Google Analytics shows most referrers as "Direct" -- What are some better tools?
-
Very often Google Analytics will show 50-90% of our referrers as (direct) which is not very helpful.
Are there other tools out there that will provide a clearer breakdown of what other websites are sending us our traffic?
Specifically, I want to be able to be able to tell who are the top traffic referrers to my top performing pages on my site for the last 30 days. (I want to be able to study this on a per-page basis.)
Thanks in advance!
-
Thank you for the amazing answer Robert. I followed your advice and checked for the GA code. Fortunately it is indeed placed in the header before the /head code, and it is only placed once.
I'm very grateful for this information, for at least now I don't have to worry that we've somehow implemented GA incorrectly on our site.
-
Killer
I saw this in the Moz Top 10 this morning and had to reach out to you. Very interesting stuff on organic traffic showing as direct. While not referral showing as direct, I think you will see the implications. Best
-
Killer,
Sorry I fell out of the loop on this one. I have to say that Takeshi has done beyond an above admirable job on laying out the issues of how referrers become direct. I think two things stand out to me: yes the installation of your GA code could be wrong and it should be before the closing head tag . Often people will put it in the footer and occasionally someone enters it more than once.
I think another thing you have to look at is you are comparing Chartbeat to GA and they provide two types of analytics. The numbers, of course, are different. Chartbeat was out before GA had realtime and, I think, it was intended for the real time needs. It claims to have better data for publishers and I have no way of comparing. Since it has been around for some time, I would think it has some real traction, but you have to be careful as you are comparing two different services.
Another thing that gives me a slight amount of concern about the issue is that they have a section on how they help you sell more ads (when you are a publisher). That type of data makes me a bit nervous. It may be the real deal, it just makes me nervous when something claims to help you prove your position.
The question re your site is a bit skewed in the way you present it. What is the other data re that site? How many total directs? How many organic visits? How many PPC visits? Etc. You cannot just conclude that there cannot be that many directs without all the data. How many total visits would be the first place to look.
As to the expert on analytics and your GA code, you can do it your self. Look at view page source (typically a right click function). When you have the page with the code on it, do a control lookup (command or Control plus F on). Then in the search box, type in UA- and see where it occurs. There should be only one occurrence of UA-XXXXXXXX-1 or -12, etc.
If you see it more than once it is likely a problem, but not always. If it is in the header and footer, typically you will see an artificially low 2% or less bounce rate. You can also do a lookup of the closing head tag and see if it is after the analytics code.)
I hope this helps you a bit.
Robert
-
You know what, that is a GREAT question. Maybe we should hire a professional to check up on our Analytics settings/scripts/installation to make sure it is all working properly.
Is there a special designation I should look for when hiring someone like that? Is there such a thing as a "certified analytics expert" or something?
-
Interesting. Do you know if Google Analytics is installed properly? Is the code in your header? Are there any javascript errors on your page? Do you have multiple analytics scripts installed?
May be some kind of tracking issue there.
-
Hi TakeshiYoung,
Thank you for taking the time to answer my question! When I was first presented this problem by my editor (who is trying to use GA), I thought the same thing as you -- that if GA couldn't find the info then it probably can't be found by any other tool.
However, my editor presented me with this example: one recent blog posts had 6,000 pageviews. Google showed referral data as:
- 5,000 (direct)
- 344 off.net.mk
- 113 facebook
- 21 feedly
Using another tool called Chartbeat on the same blog post, we got way clearer data:
- 3,410 from off.net.mk
- 1140 from fb
- 425 from feedly.
We think the 5,000 direct number from Google must be wildly inaccurate. How could 5,000 out of 6,000 pageviews be direct?
-
Hi Robert,
Thank you for taking the time to answer my question! Your rephrasing of my question is correct. Here's an actual example provided to me by our editor who is trying to figure this out.
For example, one recent blog posts had 6,000 pageviews. Google showed referral data as:
- 5,000 (direct)
- 344 off.net.mk
- 113 facebook
- 21 feedly
Using another tool called Chartbeat on the same blog post, we got way clearer data:
- 3,410 from off.net.mk
- 1140 from fb
- 425 from feedly.
We think the 5,000 direct number from Google must be wildly inaccurate. How could 5,000 out of 6,000 pageviews be direct?
I hope this helps!
-
Not really. Google only shows referrers as (direct) when there is no referral data present, and that's an issue with the referrer, not your tracking system. Referral data can be lost for many reasons:
- The user is getting to your site by typing the URL directly into their browser
- The user is getting to your site via a bookmark or browser favorite
- The user is using iOS6, which doesn't pass referrer data properly for search traffic
- The user is clicking on a link from an e-mail client which doesn't pass referrers (common if you have mailing list)
- The site that is linking to you is using HTTPS, and you are using HTTP
- Traffic is coming via a mobile app that doesn't pass referrers
- The user is using IE and they clicked on a Javascript or Flash link
- Etc.
You may be able to get around this to a certain extent by using URL parameters (utm_source, utm_campaign), but those would only be for links that you have some control over.
-
Killer
When you say referrers are showing as direct, what are you meaning by that?
If I type in your domain and enter, I will show as direct.If I go to site A (say YP.com) and click a link to your site it will show as a referral from YP.com.
