Monday, June 29, 2009

Social Media Analytics Part I

What is Social Media?

There are so many ways people define social media and some even argue that it is not really media. According to Wikipedia, Social media is content created by people using highly accessible and scalable publishing technologies. Simply put, social media refers to all the conversation and engagement that happen on networks and sites like facbook, myspace, twitter, blogspehere, youtube, flicke, messageboard, forums etc..

Since these conversations can have a big impact on your brand it becomes critical for a marketer to understand what people are talking about their brand/products etc. so that they can take appropriate actions and help in creating and fostering a positive chatter (conversation) about their brand. Social Media analytics is about measuring and analyzing Social media (content generated by people throughout the web).

Broadly, Social media measurements comes in two flavors
  1. Measuring the conversation about your brand in social media

  2. Measuring the impact your own social media efforts (e.g. the facebook widget that you spent tons of money on or links that you posted on twitter etc.)

In this post I will talk about the first point, "Measuring what conversation are happening about your brand". I will have second post to discuss about second post.

Challenge with Social Media Measurement

Social Media measurement is very different from measuring your own web site (this is what most of the web analytics tool measure). You own your own website. You can (should) measure interaction of your visitors with your website. Social Media happens with our without your active participation. It mostly happens outside the realms of your website such as conversation on twitter, blogs, forums, facebook etc. Since you don’t have any web analytics tool installed on these places it is hard to find out what’s happening. Even if you had a web analytics tool installed you won’t know what people are talking about. Which is what you would like to know? This sort of information is not available from traditional web analytics tools like Omniture Site Catalysts, WebTrends or Google Analytics.

Social Media Analytics Tools

There are new breed of tools that help you monitor the social buzz. These tools let you “listen” into the conversation about your brand
In these tools you specify a set of keywords that define your brand or are associated with your brand and then the tools do the rest. They crawl the social media networks/sites and find all the mentions of the specified keywords and bring them back to you in nicely formatted reports.
The setup in most of these tools is a very manual process. Once the data is back you will needs a human to go through and analyze the data (not any different from your web analytics tool).

What kind of information do these tools provide?

Most of these tools bring some flavor of the following information (and much more)
  • Brand Mentions - Conversation about your brand/competitor/industry (as specified by keywords). You get total mentions by day/week/month and also the ability to drill down to a specific conversations.

  • Brand Sentiment – What is the consumer sentiment towards your brand? Are they positive, negative, neutral on your brand?

  • Influencers - Who is talking about you? How influential are they and how many times have they talked about your brand.

Some of the Social Media Analytics Tools

Below are the screen shots from Radian6 and SM2. Representatives from both these companies were very helpful in responding to my tweet and providing me the screenshot of their tools (see below).



In part II I will talk about how you can measure the impact your own social media efforts (e.g. the facebook widget that you spent tons of money on or links that you posted on twitter etc.).

I might also write reviews of some of the tools mentioned above, if you are a vendor of social media analytics tool and would like me to do a review please contact me.

Looking to fill your Web Analytics or Online Marketing position?
Post your open jobs on
Latest Job: Senior Web Analyst at Vml (Kansas City, MO)
Site: AnilBatra.comTwitter:
Have you used a URL shortner with real time analytics?

Tuesday, June 23, 2009

Widemile Optimize: A/B and Multivariate Tool

In my post titled “A/B and Multivariate Testing Landscape”, I reported on a variety of tools that marketers and analytics professionals were using for A/B and multivariate testing. As I’ve predicted the interest in testing and targeting/segmentation continues to grow. This was also validated by the Web Analytics Association’s “Outlook 2009: Survey Report”.

I just got through a good discussion and demo of the newest release of Widemile Optimize with Bob Garcia, VP Business Development for Widemile. I’m impressed with the new release of Widemile Optimization product that he demoed; this is the first Widemile release for corporate users. Widemile had released an Agency Edition of the solution last fall and the new release includes advanced segmentation and a refined user interface.

You might be wondering who Widemile is as they don’t have the same brand recognition as Optimost or Omniture Test&Target. Widemile has been around for several years and was one of the vendors listed in “A/B and Multivariate Testing Landscape” survey at eMetrics. I started working with them last year as they were rolling out their platform for agencies.

Widemile’s new release should help more marketers get started with testing as they’ve focused on simplifying the process while still offering more advanced users a lot of flexibility. Their new segmentation wizard makes it easy to create and manage visitor segments based on visitors browser information, Geo information etc. Bob showed me how you can clone a test run and set up a new test against specific segments in less than 5 minutes without any tagging changes.

