According to the Online Measurement and Strategy Report from E-consultancy and Lynchpin just 18% of surveyed firms use web analytic data to make business decisions. This is both surprising and bewildering because the real beauty of online marketing is our ability to track, measure, analyze, and test the marketing initiatives but many businesses don't seem to 'get it'.

To get a clear picture of the current and future state of your online business, you need to transform your raw web data into meaningful insight that can help you understand and predict the behaviour of your customers.

In this newsletter, we would be discussing analytic data overload and how to synthesize data to deduce actionable insight.  

Drowning in Data

Conventional wisdom suggests that the more information one has concerning a particular issue or event, the more likely he/she is to accurately predict the possible outcome. If you come to think of it, this applies to diverse areas such as sports, weather and off course business.

A web analytic tool will churn out more data, which can be used to extract information, than you can possibly imagine or handle. However, there is a growing belief amongst marketers that data is now a part of the problem and not a part of the solution. This belief is fuelled by the fact that businesses are either not making use of this incredible data or are overwhelmed by the sheer amount of it. In either case, the real question remains mainly unaddressed - "How much of this data are we using to make timely and actionable decisions to maximize top line revenues?"

Too much Data vs. Actionable Insight  

One single look at the instrumentation panel of a cockpit can create a feeling of information overload, especially for non pilots. On the other hand, pilots can quickly identify the primary gauges and navigation instruments that are immediately important to fly the plane. More importantly, they know exactly where to look to verify the intended effect of the control input made by them. So if they increase power to the engines they would look at altimeters and airspeed etc. to check the intended outcome. Actionable web analytics is a lot similar to this.

Most business owners experience the same sense of information overload when looking at web analytics as a non pilot would when looking at the cockpit panel.

In absence of formal training or experience, deciphering which bits of data/information are relevant and appropriate to the objective at hand and which bits are extrinsic to immediate concerns becomes a daunting challenge for business owners.

Turning data into action - Data synthesis

Though most businesses are armed with web analytic tools, its utility is mostly confined to just numbers, reports and pretty charts. Disconnect between website statistics and its transformation into actionable insights to improve marketing efforts, conversion rate, and customer retention, still exits.  Therefore, data synthesis is all but important.

Data synthesis can be simply performed by answering some very basic questions without burdening yourself with heaps of raw data. Listed below are few questions that that will help you cut through this data maze.

• Where are they entering?
  It is fairly easy to view the top landing pages of your website regardless of which analytic tool you use. You may be surprised to see the number of visitors who aren't entering the website via home page.
  

  This crucial bit of information does not only show the key entry points but also highlights non performing pages which need to be tested for performance enhancement. This data is also critical to conversion path analysis and understanding why visitors take the action they do.

• How are they entering?
  Any analytic package will allow you to break down visitors by source i.e. organic search traffic, paid search traffic, direct traffic etc. And for organic and paid search traffic you will be able to find keywords that brought visitors, which in turn will help you sense and understand your visitors': 

  • Expectations:
    Every keyword portrays a searcher's intent and expectation, and this insight can be fruitfully used to align your services to meet their expectations. The more you meet up with their expectations, the more customers you will have.
    
    
  • Likely goals:
    The same keyword can be used by different people having different goals and it is probable that you are ignoring a few of them. A person searching for "discount book stores" may be looking for a particular book on sale while someone else may be looking for children books.
    
    Knowing the keywords and using this data to map as many likely goals as possible for each keyword will help you better understand your visitors, and thereby align your offering(s) to cater to a wider audience.
    
    
  • Brand exposure: People who land on your website after having searched for your brand name or a specific product line of yours are in an advanced buying stage and more likely to convert during that session.
    
    Keywords can tell you which stage of the buying process the visitor is in and thereby enable you to present the most appropriate marketing message.
    
    And
    
    
  • How well is your landing page synchronized with visitors' expectations and goals:
    If you are able to gauge visitors' expectations and goals based on the keywords they use, it would not be difficult to tell if the entry points/landing pages are in sync with what they are looking for.
    Weeding out disconnects and ironing out any mismatch becomes easy to achieve because of actionable information deduced from web analytic data.
• Where are they going (navigation)?
  You now know where your visitors are entering and what expectations/goals they have when they land on your website. The next step is to look at what pages they navigate to upon arrival in tandem with the landing page and ask some fact finding questions.
  • What are the most popular 'next' pages and are they the logical 'next' pages in the conversion path?
    
    
  • Are these pages linked by strong call to action elements or are accessible via top or side navigation links?
    
    
  • Do these pages answer a particular question that the visitor may have in mind?
    
    
  • Is this action congruent to their keyword expectations and likely goals?
    
  Answers to these questions will depict a pattern and you can use this information to define a conversion path irrespective of the landing page. Also, watch out for navigated pages that have high exit rate.
  
  
• Where and why are they leaving?
  Bounce rate and exit rate are the most common data sets to look at to decipher where and why the visitors are leaving. Let's look at them, one at a time.
  

  Unless visitors can perform a desired action/convert on the same page (with a few exceptions), a high bounce rate for the page is bad. This could either mean that you are attracting the wrong visitors or your landing page is terrible at assuring them that they have come to the right place.

  On the other hand exit rate could mean a few different things and would have to be investigated further.  

  • Planned and Unplanned exits:
    Some exits are meant to be - you expect a visitor to leave your site after having performed the intended action i.e. from a 'thank you' page. But you obviously don't want visitors to leave before they have converted.
    

    If you identify a high exit page in the conversion path, look at major entry paths to that page and ascertain if there is a mismatch between the two pages. Ask questions such as, "Was the call the action link misleading?" Or "Is the page not up to the job?"

  • Time on page: A person leaving the page in few seconds is not the same as a person leaving the page after few minutes. An exit within few seconds is most likely triggered by wrong product or service, while an exit after few minutes is mainly due to lack of information and/or confidence.

If you regularly subject your web analytic raw data to the above mentioned actionable analysis steps, you will find several anomalies in the form of:

• Mismatch between yours' and your visitors' expectations.
• Mismatch between what was supposed to happen and what actually happened.

And this brings us to the most important aspect of actionable analysis - testing. The mismatches that exist will give you a clear idea of elements (parts of website) that need to be tested. Form a hypothesis and get on with testing. It will not only help you to eliminate trouble elements but also assist you in enhancing user experience.

And to end our longish deliberation, we would like to share a quote from non other than the web analytic guru Jim Sterne:

"Web analytics reports are just lumber. It takes an architect, a designer, a builder and a lot of other skills to turn it into a house."

Do let us know if you found this information useful. We would look forward to hear from you.