If you think of the “brick” retail environment there are a bunch of methodologies to understand consumer behavior- these include observation, focus groups,surveys and the whole bunch of traditional market research tools.
Now consider a similar customer who visits the on-line store of the retailer. You can notice exactly what the customer looked at , where she left a category and started browsing another and exactly where the customer left if she did not buy. Even if there was a purchase you can track a great deal of things later like sending an online survey or just observing post-purchase behavior data. Some of these analytics platforms are free like Google Analytics.
Suppose you find that lots of customers leave after they see that garish red scarf you have on offer for $19.99 you can replace the scarf photo or better still try alternate experiments. You can actually put in different colors of scarves (maybe blue works better) or see if there is a difference if you change prices around.
Imagine trying to do the above at a real clothing store. It’s really hard to know that someone looks at the red scarf and seems to move on. If you are the owner you will observe but can hardly expect the floor employee to notice,report and act on the observation (i.e. try putting on the blue scarf). In fact, buyer behavior data is both hard and expensive to collect. Gaining insights and acting on them is even harder . Hey where did I put the label gun? might just make you forget about trying out a different color scarf on the retail floor.
Organizations have tons or rather terabytes of data but are not able to develop actionable insights as this McKinsey article suggests. And it is not math skills that are required because all the analytics software do the math for you. What is required is marketing and business knowledge and deep domain knowledge and a curiosity and passion for the category (you must care a little bit at least, about the question of what kind of scarves people choose and why?).
The amazing thing is that never before has consumer behavior been easier,cheaper or quicker to observe,quantify and analyze for actionable insights.You are able to actually observe what people do and not what they say they do in response to 7 point “like-dislike” scales.
Incidentally at the B2B level, going by the very cumbersome Supplier portals at big company websites, supplier web behavior analytics have barely started . Exciting times are ahead for market insights through web analytics.
Also see: From Strategy to Execution via Website: the Power of Web Analytics.