These folks are linking data from various systems. These folks don't care whether they have 100% coverage or not, they simply link what they have. If they can't link visitation information to every purchaser in the database, they don't. They work with what they have.
These folks maintain a customer database with fields like these:
- Months Since Last Purchase, Life-To-Date Purchases, Life-To-Date Items, Life-To-Date Demand/Sales, Life-To-Date Returns, Twelve-Month Purchases / Items / Demand / Sales / Returns
- Total Company
- By Physical Channel (Mail, Telephone, Internet, Retail).
- By Advertising Channel (E-Mail, Catalog, Google Paid Search, Yahoo! Paid Search, MSN Paid Search, Portal Advertising, Affiliate Marketing, Shopping Comparison Marketing, By Most Popular Blogs).
- By Merchandise Division.
- By Store Or Region
- Months Since Last E-Mail Click-Through, Total E-Mail Click-Throughs, Total E-Mail Twelve-Month Click-Throughs.
- Website Visitation Data
- Months Since Last Visit
- Number Of Life-To-Date Visits, Number Of Twelve-Month Visits.
- Average Time Spent On Site Per Visit.
- Number Of Pages Visited.
- Months Since Last Visit By Key Landing Page, Merchandise Division, Key Links.
- Months Since Last Shopping Cart Abandonment, Number Of Carts Abandoned.
- Outbound Marketing History
- Catalogs Mailed, E-Mails Mailed, Telemarketing Calls, Direct Mail Pieces.
- Self-Service Marketing Initiated By Customer
- RSS Subscriber And Items Subscribed To.
- Paid Search And Natural Search Activities.
- Visits From Key Blogs, Facebook, MySpace, Social Media.
I've worked on a few projects with data of this nature. The findings are fascinating.
- Websites are a separate channel from e-commerce. There is an information element to the website that supports other channels. There is a social element to a website that allows customers to interact with each other. There is an entertainment element to a website that causes customers to visit again. And there is an e-commerce component to the site. Customers self-segment themselves into one of these micro-channels.
- E-commerce customers have unique dynamics.
- Those from paid search have a very different relationship with your brand ... unless they use another form of advertising (e-mail, catalogs) in combination with paid search.
- Once the catalog customer tries e-commerce, they become unlikely to go back to ordering over the phone.
- Once the e-commerce customer purchases in a store, they become less likely to go back to ordering via the internet. This does not mean they become unlikely to use the internet --- in fact, they become information/entertainment/social users.
- Trigger-based e-mail marketing is ideally suited for customers who last visited the website within ten days.
- You learn that you have a couple hundred important micro-channels, combinations of advertising, self-service activity, and purchase channel. You learn not to "force" customers into micro-channels --- instead, you let customers self-select themselves into a micro-channel, and work with the customer based on her natural, subsequent behavior.
- Customers evolve across micro-channels, often migrating to a "most popular" micro-channel over time.
- E-Mail click-through customer who visits website frequently and buys merchandise in-store.
- Paid search customer who does not subsequently repurchase.
- Rural customer who buys online after receiving a catalog.
- Customer who visits site after reading a blog post about merchandise, visits frequently, does not buy merchandise.
- Customer visits site after seeing portal ad, signs up for e-mail marketing.
- Customer buys in-store only.
- Customer only shops via landing pages (i.e. needs a merchandising assortment presented to her).
- Customer only shops via site search (i.e. a self-service customer who picks and chooses what she wants).
The folks I've seen do this style of marketing are producing very interesting results.