For any organization, understanding the customer’s needs is the most important thing in order to provide better services and build healthy relationship with the customer. Customer analytics is a method of combining customer information (personalization data, customer interests, customer browsing behavioral data, and purchase behavioral data) to present a consolidated view of the customer and their activities. This involves segmenting customers into multiple buckets, understanding their purchase behavior, browsing the behavior, and estimating their next move. This helps business users (CRM managers and the marketing team) to make the right decisions in approaching and retaining customers.
Below are some of the typical things business users would like know about customers in order to make effective decisions and building the right marketing promotions.
Questions Asked By The Business
1) Who are the frequent customers?
2) Who are the high spenders?
3) Who are the holiday shoppers?
4) Who are at-risk customers?
5) Who are attrited customers?
6) Who are the first time buyers?
7) What is the primary genre of the customer?
8) What products might the customer buy next time?
9) What products might the customer be interested in?
Most of the questions listed above can’t be answered by just bringing customer personalization and purchase information together. It requires lot of effort to construct sophisticated processes to pre-calculate answers to questions like the ones posted above and feed them to the data warehouse. The marketing team usually needs this information to be refreshed on a daily basis. So it is very important to construct sophisticated data models and highly tuned ETL processes in order to accommodate the refresh process along with other nightly maintenance, batch processes, and other ETL processes.
It also requires a considerable amount of meetings with the business teams to understand the business logic of these metrics and implement. Most of these metrics/answers need to be rebuilt frequently by looking back at least a few years of historical data. For example, the ‘Next Product To Buy’ calculation needs to be based on the last product the customer purchased, so this needs to be rebuilt for the customer each time he has made a purchase and it needs to look back at least a couple of years’ sales to avoid recommending the product that customer has already bought in the past.
Customer Analytics Challenges
It is also important to denormalize customer information as much as possible and avoid redundancy. The marketing tools (like IBM Unica Enterprise Campaign) which will have a catalog constructed over marketing data warehouse, can’t perform well and will not be able to leverage its strength to the fullest if the underlying data model is not denormalized correctly. Also it causes lot of confusion to campaign builders if information is scattered across multiple tables.
It is also important to have pre-cleansed and consolidated customer data flowing into the Data Warehouse in order to provide golden record of customer with the latest personalization information. This can be achieved either having in-house master data management system or taking services from a third party vendor.
PR3 Systems consultants are experts in setting up MDM and helping to build effective Data Warehouse for Customer Analytics and Marketing. We have thorough experience in dealing with Marketing Data models that can easily collaborate customer analytics information with campaign response data. We are experts in setting up easy to navigate and fine-tuned catalogs for IBM Unica Enterprise Campaign tool, building Data Warehouses for multi branded marketing systems, building ‘Next product To Buy’ ETL logic, building campaign reports, as well as compiling campaign response reports, campaign revenue reports and campaign analytics reports (Test vs. Control group metrics comparison).
Please contact PR3Systems at email@example.com for further inquiries.