Dangers of Shortened URLs for Analytics

Twitter Facebook Linkedin Email
Shortened URL services like bit.ly are incredibly convenient, right? They allow you to take an unwieldy, lengthy (and let’s face it – ugly) URL and transform it into a short, manageable URL with just the click of a button. You can even view quick stats to learn how many clicks your shortened URL received.
Sweet, right? Not really.
What many marketers and social media experts don’t realize is that major problems can occur from mindlessly shortening URLs to share on social media. You’re impacting the central data for your website in Google Analytics, which is much more important (and telling) than any one-off analytics data you’ll receive from that URL shortener.

dangers of url shorteners

 

In fact, relying on URL shorteners to aid the social management process can ultimately hurt social media in an organization and downplay its value by eliminating important social traffic data in Google Analytics. If your boss or client sees skewed results – fewer sessions and conversions from social media than should be attributed — it may be decided that social media is something he or she no longer wishes to continue.

Problems with Shortened URLs

So, how exactly does this happen? Let’s take a look at the results from an infographic that was promoted primarily via social media with a shortened URL. The stats are segmented by Source/Medium for a specific landing page.

google analytics landing page by source

As you can see, Google Analytics shows the primary source as Direct, with only three clicks coming from Twitter (note that in this case the source showed up as twitter.com, possibly from a URL tagger as will be explained later, but generally Twitter links show up as t.co). Wow, not an impressive Twitter showing, huh?

Well, let’s take a look at bit.ly’s analytics for the link.

clicks from bit.ly

Bit.ly shows 47 clicks coming from the shared link. Granted, some differences will occur between bit.ly and Google Analytics (the GA view shown below filters out sessions from some IP addresses, and bit.ly is less restrictive about tracking bot clicks) but, it’s safe to say most of the 44 direct sessions recorded have been misattributed and are a result of users clicking on the link via Twitter, not visiting the content directly. This means the content received many more sessions from Twitter than was reported.

Because direct traffic lumps together any sessions where Google could not identify a specific source, you can see how easy it becomes for a URL shortener to hurt your ability to track a proper source. In the process of redirecting the URL from the bit.ly link to the yoursite.com link, the source gets lost and is seen simply as a direct visit. Some shorteners like Twitter’s t.co will pass on the referral source properly, but you can’t always trust the source to come through on its own. In addition to not showing correctly in referral reports, social visits from shortened URLs won’t be counted in the Social section of analytics. Marketers can use this section to get an excellent peek at sessions, conversions, and other stats from social media referrals but is useless without proper data.

The screenshot below shows referrals from Twitter in the Social > Network Referrals report, after clicking on Twitter to segment specifically to that channel. The second link on the list is the same infographic from the previous report, with just three visits reported here as well.

twitter referrals in google analytics

Ensuring Accurate Data

Now that you’ve seen the negative effect URL shorteners can have on your GA data, how can you ensure this doesn’t happen to you and that you get accurate data from links you share on social channels? First of all, you don’t necessarily need to shorten links in the first place. By using the true version of the URL, you’re most likely to see the proper source when users click the links. Twitter automatically shortens URLs via t.co, so you don’t have to worry about lengthy URLs eating up your character limit. For example, in this referral report t.co represents visits from Twitter.

referrals from t.co in google analytics

Shortened URLs also aren’t really an issue for Facebook, which allows a long enough character limit in posts to make shortening them unnecessary. Also, there’s little need to care about the extra analytics you can get from shortened URL services as long as you have access to Google Analytics which provides more extensive and more accurate data anyway.

However, you may still prefer to use a URL shortener. Maybe you’re concerned about lengthy URLs looking messy on Facebook. You can still use a shortener and get correct analytics data, as long as you tag your links with campaign parameters before shortening.

Even if you’re not using a URL shortener, tagging URLs before sharing is a good practice to establish. Referral data often gets lost for various reasons, especially for mobile visits (see this article for more details). When you tag your campaigns, you don’t have to rely on the app or browser to correctly send referral information to Google Analytics.

Because a tagged URL can look exceptionally long and unwieldy when appearing in a social media post, tagging combined with a URL shortener makes a smart approach.

To add these parameters, head on over to Google’s URL builder.

google url builder

This example shows the process of preparing a custom link to share a recent Megalytic blog post on Twitter. First, paste your URL into the “Website URL” field.

Next, insert the social network where you’re planning to share the link into “Campaign Source” (in this case twitter.com for Twitter). Medium should be broader than the specific site/source; in this example, “social” will tag any link shared to social media.

The “Campaign Content” field is not necessary, but we’ve included an abbreviated title for the article being shared to show extra info in analytics. Finally, the “Campaign Name” is a more general category we’re using for any links shared to the blog.

Click “Submit,” and you’ll see a final URL you can then copy. Paste that version of the URL into your preferred URL shortening service and generate your shortened URL. That URL (for example, bit.ly/*****), will then forward to the final link with all the proper tracking parameters tacked on. This will allow you to see visits accurately attributed to Twitter, as they should be.

