Pin It

Web Analytics and Digital Marketing Blog by Sameer Khan

Quest for solving multichannel business challenges

Buffer

Myths exist in almost every industry and web analytics is not free from it. There are several myths in the web analytics industry, and we will debunk them one by one in this post.

#1 Web analytics is accurate – Web analytics is all about direction and insights. It is funny that we are still worried about the accuracy of analytics reports. We can spend days arguing with the marketing and sales managers on why the reports are accurate even though we all know that web analytics can never be accurate. Instead of focusing on this let’s start focusing on insights.

All web analytics reports are directional and getting insights is the key to web analytics success. Focus on direction, build trends, and apply statistical significance to the trends to get meaningful takeaways.

Statistical Control Limit

#2 Site Visitors delete cookies – It is extremely important to understand the difference between the first and third party. First party cookies are cookies from the host domain i.e. your domain. Top analytics tools (Google Analytics, Omniture, Coremetrics, WebTrends) all install first party cookies in order to measure the repeat visitors. The deletion / block rate for first party cookies is atleast 50% lower than the third party cookies.

Having said that, studies have shown that web analytics software cookie (first or third party) will be deleted no matter what. This is the truth about the tag based web analytics industry, and we have to accept it. This does not mean that we should stop trusting the reports and ignore them. The analytics reports are power packed and actionable because it can provide you insights if you follow the direction. As mentioned above the directional aspect of web analytics is unmatched.

So, stop worrying about cookie deletion rate and focus on the trend.

java cookie

#3 Users disable Java scripts – A major myth that goes around the market is about the blocking of the java script and thus prevent the execution of analytics tags. Java is an integral part of web and disabling java on the web browser can limit the accessibility to a large number of services on the internet.

Still not convinced? Here is a good study conducted by few developers showing why only 3-5% of the users disable java scripts.

#4 Google Analytics data for search ranking - Google analytics data is not used by Google to evaluate search rankings. Here is a quick video from Matt Cutt’s (head of Googles webspam team). In this video, Matt debunks the myth about Google using GA data for SEO or search ranking.

YouTube Preview Image

#5 Data addiction = actionable insights – Are you lost in data collection, integrations, and data obsession? If you are a pro analyst then you know that data dump, web integrations, data warehouse are critical issues. Please don’t get carried away by data addiction and focus on analytics priorities. Ask yourself “what is the most important and high priority task / project for you in next 5 days/ 5 months.?” Is it business growth insights or painstaking integrations? Is it few key recommendations to your management or never ending analysis?

I agree, analytics integrations can sometimes make your job lot easier but in most cases integrating your CRM and analytics tool is painful. Use excel reports and MS Access for connecting the dots. Leave the integrations for your web developer and business intelligence teams.

#6 Analytics code hampers site performance – I cannot deny that java scripts can increase the web page load time. However, there several options available to minimize the code impact on the site performance.

1. Open a Google webmaster tools account. Google has recently added a cool web performance measurement tool in the webmaster tools. It is the only free tool that measures your site performance relative to other sites on the internet and displays your historical performance in a trend line.

page speed

2. Install Yslow and Google page speed Firefox plugins. These plugins can instantly provide you detailed reports on how you can improve your site performance.

3. If you are using Google analytics then switch to asynchronous tracking. Asynchronous tag allows the browser to continue loading the web page while the tag is loaded in the background. This decreases the page load time.

4. For other analytics solutions, limit the use of onclick function to reduce server call connections. Tag only those pages that can provide insightful and actionable data.

#7 Unique visitors are absolutely unique – Your unique visitor reports are not really unique. It varies depending on the time frame you are looking at. You will have different data for hourly/daily/weekly/monthly unique visitor reports because of the way unique visitor is calculated. Using visits report with new and returning visitor segment will provide you more insights that you can ever imagine from unique visitor reports.

new vs returning (1)

#8 Time on page is accurate – The time on page can only be calculated if the web site visitor navigates to another page on the same website. What this means is you will never be able to figure out how much time a web visitor spent on the last page of the website. Stop worrying about the time on page metric and focus on conversion and task completion. A task completion could be clickling the link to the next page, subscribing to RSS feeds, clicking the buy now button or anything that page is motivating the users to do.

