luni, 18 octombrie 2010

SEOmoz Daily SEO Blog

SEOmoz Daily SEO Blog

Traffic "Bait" and Ad Clicks: Perfect Market's Study Isn't Telling the Whole Story

Posted: 18 Oct 2010 01:42 AM PDT

Posted by randfish

Yesterday, Perfect Market, a company that "helps publishers create value from their online content with little effort and no risk1" released a study that's been getting quite a bit of attention. The study analyzes the relative traffic value per visit of several types of content, coming to the conclusion2 that "while the Lindsay Lohan sentencing and other celebrity coverage drove significant online traffic for major news publishers, articles about unemployment benefits, the Gulf oil spill, mortgage rates and other serious topics were the top-earning news topics based on advertising revenue per page view."

Coverage included the New York Times' Traffic Bait Doesn't Bring Ad Clicks, Columbia Journalism Review's Celebs are Loud, but Hard News Pays, Nieman Journalism Lab's Public Interest News Can Be More Valuable to Publishers than Traffic Bait and Search Engine Land's Hard News Pays More than Chasing Search Trends.

I'm worried for a few reasons:

  1. What's the branding value of those stories? Do they drive up awareness of the publications that authored them? Do they increase return visits?
  2. What other actions do those visitors take? Are they more likely to subscribe to an RSS feed? To share those stories on social networks? To get email notifications?
  3. Do these stories drive links that then help other, lower link-earning content rank well in search engines? The goal of linkbait, after all, is often to drive branding, links and sharing rather than being directly monetizable. Plenty of consultants on viral content creation even recommend removing ads to drive up sharing and linking activities.

Granted, from a personal perspective, I love the idea that writing about celebrity gossip and other "soft news" isn't profitable and therefore might be less prevalent in the future. It's purely opinion, but I suspect that many share my sentiment that the United States' major media outlets are far too focused on shallow reporting of topics (like those mentioned in the Perfect Market analysis) that deserve far less attention than, say, understanding what caused the mortgage crisis, who's spending money on elections and why, the success other nations have had in dealing with crime, poverty, drugs, multiculturalism, etc.

However, anytime a skin-deep, single-metric analysis like this makes its way into major publications, it has an effect on content publication that's not necessarily positive. If executives, editors and journalists start using singular metrics rather than deep analyses of data to make decisions, their publications will suffer and their content and marketing budgets will be misallocated.

If Perfect Market (or another source) could show:

  • The value of the links brought in from those stories
  • The branding impact of the visits generated
  • The value of sharing activities from those visits

I'd be far more inclined to agree with the conclusions the press is reporting.

If you can't fully/accurately analyze the true lifetime value to your publication of so-called "bait" (and I don't just mean celebrity-obsessed soft news, but a broader group of creative, traffic-driving pieces), that's OK. Just don't presume a single metric like "ad click value" combined with "page views" will give you the whole story. The web is all about providing data, and you're cheapening your own value when you cut corners to this extent.

BTW - I don't mean to cast all the blame on Perfect Market - they did some reasonable data analysis and shared the findings. I wish it had included a bit more caveats, but their job is promoting their work. I'm more concerned with how the media treated the story - reporting, exaggerating and not bothering to dig deeper. Just look at the opening lines of the NYTimes piece3:

Sure, articles about Lindsay Lohan’s repeat trips to rehabilitation and Brett Favre’s purported sexual peccadilloes generate loads of reader traffic, but do they actually make decent money for the Web sites that publish them? According to a new analysis, no.

That's not what the analysis showed. It showed one metric and it's impact, but it didn't explore the overall value of the page views, visits and CLTV (Customer Lifetime Value) of the stories it examined. Let's hope the publishers do a more thorough job and that we, as content creators & marketers, think carefully about how to value the content we create and the traffic we attract.


Do you like this post? Yes No

5 Quick Google Analytics Hacks

Posted: 17 Oct 2010 12:52 PM PDT

Posted by Tom_C

1) Regex for Counting " " and "/"

Regex is awesome. I don't claim to be amazing at it but there are a few common regex strings I use all the time in my analysis.

