Blog III: Web Analytics

Analytics … Analytics … A word that we have been hearing quite a lot lately. So what exactly is Analytics? According to Wikipedia, analytics is defined as the discovery and communication of meaningful patterns in data. If that what analytics is defined as, then what is Web Analytics? Is it as simple as the discovery and communication of meaningful patterns in data on the web/Internet? According to one of the Web Analytics guru, Avinash Kaushik, web analytics is the analysis of qualitative and quantitative data from a website and the competition to drive continual improvement of current and potential customer’s experience.   

Web analytics is usually part of customer relationship management analytics, or short CRM analytics. The Web Analytics software tracks every instance and action that is happening on a website and it is recorded in real-time. The analysis can include items such as re-designing the website to make it more personable to frequent visitors/customers; monitoring purchases and volume by specific customers or group of customers; determining likelihood of a customer re-purchasing a product; exploring the demographics of customers, such as what are the regions from which least and most customers visit the site; predicting what items are customers more likely to purchase in the future; and so on.

All this is mind-boggling isn’t it? Where is this technology going? We can track and review everything? You know when Facebook suggests some sites for you, like all kinds of shoes and clothing sites on my FB page of course? I used to wonder, how does Facebook knows what I like, why is it showing ads where I will most likely spend all my money onJ? Now, thanks to this week’s class I know how Facebook and other sites do it. All my actions on the web are tracked! Hmmm, this could be good or bad … And I will talk more about the bad side of Web Analytics in a bit.

Next, since we now know what Web Analytics is and what it can do, lets focus on Web Key Performance Indicators or KPIs (note: the list is not exhaustive). First of all, to avoid confusions, we should define KPI so some of you are not thrown off by the word. KPIs are measures that help an organization track its successes and failures in accordance to the organization’s already defined objectives.  

·        Conversion Rate – proportion of visits that result in goal achievement. For example, if Google’s goal is for a web user to click on an ad campaign, then you will calculate how many visits on the Google site result in achieving that goal, which is the conversion rate. This metrics is very valuable KPI as it steers the organization’s focus on Objectives.  

·        Task Completion Rate – percentage of visitors that successfully completed a specific task on the site. For example, if Business Insider’s goal is site visitors to download an article then the Task Completion Rate will be the percentage of visitors that successfully complete the download of articles. This metrics will show how easy is for visitors to perform actions on the site and it will give suggestions for web re-design, on how to make it more visitor-friendly.

·        Average Order Value – monetary value of sales per conversion. For example, if Fabletics’s visitors click on yoga pants and buy them, what is the revenue Fabletics gets from each conversion? AOV goes hand in hand with the conversion rate - it will help an organization more clearly understand why the revenue is down when the conversion rate is high and vice versa.

·        Exit Rate – The percentage of visitors that leave the website from a particular web page. The exit rate is calculated for a particular web page. For instance, the percentage of visitors that leave the New Yorker website after visiting the Business web page.

·        Bounce Rate – Percentage of visitors that leave the website from a particular page after a visit to a single page. It is based on visits that start with a particular page (i.e. Business section) and they leave the website completely.

·        Days & Visits to “Purchase” (it can be any outcome) – The days and visits that lead to “purchase” measure the true customer behavior on a website, or how long and often it takes a customer to make an outcome on an organization’s website. This measure has a lot of bearing in terms of perfecting the marketing messaging on the organization’s website.

·        Share of Search – Percentage of searches that leads to a website visit. This metrics also allows an organization to see specific keywords that lead to the website. For instance, for Southwest Airlines it may be cheap flights, free checked-in baggage, top ranked airlines, flights, etc.

As I noted, this list is not exhaustive … There are many other KPIs that an organization can use that will help better measure the objectives set for by the company.

And now as promised, I want to share few thoughts on the dark side of web analytics which is mostly based on a recent article I read. Are you ready?!? Well, here you go … As of November 2015, FireEye, a cyber security and malware protection organization, has identified about 14 websites that hosted a profiling script that was collecting and extensive information from the Internet. What does this mean? The backbone, is that threat actors with support from the Russian government, used web analytics to gather information about desired victims and computers owned by the victims in order to track, profile and infect the computers with specific malware. As per FireEye, the attackers are interested in gathering data from diplomats, executives, government and military personnel from US and Europe.