#9 Free Analytics = Enterprise Analytics – Investing in paid analytics solution is not a bad decision atall. You might have been motivated by someone on why you should never pay for analytics. The truth is far from accurate. Google Analytics is one of the best free tools out there, but it does not provide enterprise level capabilities that are required for advanced business analytics.

Here are the top 5 reasons why paid enterprise solutions win over free solutions-

1. Support – Paid solutions comes with support hours so you don’t have to spend time fixing analytics issues.

2. Custom Metrics – Paid solutions allow you to create custom metrics on the fly. You can create your own task completion and conversion metrics.

3. Multi-dimensional Timeline – Advanced solutions like Omniture Discover and Coremetircs Explore allows you to compare data from more than two time segments. Google Analytics can only provide two dimensional timeline comparison.

4. Analytics integrations – Web integrations become easier with the enterprise analytics provider because of their partnerships with email, crm, live chat and other service providers. Usually they provide integration hours or detailed documentation.

5. Campaign Attribution – Enterprise analytics providers store all the data on the data warehouses. The data can be accessible anytime and both first and last click is captured in most cases (depending on your settings). You can use this data to design a campaign attribution dashboard and allocate funds efficiently.

#10 Dedicated web analyst is not required – Paying $100K to your enterprise web analytics provider and expecting them to provide you all the answer is not going to work. Web analysis is the core of online marketing and without proper analysis and recommendations your business could come to a dead end.

dead end

Even though I love enterprise analytics I cannot stress enough that a dedicated analytics team is a must. Intelligent recommendations may be the future of web analytics but without an expert eye to decipher data and implement recommendations you won’t get results.

Hiring a web analytics expert is more important than investing in paid analytics tools and services. Period!

Please share your thoughts, insights and comments below.

Here is a list of the related top blog posts on this blog -
Ultimate Web Analytics Training Guide: From Click to Close
Top 5 Coolest Web Analytics iPhone Apps
Copy My Actionable Google Analytics Advanced Segments
Top 3 Worst Web Analytics Metrics & Reports

8 Responses so far.

  1. Interesting post.

    Your assessment of accuracy is spot on. It looks like you are already aware of my whitepaper on the subject of web analytics accuracy, though I hope you don’t mind if I post a link here to it for your readers.

    Its a vendor agnostic document that details all accuracy issues when considering web analytics:
    http://www.advanced-web-metrics.com/blog/2010/04/23/understanding-web-analytics-accuracy/

  2. As a follow up, I wanted to comment on item #9 – Free Analytics = Enterprise Analytics.

    As you might expect I disagree (quite strongly!) on this…

    In fact, many enterprises are using GA at an enterprise level. That is, doing great things in terms of optimising their online marketing and visitor experience. Here is a list of some of them:
    http://www.advanced-web-metrics.com/blog/who-uses-google-analytics/

    For example, P&G have been using Urchin and GA for many years…

    I am *not* suggesting GA can do everything – but then no web analytics tool can – free or paid. However, I wanted respond to your arguments for your hypothesis:

    1. Support
    Free Analytics refers to the data collection and reporting of web visitor data. It has never been intended to include free training, consultation, optimisation or any other form of professional support services.

    As you would expect, you hire expertise for this. However, the DIY option is also available with Google Help Centre (the world’s largest resource on web analytics – and free), the GAIQ learning centre (something I have a personal interest in) and of course numerous books.

    2. Custom Metrics
    If you wish to create a custom metics (one number added/subtracted/multiplied/divided by another), the recommend route is to pull the data via the Export API, then do as you wish.

    3. Multi-dimensional Timeline
    Using Advanced Segments, GA allows you to compare up to 4 metric timelines

    4. Analytics integrations
    GA integrates very tightly with Adwords, Adsense and Feedburner. Being Google, no other vendor can really compete with that. Look out for future integrations with pretty much all of G’s products – as I predict here:
    http://www.advanced-web-metrics.com/blog/2009/04/16/google-analytics-fours-years-on/

    5. Campaign Attribution
    GA can use the first click attribution method for paid campaigns. However the correct assignment of attribution is a multi-dimensional problem for the entire industry. No vendor has cracked this and to be honest, without microchips implanted into web visitors (or forcing users to log in to every website), this is unlikely to ever be solved.