The way you are saying this, if you are getting 1000 visits in a fixed period it shows as 500 organic, 450 direct and 50 referred. You believe the 450 direct are actually more like 50 direct and 400 referred, is that correct?
Best,
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
I am confuse with google analytic custom and segment report
Hi All, In google analytic when I create custom report for my ecommerce site then figures go mad. I really not able to judge peformance of device, browser and it's version, conversion, ecommerce conversion rate etc. same way if I add secondary dimension in report then also figures are not accurate. Again when I create different segment like desktop, mobile, tablet then in tablet segment mobile devices comes and in mobile device tablet appear why segment also not accurate? Is it because I am using free version? Also do we have alternative of google analytic which give same report like google analytic either device, browser, os, segment, enhance ecommerce etc? Thanks!
Reporting & Analytics | | dhisman0 -
Google Analytics goals only show unique event triggers
I have some goals in GA which is linked to different event triggers. However I've found out that the goal is only showing the unique event triggers and not the total. I need it to show the total in order to be accurate. Is there any way to switch that option? I can't find any options like that in GA or guides to do so.
Reporting & Analytics | | WebGain0 -
Visits from Google with ccTLD are showing as referrals
Hi!I was seeing on of my clients Analytics report and it shows that some of the main sites that send visits and get tagged as referral traffic are google.com.br, google.cl, google.com.ar, among others. Do you know why is this happening? Shouldn't they get tagged by default as organic?
Reporting & Analytics | | arielbortz0 -
Identifying Bots in Google Analytics
Hi there, While you can now filter out bots and spiders in Google Analytics, I'm interested in how you identify a bots and spiders in the first place. For example, it used to be thought that Googlebot wouldn't appear in GA as it 'couldn't process Javascript' but now Google has announced new developments for its crawler with regards to interpreting javascript and CSS, this argument isn't as cut and dry. I'm not suggesting Googlebot appears in Google Analytics, but I am saying that you can't make the case that it won't appear only because it can't interpret JavaScript. So, I'm interested to see what metrics you use to identify a bot? For me, the mix of Users > Browser, Users > Operating System Version is still quite handy, but is it possible to identify individual bots and spiders within Google Analytics? And would Googlebot appear?
Reporting & Analytics | | ecommercebc0 -
Google Analytics Organic Accuracy?
Hi, On a cool summers day, I was quenching my thirst for SEO Knowledge in the depths of Google analytics, When I stumbled upon a most troublesome sight. I was Browsing the data inside Acquisition->Keywords->Organic and comparing them to Acquisition->Search Engine Optimization->Landing Pages, to find the results quite questionable.. Since Keywords->Organic measure the visits from the organic results, and Search Engine Optimization->Landing Pages measures the clicks onto one of my landing pages from Google, I assumed they should be the same or very close, (besides Keywords->Organic having visits from other search engines) but they were not the same, as a matter of fact it was telling me that Keywords->Organic visits = 1,474, 27.19% of total(5,422) and Search Engine Optimization->Landing Pages, clicks = 1,548 154.80% of total (1000). Even worst the Search Engine Optimization -> Queries clicks = 1,123, 160.43% of total(700). I have been searching, and come up empty handed in the answer so does anyone know why Google is showing me different results and strange numbers that don't make sense? Any information is greatly appreciated?
Reporting & Analytics | | KBB_Digital0 -
Why am I showing a difference in Goal completion for Adwords and Analytics?
I have conversion tracking setup for Adwords and Goal tracking setup in Analytics. There is a large discrepancy between the two. Anyone have any ideas?
Reporting & Analytics | | EcommerceSite0 -
How "Assisted Conversion value" is calculated on Google Analytics?
Hi guys, Can any of you tell me how the Assisted Conversion Value is calculated under Assist Interacion Analysis? For the channel Direct, we have Assited Conversions: 6 Assited Conversion Value: 451,59 Last Click or Direct Conversions: 10 Last Click or Direct Conversion Value: 786,54 Assisted/Last Click or Direct Conversions: 0,60 How the value 451,59 is calculated? Thanks. Regards
Reporting & Analytics | | footd0 -
Google Analytics Site Search to new sub-domain
Hi Mozzers, I'm setting up Google's Site Search on a website. However this isn't for search terms, this will be for people filling in a form and using the POST action to land on a results page. This is similar to what is outlined at http://support.google.com/analytics/bin/answer.py?hl=en&answer=1012264 ('<a class="zippy zippy-collapse">Setting Up Site Search for POST-Based Search Engines').</a> However my approach is different as my results appear on a sub-domain of the top level domain. Eg.. user is on www.domain.com/page.php user fills in form submits user gets taken to results.domain.com/results.php The issue is with the suggested code provided by Google as copied below.. Firstly, I don't use query strings on my results page so I would have to create an artificial page which shouldn't be a problem. But what I don't know is how the tracking will work across a sub-domain without the _gaq.push(['_setDomainName', '.domain.com']); code. Can this be added in? Can I also add Custom Variables? Does anyone have experience of using Site Search across a sub-domain perhaps to track quote form values? Many thanks!
Reporting & Analytics | | panini0