Unique Visitors In a Multi-Device World


How important is the  Unique Visitors metric to you? What if I told you it’s not as accurate as you thought it was? What if it was off by 50%?

As more people shift toward browsing not just on their computer, but on their phones and tablets too, visitor-based metrics become less accurate. This is because a visitor in analytics isn’t an individual person (that’s what we’d like!) but a unique browser on a unique device. Each browser on each computer/device uses it’s own set of cookies, so if I visit your site from my work computer, then visit again at home on my laptop, I’ll show up as two unique visitors. Or if I go to your site using Firefox, then again with Safari, and a third time using Chrome (all on the same computer), you guessed it – three visitors.

Let’s look at another example. You just launched a new product or service and have a hilarious video on your homepage to get people talking (think Dollar Shave Club). I hear the buzz on Twitter, so I go to your site to check it out on my work computer, using Firefox (1).

Later at lunch, I tell my coworkers about this hysterical video and pull up your site on my phone (2) to show them. It loads, but the screen is too small. So I grab my iPad (3) and go to your site so everyone can see it clearly.

Later, I find myself in the after-lunch sleepy period and need a wake-me-up. So I grab a cup of coffee and go back to your site again, this time using Chrome (4).

When I get home, I tell my wife about the funny video and grab the laptop to show her. I type the address in the browser that’s already open, unfortunately it’s IE6 (5) and the site doesn’t load quite right. So I open Firefox and go back to your site (6).

We have many laughs. But you’re not laughing because I just racked up six unique visitors in your analytics. Six. But I’m just one person.

This isn’t a new problem, but one that is commonly misunderstood. Unfortunately, there’s another twist that is starting to exacerbate this problem. Apps.


Have you ever tapped on a link from within the Twitter app and it opens the page – not in Safari, but still within the app itself? That app uses a  browser that uses its own cookies too.

And if you’re like me, you prefer to view the page in Safari, not the Twitter app. So you go to the bottom right corner and tap on the little box and arrow, then tap “Open in Safari.” Oops, now I’m a new visitor again. Bummer.


Traffic Sources

Here’s the double whammy of app traffic (traffic to your site from apps, not traffic to your apps). Many of your favorite apps show up as Direct visits in your analytics. In the scenario above, tapping on the link in the Twitter app will show up as a direct visit in GA instead of a referral visit from Twitter. Which apps behave this way, you ask? Check the table below:

App Medium/Source (iPad) Medium/Source (iPhone)
Twitter Direct Direct
Facebook Direct Referral / m.facebook.com
Pinterest Direct Direct
LinkedIn Direct Direct
Google+ Direct Direct
Google Reader Direct Direct
Gmail Direct Direct
Google Docs Direct Direct
Evernote Direct Direct
Pocket (formerly Read It Later) Direct Direct
StumbleUpon Direct Direct
HootSuite Direct Direct
TweetDeck Direct Direct

Note: the test above is not all inclusive – there are hundreds of thousands of apps and I didn’t have time to check them all! Also, I only checked from an iPhone and an iPad (both running iOS 5.1.1). I did not check an Android device – I’ll leave that to you, our faithful readers. Please share in the comments what you find!

In a recent post by Josh Davis, he reports that as much as 38% of Pinterest traffic is not showing up as coming from Pinterest. Although the amount of under-reporting depends highly on the traffic source and industry, this is a trend that is likely not going away.

Jim Gianoglio is a Manager for the Analytics & Insight department. He works with implementation, analysis and training of Google Analytics and Google Tag Manager. Before focusing on analytics, he led the SEO campaigns of Fortune 500 companies in the insurance, retail and CPG industries. Things you didn’t know about Jim: he’s biked from Pittsburgh to Washington DC in 41 hours, roasts coffee beans and has done voiceovers for TV commercials.

  • I don’t know how it works with other apps, but if I click on a link in Tweetdeck it will send me to a t.co URL. And from there to the final destination where t.co will be the reffering source in stead of direct.

  • @André – are you on an Android device? When I tested Tweetdeck on iPhone and iPad, it showed up as direct.

  • I always presented that metric as “Unique browser identifiers”. What is true about device, is also true about browser. Two persons using the same browser. One person using two (or more) browsers. Without authentication, their is no way to get a count of unique visitor. The trends was still usable, but now, with the proliferation of devices, maybe it’s not so true anymore.

  • Thanks for the post, Jim. But in my opinion the describe issue with the unique visitor metric is just the tip of the iceberg. The real issue in a multi-device world is that all classic ROI measurements and all classic customer journey calculations (Or multi-channel attributions) are off.

    Why? Because they are device (or even browser/application) centric. But if I’m a performance based marketer and I see traffic acquired through mobile campaigns, but people convert later on their work PC, then I have a really hard time understanding if my traffic acquisition costs on the mobile channel are justified.

    This is particularly hard for retailers and sites that don’t have a login functionality. And it could be one of the main drivers, why unobstrusive logins like Facebook Connect can drive real value, because they allow unique ID tracking across devices.

    This all comes with real privacy issues again and the industry will have to figure out a way to stay attributable, without intervening with users privacy expectations.

  • Sébastien and Bjoern – thank you both for your thoughtful comments!

    Sébastien – I agree. I think the overall trends are still useful, but I wonder how long that will be the case.

    Bjoern – you hit the nail on the head (and provided much better examples than I did in my post – thank you!). I think this is one of the top issues digital analysts and marketers are facing today. I know there are a lot of brilliant minds working on this, and I look forward to their solutions.

    In the meantime, Facebook connect seems to be the best we have.

  • Jim,

    I will be writing an essay on the issues of multi-device tracking in the “100 Experts Book” due to be out later this year (http://100experten.de/). If you’re interested, we could probably translate that for the English speaking audience as a guest post on Lunametrics.

  • Yes, of course we would all be interested. Thank you for you insite.

Contact Us.

Follow Us




We'll get back to you
in ONE business day.
Our Locations
THE FOUNDRY [map] LunaMetrics

24 S. 18th Street
Suite 100

Pittsburgh, PA 15203


4115 N. Ravenswood
Suite 101
Chicago, IL 60613


2100 Manchester Rd.
Building C, Suite 1750
Wheaton, IL 60187