Automatic URL Tagging

Note that some social sharing services offer an option to automatically tag URLs. However, Annie Cushing points out that some tools add inaccurate tags, causing just as much damage to your data as not tagging URLs before shortening. You’ll want to ensure that the tool you use adds the proper tags; if not, you’re better off turning off auto-tagging and manually adding the parameters.

Conclusion

In summary, you’ve seen the dangers of using shortened URLs without properly tagging them first. Don’t make the same mistakes with your social sharing, and ensure that your analytics data is coming through accurately. Make a plan with whoever manages your social media to either stop shortening URLs or else add tags before shortening.

Subscribe to us

  • I have tried to solve this problem caused by url shorteners by building a shortener with built in support for GA utm tags. http://utm.io

    A cool feature I am proud of is that it doesnt require you to create separate short urls for each website. Instead you can choose to have the utm_source parameter updated for each user based on the site the url was clicked on.

    • megalytic

      That’s cool – I see that the source is automatically set to the referrer. Nice.

  • Pingback: URL Shorteners are Messing up Website Analytics()

  • OURL.in is a quick url shortening website. Here you can shorten your url,

    which are easy to remember. Make short links or url using the form above.Create Short URL

  • Fastest and Shortest one I have found is http://snipac.com/ and it has a lot of features like link life duration, password protection, Custom Alias, Description, Geotargeting, QR Code and more.

  • While I agree that you should always use UTM parameters, URL shorteners are getting better at passing the parameters from the source. And they come in handy for improving the readability, branding, and tracking of links. Another risk of trusting the click data in Bitly might be that they aren’t removing bot traffic properly. It took us a while to figure out how to do it right with Rebrandly, and Bitly seems to want you to have to pay to scrub the click count properly. In that respect, I would trust GA data over Bitly data for real website visitors, as you can turn off bot and spider traffic in your GA settings.

    • markdhansen

      I don’t think it makes a lot of sense for a URL shortener to filter bot traffic. The shortner should do 1 thing – pass through the clicks. Bot traffic filtering should be handled elsewhere in the stack so that you can apply same technology across all sources of traffic.

      • Sorry, I mean filtering bot traffic from click reports. When you share a shortened link on Twitter, you want to know how many real humans clicked on it. If you report 45 clicks, when 40 of them were bots, you will be way over reporting to your boss. Lots of bad things stem from that.

        Bot traffic filtering, as in blocking a bot from passing through a shortened URL to the destination source, could have some problems, so I agree with you there.

        • markdhansen

          I get your point. Thanks for clarifying.

      • Just did a little test and it does appear Bitly is removing bot traffic properly for free accounts.

  • Peer Hits

    Some short links actually get you paid while visitors go through it. This is a great boost to income if you’re sharing income, and can be done with no harm.

    If you don’t want to shorten your website links, shorten a random link. I’m actually here to talk about getting paid from shortened links, and what my website offers- so feel free to read on if you would like.

    If you are using paid url shorteners to generate extra income, then it’s hard to get enough traffic to that link to generate any sort of income. We know, and now we have a solution!

    Peerhits.com is a website that lets you submit your paid link for others to click once you have been to 10 of your peer’s links. It’s completely automatic, and it’s free to use- so you can try it out right now to see how it works.
    Our site won’t be harmed by the links you upload, and you will get clicks on the URLs you put in forever. A one time browse will get you thousands of clicks, fully autopilot and for life, and it’s done with our website.

    However, we need people to visit the website, or we won’t have people to click your links, meaning you won’t earn any money. If there is a lot of people, you earn a lot of money, and so promoting the website or even visiting it (if everyone reading this went to the website, everyone would be able to get hits on their link)

    If you want hits to your paid URL, visit our website. You have nothing to lose, and after, you will gain traffic for life!

    http://Peerhits.com

  • Montana Ruark

    I’m noticing an issue with the GA UTM Link Builder’s readings versus the campaigns being tracked within GA. https://uploads.disquscdn.com/images/7551557b36184697e573042b66a738fec636655f6ac14c2b477a883eb5e8fffe.png

    • markdhansen

      Clicks (URL Shortener) are not the same as Sessions (GA). A Session (see: https://support.google.com/analytics/answer/2731565?hl=en) is a “visit” and consists of multiple page views. So, a person could click on the URL 3 times within 10 minutes and that would be 3 clicks, but only 1 Session. Here is an article about the discrepancies between AdWords clicks and Analytics Sessions: https://support.google.com/analytics/answer/1034383?hl=en The concept applies to the URL Shortener “click” metrics as well.

      Having said that – the Clicks here are > 3x the size of the Sessions – which is a big difference. Also, make sure that the date ranges are the same in both cases.

  • Victor Nwokedi

    What happens in a case where you manage a bbm channel knowing fully well that the characters can’t contain every thing and you need to insert the URL that will link back to your site when clicked??

    • markdhansen

      I’m not sure I understand the question, but sound like a you can use a link shortener to encode a URL that links back to your site. Just like you would on Twitter or anywhere you don’t want to or can’t use a long URL.

  • Angel Smith

    Have any of you used Taglynx to build utm links?

  • Musadhik

    i had same kind of issue with my e-mail campaign with two different results from Google URL Shortener (3K clicks) and Google Analytics (just 17 clicks)