Widemile uses fractional factorial methodology which allows for tests that use a subset of the combinations used in full factorial tests. Using a fractional factorial methodology allows the tool such as Widemile to reduce the amount of time required to get statistically significant test results.

Widemile must have read my A/B and Multivariate Testing Landscape post I mentioned earlier, which cited lack of budget as an impediment to doing testing, as they also announced an aggressive promotion (up to 50% off) for the first 20 customers that sign on. In closing, if you’re investigating your options for a professional testing solution Widemile should be on your short list.

Feel free to email me if you need help to figure out which testing solution will work best for you or need help in getting your testing going.

Looking to fill your Web Analytics or Online Marketing position?
Post your open jobs on
Latest Job: Senior Web Analyst at Vml (Kansas City, MO)
Have you used a URL shortner with real time analytics?

Thursday, June 18, 2009

Hits, Page Views, Visitors and Visits Demystified

This article is an introductory level and the intention of this article is to clarify few terms that you constantly hear in Web Analytics. Why am I writing this article? I hear some confusion about these terms from people new to field, so I thought I will write this blog post to clarify some of the common terms.

I am going to explain, Hits, Page Views, Visitors and Visits in this blog post.


Back in the early Internet days, Hits was a term commonly used to measure websites traffic. This term was mainly used by IT folks, early users of web analytics tools, to get an idea of the load on the server. As Web Analytics has moved into marketing and we have move to JavaScript based solutions, this term does not hold much meaning today as terms such as Page Views, Visits and Visitors have taken over.

So what is a Hit anyway? Let’s take an example of a simple web page shown below

This page is an html file with one image embedded in it.

When a person browses to this page (in her internet browser), she is requesting this page from the server to be downloaded to her internet browser. She views this page as one entity. In return browser is actually requesting 2 items from the server
  1. The actual HTML page

  2. The image embedded in it

When server returns these items, the browser assembles them and makes them look like one page to the person browsing this page.

This is what the log file of the server might look like (I have removed several items to make it simple) - - [16/Jun/2007:11:17:55 -0400] "GET /samplepage.html HTTP/1.1" 200 3225 "" "Mozilla/5.0 (Windows; U; Windows NT XP; en-US; rv: Gecko/20070914 Firefox/" - [16/Jun/2007:11:17:55 -0400] "GET /batman.jpg HTTP/1.1" 200 3225 "" "Mozilla/5.0 (Windows; U; Windows XP; en-US; rv: Gecko/20070914 Firefox/"

That means there were 2 hits on the server, one for the html page and one for the image. So with one page request there are 2 HITS (in this example)

All the above items will show up in your analytics reports if
  1. You use log file based solution

  2. You do not filter them out when setting up your reports

If you use a JavaScript solution then the only thing which is tagged (contains the JavaScript code) is the HTML page and that’s the only thing which will show up in the Web Analytics report.

Now let’s take a look at this sample again but this time we will look at the source to make sure there are no items hidden behind the HTML code. Sometimes (read most of the time) there are files that are not visible to the individual but still need to be downloaded from server and count towards the hits.

Here is what the source code looks like:

You will see there are two more files that are embedded in the page. One is a style sheet (stylesheet.css) and the other is a JavaScript (myjavascript.js) file.

So when a user requests this page, a total of 4 files are being requested from the server
  • The actual html page

  • The image embedded in it.

  • The .css file (stylesheet)

  • The .js (JavaScript File)

This is how the log file will look like - - [16/Jun/2007:11:17:55 -0400] "GET /samplepage.html HTTP/1.1" 200 3225 "" "Mozilla/5.0 (Windows; U; Windows NT XP; en-US; rv: Gecko/20070914 Firefox/" - [16/Jun/2007:11:17:55 -0400] "GET /batman.jpg HTTP/1.1" 200 3225 "" "Mozilla/5.0 (Windows; U; Windows XP; en-US; rv: Gecko/20070914 Firefox/" - [16/Jun/2007:11:17:55 -0400] "GET /stylesheet.css HTTP/1.1" 200 3225 "" "Mozilla/5.0 (Windows; U; Windows XP; en-US; rv: Gecko/20070914 Firefox/" - [16/Jun/2007:11:17:55 -0400] "GET /myjavascript.js HTTP/1.1" 200 3225 "" "Mozilla/5.0 (Windows; U; Windows XP; en-US; rv: Gecko/20070914 Firefox/"

If you are counting the Hits then there are 4 Hits on the server. It is evident it does not make a lot of sense to count Hits. Let’s look at what make sense (at least for now).

