Analyzing return on investment of advertising campaigns using cross-correlation of multiple data sources

A computer-implemented method for processing data, the method comprising:

• electronically receiving data in an advertising measurement system from one or more data sources including at least a program delivery source associated with a household of a consumer, the received data including a first identifier associated with the household assigned by the data source, the advertising measurement system including an electronic computer processor and at least one electronic data storage medium;
• and
• electronically using a thesaurus for:
• producing data without personally identifiable information, and
• indexing the produced data by a second identifier, wherein the thesaurus relates each first identifier of the household to the second identifier.

A computer-implemented method for storing data in a data access tier of an advertising measurement system, the method comprising:

• distributing the data access tier of the advertising measurement system into multiple shards, wherein each shard is configured to maintain the same data structure while storing different data, wherein the advertising measurement system includes at least one electronic computer processor and at least one electronic data storage medium;
• confining all information pertaining to a specific household of a consumer to a single shard;
• executing multiple instances of an aggregate query on each of the multiple shards;
• and,
• merging results obtained from the multiple queries into a single query result.

A system for storing data, the system comprising:

• an advertising measurement system including a data access tier distributed into multiple shards, wherein each shard is configured to maintain the same data structure while storing different data, wherein the advertising measurement system includes at least one electronic computer processor and at least one electronic data storage medium;
• each shard being configured to confine therein all information pertaining to a specific household of a consumer to a single shard;
• and,
• the data access tier being configured for:
• executing multiple instances of an aggregate query on each of the multiple shards;
• and,
• merging results obtained from the multiple instances of the query into a single query result.