Thursday, March 18, 2010

Monetization Models in Web Analytics

One way or the other, most web sites in place have the same objective: making money, generating revenue. The road leading up to this objective can be different for different sites, for e.g. some sites may generate revenue by getting leads, e-commerce sites get revenue by direct online sales and directing traffic to offline stores, content sites generate revenue through ads by getting more and more traffic to their sites.

Now web analytics helps most online businesses (content sites, ecommerce sites, etc.) by helping them understand what drives traffic to the sites, which combination of content is most efficient in retaining traffic, which assortment of products sell more online, etc. and then making the appropriate changes. But to understand the true value derived out of making any changes to a website based on analytical recommendation is most effectively realized only when you start working on monetization models for your site and believe me it can be done for almost any genre of web site out there whether it be a lead generating site, an ecommerce site, a content site, you name it.

I recently came across this article in the book Actionable Web Analytics: Using Data to Make Smart Business Decision which explained in detail the concepts around the monetization modeling with very concrete examples for different kind of sites. It’s really simple to understand and has very great effects (as far as convincing business to make changes based on recommendations), but the catch is it can be a bit tricky to implement. Instead of trying to guess how valuable a change to the site will be, monetization models show you exactly what will happen that too in monetary terms, so that you won’t have hard time convincing the business.

Web-monetization models assign dollar values to desired site behaviors, most often those directly involved in driving your overall business. The behaviors they quantify are different for every type of company and range from easy-to-measure e-commerce sites to more difficult ones like consumer-product-branding sites.

To understand how monetization models work, let’s see how a web site can affect a company’s profitability. It can primarily drive revenue or reduce cost. In addition, two kinds of visitor behaviors can affect your business: direct and indirect. 

Direct Behaviors:

Lead Generation
Customer Service

Indirect Behaviors:

Offline Sale

Now let’s take the case of a typical ecommerce site that sells products online to generate revenue and explore how the monetization model can help drive changes and understand their effect easily.

It is generally seen that a majority of site visitors who visit a product detail page are more likely to make a purchase online. So in order to assign a monetary value to this page all you have to do (very crude methodology) is find out what percentage of visitor(say x%) who visit the product detail page make an online purchase (say during a one month period), find the average order value for your site ($Y) which contributes to a profit margin $Z. Now if your site has a monthly visitor count of N then we can assign p =$(N*x*Z/N) as the average profit generated by each visit to the product detail page. This is off course just a very simple example and there are a lot of factors that need to be considered to find out the final value of p.

So moving ahead with our example, say you are able to increase x (percentage of visitor who visit the product detail page) to x+ ∆ through some change to the product detail page (better site navigation, better internal search, there can be a myriad of ways), then you can effectively suggest that due to this increment ∆ your profitability will increase by $(N*∆*Z) which can be a large increase if N is large. So what you have done is related an increase in the conversion to a particular page to direct increase in the profit for your site. This can be very helpful in deciding which changes to make and which not based on results of A/B or multivariate testing.

P.S : The monetization methodology suggested in this post is meant as an example to drive home the point. Monetizing modeling can be complex depending on the kind of site and has to take into account a lot of factors to give you an accurate value which can help in decision making

No comments:

Post a Comment