google Analytics
You arrive at the office one Monday morning, expecting it to be like any other day. You grab your coffee, chat with your coworkers, and finally make it to your desk to do your daily morning activities.
That’s when you notice… Your traffic dropped off. *screams in the distance* A LOT of your traffic dropped off.
How could this happen to you? You’ve heard about it happening to other people but you never thought it would happen to you.
First off, don’t panic. This fictional scenario (yes, it was fictional) is actually more common than I made it seem. Websites are fickle mistresses and they can break for a million different reasons. If you do happen to have an inexplicable drop in traffic, follow these steps to discover the source and you’ll have your site traffic back up in no time.
Go Straight to the Source
The best place to start is at the source… literally. The source of the traffic will tell you exactly where to begin while digging into analytics. Each source has vastly different characteristics and these characteristics are the starting point. You’re going to need to go through some very different paths if your paid traffic went down compared to your organic traffic.
Start by making sure that your segment is set to “All Sessions” as this will make sure that all of the traffic your site received is accounted for. Then click on Acquisition on the left side and Overview. Once you’re there, scroll down and you’ll see a breakdown of all of the channels that brought traffic to your site.
Now we need something to compare against.
Of course you don’t want to wait a long time after a traffic-catastrophe but the longer it’s been, the more data you will have to work with and the easier it will be to discover what’s wrong. Set the date to the beginning of the traffic drop all the way through the last full day. Check off “compare to” and choose “previous period” from the drop down.
At this point you should have a graph that shows the performance of each source compared to a time period from before your fateful day.
You should also be able to tell which ones fell and which ones didn’t. You can’t proceed without knowing which channels were affected.
Is there just one channel that was significantly impacted? You can breathe, this is actually a good sign. When one channel is significantly impacted, it is a lot easier to find out why.
If all of your channels dropped, we are basically back at go and we didn’t collect $200. We do know that it’s not a specific channel so it doesn’t mean we haven’t gone anywhere yet.
Once you’ve discovered which channel is affected, read the relevant section below. Each section will tell you what you should be looking for and guide you through the symptoms to the ultimate problem that caused a drop in traffic.
All Mediums –
If all of your mediums dropped it’s likely a technical issue. Each channel has such varied reasons that they would drop but if it’s all of them it’s most likely a technical issue that is preventing everyone, no matter which channel they come from, from entering the site.
Once you know that all of the mediums dropped, there are a few things you can check that might lead you to the answer.
Google Trends: Sometimes, it really is this simple. There are fewer people interested in your product now than there were before. This generally shows through a gradual decline but seasonality or an event in the industry could cause this.
Affected Pages: See which pages were affected. If all the pages were affected you likely have a technical issue preventing people from accessing your site. If only a section was affected, there might be a widget or code tweak that caused the page to break.
Annotations: Check your annotations. You’re taking annotations right? (I’ll harp on this frequently). If you’re taking careful annotations for changes on the site, you should have some inkling as to what may have caused this issue.
Google Webmaster Tools: Google webmaster tools has a lag in its data of about 2-3 days. If you haven’t figured it out by then, go check out webmaster tools (once there’s data, of course) and see if there are any errors or indexing activity.
Organic traffic dropping is the toughest to diagnose. You have no control over the source of traffic and sometimes the symptoms for the drop in traffic are hidden deep in the analytics data. Don’t worry though, it’s there! You’re just going to have to work for it.
Here are a few things you can look at for a drop in organic traffic but if you need a deeper dive, read this article for some additional steps.
Landing Pages: Which pages were affected. Be sure to check landing pages because it’s the loss of people entering the site that we are worried about. Is there a pattern? Was one section of the site affected while the rest remained unaffected?
Secondary Dimensions: Secondary Dimensions are where the trends tend to show up for drops in organic traffic. Unfortunately, picking the right one can be a guessing game sometimes. I prefer source and device category, but others may have revealing data as well.
