Online and offline become 1: a new era has begun (part 1)

I recently came across two interesting articles that are closely related to our Customer Online Targeting (COT) tool. Both are from Information Management. The first one, “Online Analytics in Action” by Roman Lenzen, deals with web data and how to manage this huge amount of information. The second one, “Bridging the Gap Between Online and Database Marketing” by David M. Raab focus on linking online with offline data at the customer/visitor level.

I recently came across two interesting articles that are closely related to our Customer Online Targeting (COT) tool. Both are from Information Management. The first one, “Online Analytics in Action” by Roman Lenzen, deals with web data and how to manage this huge amount of information. The second one, “Bridging the Gap Between Online and Database Marketing” by David M. Raab focus on linking online with offline data at the customer/visitor level.

Online Analytics in Action

The first interesting point, that we have also noticed at FinScore, is that online data are usually not integrated with other offline data. To enable true analytics, Lenzen defines three steps:

“First, the online data must be integrated. Second, it must be analyzed; and finally, the insights must be made actionable within all channels.”

Lenzen also lists four initial requirements. Whenever possible, I give examples using COT:

1. “Determine the online data that is available and develop links“. In COT, the link can be made at the customer level (when he is identified) or at the cookie level (when non identified or when the visitor is a prospect). This link is extremely important since this allows to have an Extended Customer Profile (as named in COT). This profile contains both offline (CRM) and online (behavioral) data about each user.

2. Preprocess online data. Although it may be huge, online data are still data. It is very important to aggregate the online raw data as soon as possible to reduce the disc space needed. With typical web logs, data aggregation can reduce data amount by a factor of ten. An important step is to find which data to aggregate, into which granularity level and on which time basis.

3. “Do not discount anonymous user data“. On most website, there are more anonymous visitors then identified ones. Working at the cookie level (as long as visitors don’t delete them) allows a targeting even for prospects. In COT, identified customer habits can be used to predict anonymous visitor behaviors (ads they are more likely to click, interests, etc.).

4. “Determine an effective and efficient way to capture and load the online data within one integrated environment“. COT produces recommendations (scores) that are delivered to the client ad server. In addition, extended customer profile (obtained in part due to web log aggregation) can be loaded back to the CRM or any data warehouse. This information can then be used by the company for 1-to-1 marketing or further data mining. The loop is thus closed.

The rest of the article deals with possible solutions with SAS, SPSS, as well as open source tools such as R, mySQL, etc. to put these steps into practice. In the next post, I will write about the second article by Raab. If you’re interested, you can read the full article: Online Analytics in Action.