    Thanks for allowing my comments.

    Best regards, Brian
    Author, Advanced Web Metrics with Google Analytics

    • Sameer says:

      Hey Brian,

      Thanks for the detailed comments and your comments are always welcome. I am your fan and have read your Advanced Web Metrics book twice. I totally respect your suggestions on why someone should go for free analytics. Quite honestly, I love both solutions so for me personally, I don’t mind using either one for day to day analysis. However, I still think GA has to cover a lot of ground areas before they can compete head to head with enterprise solutions. Most of the counter suggestions you provided are focused on manual work and are not available out of the box.

      1. Support – I agree that GAIQ is the largest web analytics learning center. There is no question that the number of resources Google provides can be compared to anything. Having said that I think support piece is bigger and mandatory. I cannot think of operating enterprise scale analytics without having access to developer support. We can either hire a dedicated web developer or hire a company to help with coding. Analytics implementation is never complete and it scales with the organization.

      2. Custom Metrics – Pulling data from API and applying to excel calculation is not feasible when it comes to handling 20 different metrics. Custom metrics is absolutely required to generate reports on the fly for each metric. Small calculations like task completion rate, ROI, and goal to visit are critical for decision making process.

      3. Multi-dimensional Time line – To clarify, I was referring to comparing more than 2 time segments in the single screen. Google Advanced segment only allows comparison of 4 metrics for the same time line. GA does not allow the functionality to change the time line for any of 4 metrics. Tools like Omniture Discover and Coremetrics Explore are built for changing time lines on the fly for each metric individually.

      4. Analytics Integrations – Integrating with non Google products are the most important element of multi channel analytics. We currently integrate our web analytics software with impression attribution tool, Salesforce, third party SEM software and chat tool. This allows us to look at the web performance in a holistic view and allows us to build end to end reports.

      5. Campaign Attribution – I agree that campaign attribution is a big problem for the entire industry but with the help of analytics integrations, we are able to solve a lot of attribution issues. We recently published our successful integration story of media server Eyeblaster and Omniture. This integration has allowed us to double our media buy budget.
      http://www.omniture.com/offer/711

      Best
      Sameer

  3. [...] Top 10 Outrageous Web Analytics Myths..Debunked Ultimate Web Analytics Training Guide: From Click to Close Facebook vs Google & Google Analytics – Final Round?? Top 3 Worst Web Analytics Metrics & Reports swfobject.embedSWF("http://www.youtube.com/v/8LQrSiqw_Zs&rel=0&fs=1&showsearch=0&showinfo=0", "vvq-725-youtube-1", "425", "344", "9", vvqexpressinstall, vvqflashvars, vvqparams, vvqattributes); Visitor Behavior, Web Analytics Basic, Web Analytics Integration, Web Analytics Segments, Web Analytics Tools, Web Analytics Training [...]

  4. emt training says:

    Great site. A lot of useful information here. I’m sending it to some friends!

  5. [...] no one understands and present insights with actionable key takeaways. Web Analytics is full of myths that originate from unnecessary data [...]



nine + 2 =

Subscribe to email feed

  • RSS
  • Linkedin
  • Facebook
  • Twitter
  • Google+

Solving Digital Conv

Last time I wrote a post about the customer lifecycle ...

Forrester Web Analyt

Every three years, Forrester Research conducts industry's most thorough study ...

5 Digital and Analyt

If you are a people person and a learner like ...

How to Overcome Goog

Google is known to cause massive stir in the online ...

Top 5 Strategies for

Here are the presentation slide from my latest speaking engagement. ...

Solving Digital Conv

Last time I wrote a post about the customer lifecycle ...

Forrester Web Analyt

Every three years, Forrester Research conducts industry's most thorough study ...

5 Digital and Analyt

If you are a people person and a learner like ...

How to Overcome Goog

Google is known to cause massive stir in the online ...

Top 5 Strategies for

Here are the presentation slide from my latest speaking engagement. ...