Direct Traffic –
Direct Traffic would be a very odd one to see. Particularly because direct traffic rarely consists of only direct traffic. Most of the time, if you see a drop in only direct traffic, you’re probably actually dealing with an attribution error (or an attribution solution if a problem was fixed).
Some things that you could look into that would explain a drop in direct traffic are:
Tagging Issues: Some links or sources of traffic don’t automatically pass a referrer. This is the most common in email marketing but can happen elsewhere. It’s possible your direct traffic dropped because someone decided to start tagging the right way.
Offline Marketing: If your company markets offline and integrates digital into their offline (these companies are extremely rare) then you may experience a jump in direct traffic. People would be far more likely to manually type in the URL.
Ghost Traffic: With the proliferation of spam traffic going around, it’s possible that you may have had a lot of false traffic. Once the appropriate analytics filters are in place, it may appear that there was a significant drop in direct traffic when really, you’re simply filtering out visits that weren’t people.
Paid Traffic –
Paid Traffic will be fairly easy to identify since you’re only getting the traffic that you are paying for but there are still places you can look to see what may have caused such significant changes. In many cases, a mistaken setting in your adwords account will be preventing you from succeeding such as negative keywords blocking targeting or changes in bids.
Here are a few places you can look. If you don’t find what you’re looking for here, try this more in-depth article that will help you get to the bottom of your problem quicker.
Negative Keywords Blocking Targeting: Sometimes, when you’re too aggressive with phrase match keywords, you can experience some overlap with your negative keywords. Run the keyword diagnosis tool to see if this is your problem.
Drastic Increase or Decrease in Bids:
Check your notes to see if you made any changes (you’re taking notes, right?). If you didn’t take notes you can check your change history. If you upped your bid on a high volume keyword with a low budget or any similar issue, you may be using up your budget before you ever get any good traffic.
Email Traffic –
Email traffic is pretty easy to diagnose as well. With email, you are only receiving traffic from the emails that you sent out. If you are using annotations (remember how I said I’d be harping on these?) you can likely pinpoint which email blast didn’t do what it was supposed to. Still, there are only a few things that it could be with Email.
Tagging Issue: If you haven’t realized at this point, tagging is pivotally important to tracking your email. Make sure you tagged EVERY link in the email, otherwise your traffic could drop quickly.
Change in Strategy: Make sure that there isn’t a change in the email template (that could have led to tracking issues). Also, a change in strategy could lead to lower engagement such as open rates, click through rates, and subsequently traffic.
Referral traffic is fairly easy to diagnose as well, and much less likely to drop suddenly. This traffic is simply people who came to your site from other sites and is fairly easily tracked. If you go into your referrals in analytics, it shows you all of the domains that have sent you traffic recently. This is likely the best place to start.
You lost an important link: If you used to have a link on a high trafficked site, and then they removed, updated, or repurposed the page you’re going to lose a significant amount of traffic. Look in your referral traffic under acquisition to see where the lost link might be
Spam Referral Traffic: Similar to the Ghost traffic from direct, there can be a lot of spam referral traffic. This may have recently stopped hitting your site (not likely) or you recently implemented a filter to block it (hopefully this is the case).
There you have it…
If you haven’t found it by now, it’s likely I can’t help you. But if you’re on your own now, I will leave you with some advice on how to find what might help you discover the issue.
Look for changes – If you saw a huge drop after a period of average traffic, it’s likely that something changed. Finding what changed will likely lead you to why your traffic dropped.
Find a trend – Traffic doesn’t just drop for no reason, there is something behind it every time and that something is often given away by a trend no matter how hard it hides. Was it everyone on phones? Was it everyone with a specific screensize? Did you lose all your traffic from the UK? Finding these trends will lead you directly to what happened to your site. (PS every medium being affected counts as a trend)
ANNOTATE EVERYTHING – Okay, this one may not be as helpful after the fact but I hope you’ve learned your lesson. If you annotate all changes, the next time this happens you’ll most likely have a clear change that will give you an idea of what trends to look for and it will save you all this trouble and heartache.
Good luck and happy digging in analytics! If you have any suggestions on other things to look for please comment!