Ideally, for the measurement we combine the more traditional click-stream analysis approach (number of pages viewed, time spent on site, etc) not just with information gained from our backend systems (number of sales or leads, service calls, etc) but also with our understanding of the customer experience (surveys, user testing etc). More on this in the paper Winners and losers in a troubled economy : how to engage customers online to gain competitive advantage.

Avinash Kaushik, author of Web Analytics: an hour a day has termed this approach Trinity.

The combination of behavioral understanding, business outcomes and customer experience is the basis of the segmentation criteria of Netmining. Kaushik describes the benefit of this approach as :”… the Trinity mindset drives the fundamental understanding fo the customer experience so that you can influence the optimal customer behaviour that will lead to win-win outcomes for our company and your customers.” (2)

We see that companies are investing tons of money in advertising that drives traffic to the site (banner, adwords, …) and we see also that the same companies are investing in CRM to follow up leads. But were is the bridge, to motivate people to leave voluntarly that world of the anonymous click on the site and to enter the CRM environment where we need to know his contact details.

Ok, you have an online form, a car-configurator where you can leave contact details, you have toll free numbers, call me back, chat, skype, … that help you bridging the gap. It’s a first step, but we have to admit that conversion ratios from online visitor to visitor who interact are still very very low.

So the right interation online is just not good enough. You need more! So let’s take a pause and let’s think once out of the box. Let’s consider the automotive website as your largest showroom. To a certain extent this is correct. Most of the websites of car brands and dealers are attracting more visitors daily than the the physical showrooms.

So imagine now you have a sales person in your physical showroom and he only starts talking when you tap on his shoulder. Sounds silly isn’t it, but that is exactly what happens online. A webvisitor has to click on a chat link (if you find it) or a form to get in touch with the sales environment of the automotive dealer.

So let’s continue with our website as a metaphor for the physical showroom. My questions are now: What is a good sales person ? For me, a good sales person is certainly not a passive person. No, it’s someone who acts at the right time, with the right interaction and the right message.

And that is now what Netmining is doing for +60% of the automotive brands in Europe. We deliver not only the right online conversation solutions but more important we do this at the right time and to the right people. At that’s the point where we go beyond the solutions of traditional chat. We add knowledge to the interactions. Every visitor is segmented in real time. Based on click behavior we calculate the propensity to buy off-line. If this is high enough we present an interaction.

The results are phenomenal compared to the traditional enviroments. More leads, higher quality, better conversions and more sales.

Co-browsing is a solution that enables a web agent to reactively or proactively monitor the online behavior of the individual visitor and additionally take over control to assist the visitor (to complete the buying process).

In my opinion co-browsing can only be effective, in other words a positive experience to the website visitor and a positive return on investment for the site owner, in the case that it is applied selectively. The first reason is that not every visitor will appreciate this kind of service, as Marcel Feenstra remarks correctly. Compare this to a real situation: when you enter a shop, you often want to look around first before a salesman taps you on the shoulder with the question “Can I help you?.”

A positive online experience by means of co-browsing will strongly depend on the ability to detect visitors who will appreciate this kind of interference as well as the appreciated timing of this interference. An appreciation that is strongly correlated to the individual visitor’s interest level related to a specific product or service. The second reason is, as Martin Hellendoorn brings up this argument, that the cost of sales will increase due to the need of incremental personnel. So from a sales point of view, you only would like to apply co-browsing on those visitors where the conversion probability times the contribution margin exceeds the incremental costs of personnel. In case you are not able to assess the quality of each individual visitor in terms of his interest level, the chances will be minimal that co-browsing will bring you additional revenue. For this reason Frank Leegstra is right that co-browsing will not be a value adding solution for every web shop. Hence, selectively co-browsing, in other words only applying co-browsing at visitors with a product interest that is above average, is a required condition in order to realize a proper return on investment with this marketing instrument.

The key to successful co-browsing is the ability to identify visitors that will appreciate the interference on one side and have a high probability on sales conversion on the other side. The technique for this ability already exists and is applied internationally by Netmining. Combining this technique with co-browsing technology seems the logical next step.

Even though Netmining’s business is not limited to analytics, it made me think… We have written earlier about the rather unique position of Netmining between website optimization on the one hand and CRM on the other hand. Nowadays the main focus is on web analytics and to understanding customer behavior. In the interview Jim Sterne mentions that the next big battle will be about integrating all this data and push the data to knowledge.

At Netmining we want to push things even a bit further. Whereas the step from data to knowledge might seem a big step at first sight, it is clear when one looks at a business, knowledge is not the end point. ROI - or sales- is what makes a company healthy.

The Netmining algorithm will help you to interprete the real time data and put this into knowledge about the customer - eg. predicting his behavior- and on top of this, trigger interactions with certain customers. This is why we are convinced that realtime, onsite behavioral targeting as Netmining proposes, is a next level of web analytics. The data about customer behavior is not seen as an endpoint but it is used in a process to make critical decisions, it is used in the same way as your most senior sales guy in showroom uses his instinct whether a prospect is worth being approached for a sales pitch or not.

“We have over 300 different clients and probably 600 different sites, and serve about 9 billion impressions per month. We’re able to use BT and retargeting, so when someone registers for a particular form or takes an offer or clicks on a banner, or doesn’t click on a banner - we record all of that and then, based on the vertical they are clicking on, we will use that to retarget additional offers. [...]“

Behavioral Targeting is nowadays too often described as a fast growing segment of the online advertising business. I guess the wave of consolidation in the online advertising industry we have seen over the last months has to do with it. The acquisition of Tacoda by AOL in particular. But pushing content and ads based on the surfing behavior of a visitor on a set of sites is only one part of what behavioral targeting can do.

So I was happy to finally read an interview where BT is described as a source for lead generation. However, with Netmining, we go much further than identifying interest. Behavioral Targeting can be the starting point to bridge the consideration gap as Forrester once named the blur moment between identified online desire for a product of service and actual offline sales of that particular product or service.

Our expertise on BT shows us that onsite behavioral targeting over multiple sessions can deliver greater ROI than behavioral targeting over different sites. In fact on-site behavioral targeting starts where the traditional BT stops. That means that a site owner pushes personalized and relevant content or interactions, based on the visitor’s web activity at his website. When someone shows a clear interest in a product or service, offline sales can than be activated through different ways: a contact form, a chat, or a phone call. All of this happens at one single website. That creates a seamless combination of a visitor’s online interest and your offline sales. This results in more qualified sales leads and contacts from your website resulting in a better return on investment. And after all, that is why we are in business, aren’t we?

Other panelists are :

• Rob Gratchner, Director of Privacy, aQuantive, Inc. a Microsoft Corporation Subsidiary
• Mark Westlake, Executive Vice President, Sales and Content, Howstuffworks
• Ralph Terkowitz, General Partner of ABS Capital Partners and Chairman of the Board of TRUSTe
• Todd Terisi, Vice President, Yahoo! Inc, Dave Morgan, TACODA/AOL
• Carlos Jensen, Assistant Professor, School of Electrical Engineering & Computer Science, Oregon State University.

Netmining will be represented by Chanterria McGilbra, VP Business and Channel Development the Americas.