MarketWide vs. Conventional SQL-Based
Campaign Management Systems
MarketWide is the only open-architecture product that lets you perform Set Operations against SQL databases. For instance, in the example below, MarketWide can tell you things like:
- Which customer has placed six orders lifetime
- Product 547 was purchased three times (on three separate orders)
- Total dollars spent on 547 is $797.30
- Total dollars lifetime is $1,685.30
- Product 547 was purchased on the most recent order
The MarketWide Approach
Here is a sample MarketWide segment, selecting customers who:
- Have placed at least six orders
- Have ordered product "547" at least three times, for a total of at least $750
- Paid at least $1,500 lifetime
- Purchased product "547" in their most recent order
The underlying SQL (not just for this segment, but for the entire campaign) generated by MarketWide looks like this:
This is an exceedingly simple query for SQL databases to perform—but the query must be performed for every segment. MarketWide runs one query, regardless of the number of segments in your campaign.
The SQL approach (SQL Server syntax)
With traditional SQL-based products, multiple complex queries are submitted, one for each segment of the campaign. The SQL necessary to perform the same MarketWide selection as above looks like this:
This is a much more complex query that causes the server to work considerably harder and use more temporary space. Further, in a small 10-segment campaign, it is necessary to run ten similar queries in parallel on the server, producing ten result sets. Finally, these result sets must be deduplicated against one another in order to ensure that each customer is chosen only once.
Overview | General Information | Reporting | Campaign Management
| 