As a finish … Web Analytics is a about collecting data on visitors on an organization’s website and understanding what they are doing on the website in order to improve the design of the website which will lead to ACHIEVING the OBJECTIVES set by the organization!

Hope you enjoyed this week’s blog choice and of course blog content!!!

References

Kaushik, A. (2008, September 16). Six Web Metrics / Key Performance Indicators To Die For. Retrieved from Kaushik.net: http://www.kaushik.net/avinash/rules-choosing-web-analytics-key-performance-indicators/

Staff, F. (2015, November 16). Russia-led cyber attack campaign shows the dark side of web analytics. Retrieved from FirstPost: http://www.firstpost.com/business/russia-led-cyber-attack-campaign-shows-the-dark-side-of-web-analytics-2508552.html

 Ram, S. (2015). Introduction to Web Analytics. Module 5

Blog II: Dashboard Design

Dashboard Design

So … I had been thinking for few days now what will grab my reader’s attention, what should my BI blog topic be? My thoughts ranged from star schemas to balance scorecards to data warehousing, and finally to dashboards. Truly, each and every one of these topics would have been a hit as all are trending topics and are raising interest in the tech and business world. But, I had to decide on one … And can you guess what I picked? Well, maybe if you know my current project/task load, then you would know that I picked Dashboards as the winner. Just a short story of why I decided on dashboards. I was recently asked at my current job to come up with a dashboard design that displays departmental KPIs which need to have the capability to be sliced and diced and drilled down to be able to examine and see more detail when anomalies and inconsistencies arise.

I would bet that most of us have heard the word dashboard – for some of you is a car or airplane dashboard, and for others is actually an operational or analytical dashboard. But what is the definition of a dashboard? According to Stephen Few, a dashboards is “a visual display of the most important information needed to achieve one or more objectives, consolidated and arranged on a single screen so that information can be monitored at a glance.”

I can attest, from my own experiences and literature research that dashboards are gaining their popularity and every organization wants or has one. Dashboards are a great tool that gives executives and other interested parties great visibility and insight on what exactly is going on in their business. Dashboards can show trending of overall expense/revenue/profit of an organization compared to budget or benchmark, product lead times for manufacturing organization, patient wait times for hospitals compared to benchmark, etc. So, all of these portray quantitative information … So, do dashboards only contain quantitative measures?  Isn’t that boring? Well no, they do not only contain quantitative measures. Dashboards can have various widgets, such as spark lines, text labels, gauges, etc.

Now, let’s discuss characteristics of an effective dashboard and review some of the pitfalls that need to be avoided.

The big point here is that dashboards need to provide the big picture of the company’s performance. Prominence needs to be given to major metrics and attention should be easily drawn to measures that show poor performance in comparison to the targets.

Now, let’s mention few of the common pitfalls discussed by Stephen Few:
1. Exceeding the boundaries of a single screen – It is VERY IMPORTANT that all of the information should fit on a single screen
2. Displaying excessive detail or precision – Information should not be displayed in more detail and precision than necessary
3. Choosing inappropriate media to display – Think about what media is the best way to represent your performance, do not just go for the fancier widgets that do not easily portray the picture
4. Using poorly designed display media – Design the components so they communication information efficiently, effectively, and clearly, without distractions
5. Expressing measures indirectly – If you want to portray the variance between actual and budgeted revenue, then rather than showing the two attributes separately and having the viewer do the calculation, display the variance directly on the dashboard

I hope you, the reader, have gained a better understanding of dashboards, several important characteristics and pitfalls and how they might benefit you and your organization.

References

Few, S. (2005). Common Pitfalls in Dashboard Design.

Few, S. (2005). Dashboard Design: Beyond Meters, Gauges, and Traffic Lights. Business Intelligence Journal , 18-24.

Ram, S. (2015). Dashboard design and its use for analysis. Module 4