Page Views

According to Web Analytics Association Standards, “Page is an analyst definable unit of content”. Page Views is the number of times a page (an analyst-definable unit of content) was viewed.

So what does it mean? It means you can define type of file, Module, Flash interaction, PDF etc as a page and when a user views them they can be counted as Page Views.

Let’s use the above example and define a valid page as the files with .html extension only. When using a log file solution we configure the tool to filter out the other types of requests and only count pages with .html extension as valid pages. In a JavaScript based solution, all other types of files mentioned above (except .html in this case, if it has the JavaScript tag) will be automatically excluded from the Page View count.

So how many pages will the analytics report show? One, as there is only one html page. (You can configure your JavaScript based web analytics tool to track other forms of files as page views too but that requires customization).
The one page that is showed in the reports is a page view.

Visitors or Unique Visitors

Visitors or Unique Visitors, sometimes also referred as Unique Users is the number of unique individuals visiting a site. The most common way to identify an individual is via an anonymous cookie. Keep in mind that this is a close estimate of unique visitors and not an exact measure. Here are four examples on how unique visitor count can be wrong

  1. If two people use the same computer and same browser to visit a site, that identifies users by an anonymous cookie, both of them will be counted as one unique visitor since their cookie will be the same.

  2. On the flip side, if one individual uses two different computers to access the same site, the individual will be counted as two unique visitors because the new anonymous cookie will be issued on both the computers and show up as two different cookies in the analytics tool and hence will count them as two different visitors.

  3. If an individual uses the same computer but two different browsers (say IE and Firefox) then the person will be counted as two unique visitors because each browser will have its own cookie.

  4. If the individual visits the site, she will be counted as one visitor. Then if she clears her cookie and then visits the site again, she will be counted as two visitors.

Note: Visitors are calculated over a period of time e.g. day, week, month, year etc. and a visitor count from two periods can not be added together to get a total visitor count. Let’s take the data for following 2 days
Day 1 - 30 visitors
Day 2 – 45 visitors

The total visitors count for day 1 and day 2 is NOT the sum of the visitors count for the two days i.e. it is not 75 (30+ 45). Why?

For simplicity let’s assume that all the visitors who came to the site on day 1 also returned to site on day 2. In that case we will have 30 visitors from day 1 and 15 (45-30) on day 2 as unique between those two days, making the total unique to be 45 for the two day period and NOT 75.

The calculation I showed above has been simplified for this example. My advice is to let the analytics tool do the calculation for you and not sum the visitor count from separate period to come up with the total count of unique visitors.


Visit is also known as session. Visit starts when a visitor interacts with this site. In most case the interaction is the first page view by the visitor. The visit ends when user does not interact with a site for specified period of time. Most of the web analytics tools set 30 mins of inactivity as the end of the visit, however in most tools it is configurable and you can set it to whatever makes sense for your business.

Unlike, unique visitors, total visits to the site can be summed across time periods to get the total visit count for the period.

Hope this clarifies some of the confusion surrounding these terms.

Questions? Comments?

Looking to fill your Web Analytics or Online Marketing position?Post your open jobs on

Latest Job: Senior Web Analyst at Vml (Kansas City, MO)
Have you used a URL shortner with real time analytics?

Monday, June 01, 2009

Adding “bing” to Organic Search in Google Analytics

A while ago I wrote blog post on how to add Twitter searches to appear in organic searches. The same principal can be applied to include Microsoft’s new search engine “bing” which is not yet recognized by Google Analytics as a search engine. Till Google Analytics recognizes it as a search engine, you can capture the data in organic searches by a simple one line of code.

GA provides the following function to allow you to add your own search engines to the list of search engines that are already tracked by GA.

_addOrganic(newOrganicEngine, newOrganicKeyword)

You simply call this function right after var pageTracker = _gat._getTracker("UA-XXXXXX-X"); to track any custom search engine.

NewOrganicEngine is the words that identify the search engine; in this case we will use “”

newOrganicKeyword is the query string that contains that keywords, in this case it will be “q” as “bing” uses “q” as the query string that contains the keyword.

Here is how your final code will look like

var pageTracker = _gat._getTracker("UA-XXXXXX-X");
pageTracker._addOrganic("", "q")

Questions? Comments?

Looking to fill your Web Analytics or Online Marketing position?Post your open jobs on