Evaluation of User-intent on the web

Posted: January 2006 C: Customer Intelligence Posted by: K. Panayotakis

The ability to identify Web users whose need may lead to a purchase of a product or service, and ‘redirect’ them to a product or service, is central in the search engine marketing world.

CPC (cost per click) on Google search engine result pages (SERPs) is higher than CPC on any other web page, because Google is currently the major search starting point, thus at this point there is a much higher degree of intent.

Yahoo is working on a new generation of task-centric search that will focus on the analysis of user-intent. (source)

In an effort to differentiate its search engine marketing product with a ‘behavioral targeting component’, MSN adCenter is currently testing commercial intent scoring algorithms. A test on alternative keywords, using the tool titled ‘Detecting Online Commercial Intention MS tool’, gives the following results: datawarehouse solutions - 0.2489 Probability for Commercial Query datawarehouse products - 0.39644 Probability for Commercial Query, while a test on a Web / URL www.pleroforea.com gives the following result: NonCommercial (Page) Probabilities for each commercial intent type: NonCommercial Prob.: 0.70 Commercial-Informational Prob.: 0.22

However, one would wonder how is this ‘commercial intent score’ derived and how valid it is. Why ‘datawarehouse solutions’ has lower score than ‘datawarehouse products’. Certain keywords lead directly to products marketed or sales channels, while others relate to needs. There is commercial value or revenue potential not only in product sales but also in services: a search on a marketing technique may lead to the purchase of relevant consulting services or an ebook. Moreover there is referral value in many web pages, in which contextually relevant topics are analysed, and authorities or recommended solutions are presented and linked.

After all, sales methodologies & sales experience tell us that a Customer is looking to satisfy a need not to purchase something. Therefore everything starts from the need.

It is valuable to evaluate commercial intent based on click-streams which resulted to an e-commerce order. This would require log analysis in a large scale. Based on that, user intent can also be targeted in an efficient way on PPC campaigns, with the appropriate keywords and negative keywords.

Search engines and PPC engines have to get much smarter to evaluate efficiently the User's intent and deliver content that can provide value to the User and the Professional.