3 Common Problems That Might Be Ruining Your Analytics Data

3 Common Problems That Might Be Ruining Your Analytics Data

By Melanie Datz –

How reliable is your Analytics data?

This is a question we ask frequently, as we’ve seen many problems in Analytics accounts which lead to reporting nightmares. From duplicate transactions to misattributed paid traffic to double-counted traffic, once there’s a problem with your data, it stays there.

Problem #1 – Duplicate Transactions

Perhaps the worst data problem is duplicated transactions. The longer it goes on, the worse it becomes. Not only do you have to go to your boss and say, “You know that really great year we were having? Yeah, we’re not,” but until you discover when the duplication started, all of your revenue and transaction numbers are suspect.

The good news is that checking for duplicated transactions is easy. You can download a Custom Report from the Analytics Solutions Gallery. The one we use filters transaction IDs by the number of times they are used. Anything more than one is a red flag. Frequently duplicate transactions are caused by users returning to an order confirmation to check their order – sometimes obsessively, particularly if people are returning to track shipments before a holiday.

Duplicate Transactions

Click to enlarge

If you find duplicated transactions, get your web team to investigate and fix. We’ve had clients fix duplicate transaction issues by preventing customers from returning to the page that triggered the transaction, but server-side logic to prevent the Ecommerce tag from being delivered more than once to any URL is also an option.

Once the fix is implemented, make a note in Analytics that the fix was applied. It’s also a good idea to try and pinpoint when the duplicate transactions began. Moving forward, it may be simpler to asterisk the previous year’s transaction and revenue data than to try and determine a more accurate number, especially for large retailers with thousands of organic transactions in a week. But if you do want to try and approximate a true number, you can download an Excel report of transaction IDs by date and remove the duplicate transaction IDs. This isn’t a foolproof method, but it will get you closer to an actual number.

Problem #2 – Misattributed Paid Traffic

Something else we see frequently are large amounts of paid traffic (with transactions and revenue) misattributed to organic traffic. This can happen when there are keyword tagging errors, when parameters are incorrectly identified in Analytics, or when Analytics and AdWords are incorrectly linked. Once it happened because Analytics and AdWords weren’t linked at all, since ads were run through a third party platform.

A good time to check for misattributed paid traffic is when there’s a large and surprising spike in traffic, conversions and revenue.  Look for landing page URLs with obvious paid parameters or tags like “camptype” or “GooglePLA”. One or two landing pages getting a couple of sessions may not be a problem, but something like this definitely is:

Misattributed Paid Traffic in Organic

Click to enlarge

Once you discover misattributed paid traffic, be sure to look back and see if this is a one-time error or if your organic data is loaded with these landing pages.  If you can see frequent periods of paid misattribution, it’s a good idea to check year-over-year landing pages every reporting period.

Thankfully, paid misattribution is easier to correct for than duplicated transactions, as long as you’re sure you’ve found all the paid tags.

Correcting for Paid Misattribution

Problem #3 – Duplicated Traffic

A third, and thankfully less common, problem, is duplicated traffic involving two separate Analytics tags. We saw this with a client running a separate mobile site with its own Analytics tag. When the main site’s Analytics tag was added to mobile pages, mobile traffic and revenue was reported twice, as numbers from both Analytics tags were rolled into one report.

Protect Your Analytics Data

In all of these cases, the effect of the data problems was to make reporting accurately impossible. Was revenue up or down? Hard to say because of the duplicated transactions. Were traffic drops year over year a result of a site problem, or due to large amounts of paid misattribution? In the case of duplicated transactions and the double counting of mobile traffic, it took a full year from the implemented fix to know for sure what traffic and conversions really were.

Protect your Analytics data. Proactively check (or be sure your SEO agency is checking) for duplicated transactions and misattributions, and that data from multiple Analytics tags isn’t rolled together incorrectly.

Leave a Reply

Your email address will not be published. Required fields are marked *