Length of Keyword

To quickly filter your keywords report by the length of keyword, I use some regex to count the number of spaces in the keyword like this:

^([^ ]+ ){5,50}[^ ]+$

The above regex searches for keywords that have between 5 and 50 spaces in them. You can also search across a single number as shown below. This image is a search for all keywords with 6 spaces in them for the distilled site (i.e. 7 words):

Depth of Page

Very similar to the above regex, but when I'm looking at top landing pages I use regex like this to count the number of slashes in a URL:

^/([^/]+/){3}[^/]*$

Note that because I'm not a full regex ninja this actually counts those URLs that have 4 slashes in (i.e. n+1). So the following image is showing all traffic to those pages with 5 slashes in them:

Note how useful this search is? Pretty much all of these pages are low quality like pagination or blog pages that have multiple categories assigned. For large sites if you construct the regex correctly this can be a great way to analyse where traffic is landing on the site and identify low quality pages to remove from the index.

If you're new to regex - this is my goto guide for using regex in Google Analytics (PDF).

2) Check Your Analytics Code Is Correctly Installed

This is a super easy one, but definitely one worth running on any new site you take a look at. SiteScan will crawl your site and check for the analytics code which is pretty nifty. It even intelligently checks for the old and new versions of the GA code. Nice. Unfortunately the free version only checks 100 pages but it's definitely a solid resource for smaller sites:

Another quick check for correctly installed Google Analytics is to look for referrals from your own domain. Any referral from your own domain indicates that there are pages not correctly tagged (and will even show you which ones!). Nice.

3) 5 Ways to Segment your Funnel

Segmenting your funnel is not something you can do natively in Google Analytics which annoys the hell out of me. I'm hopeful that Google will be adding this feature sometime in the near future. In the meantime, there's a few ways to segment your funnel:

Why do you care about segmenting your funnel? Well I give a detailed run-down of why this is important over here but hopefully this image should explain itself (the output of segmenting the funnel using my method):

4) Track SEO Variables In Google Analytics

This is a nifty use of custom variables which I recently started using on a few sites. Imagine you're running a hotels reviews website. Some of your reviews have 100s of reviews and are lovely content-rich pages. But some of your hotels are awaiting their first review. In that case, your hotel page might be very light on content and might only have the name and address of the hotel on the page (which is duplicated on 100s of other sites). Wouldn't it be nice to be able to segment your Google traffic by how many reviews your hotel page had? Well using page level custom variables this is as easy as the following code:

_gaq.push(['_setCustomVar',
          1,                   // This custom var is set to slot #1. Required.
          'Num_Reviews',       // The name of the custom variable. Required.
          0,                   // Sets the value of "Num_Reviews" to 0. Required.
          3                    // Sets the scope to page-level.  Optional.
         ]);

You don't have to limit yourself to just using this for number of reviews, you could look at other factors that you think might be affecting your pages ability to rank and pass those into GA. For example, you could pass the length of the description of a page. Or the number of tweets it has or anything you can think of really!

Learn more about page level custom variables over here.

5) Track Form Abandonment

This one comes from a blog post Duncan wrote a little while back, but I love how simple this is to use and how useful the insight is. Basically, using jquery it becomes very easy to track how far through a form people get. The idea was prompted by Sam's post from some time ago, but uses events instead of virtual page views.

You should read the full write-up on Duncan's post but the code looks something like this:

1.  $(document).ready(function() {   
2.      var currentPage = jQuery.url.attr("path");  
3.      $(':input').blur(function () {  
4.          if($(this).val().length > 0){  
5.              pageTracker._trackEvent("Form: " + currentPage, "input_exit", $(this).attr('name'));  
6.          }   
7.      });  
8.  });

Bonus!

While writing this post, one of Dave Naylor's gang posted about a new interface for in-page analytics which replaces the old site overlay. I'm quite excited about this, I think it paves the way for all kinds of cool things (not least of which is heatmaps as David points out...)


Do you like this post? Yes No

Publicat de 1 la 17:50

Niciun comentariu:

Trimiteți un comentariu

Abonați-vă la: Postare comentarii (Atom)