Data Integrity

The differences between the error generated by individual stores within a company are generally consistent, and may therefore the first step is to apply a company-wide correction factor. We refer to this as the Base Correction Factor (BCF).

Base Correction Factor (BCF)

The BCF is a reflection of the conditions typically associated with a specific type of retail and customer environment. The BCF is generally the average of observed error from a sampling of stores. Minor adjustments applied to individual stores can refine the BCF by developing a mathematical model called the Correction Factor Index.

Correction Factor Index (CFI)

The CFI refines the BCF to include characteristics of individual stores within the same company. The CFI will serve as a predictive tool when field observation is not feasible at all stores. The CFI would use a multiplier to variables that are found to affect the probability of error. Multiplied together, the CFI will increase or decrease the estimated error to better approximate each store's actual traffic.

Together, these correction factors will greatly increase the reliability that the traffic data, and standardize the measurement of sales opportunities for all locations. Counting Technologies would maintain these Correction Factors and properly apply them to the data collected by the traffic counters.

Phase	Description	Activity	Goal	Field Cost
1	15-Store Manual Audit	(a) Collect actual vs. electronic counts (b) Calculate an initial Base Correction Factor	Estimate BCF	~$35,000
2	Identify CFI-related data	(a) Identify sources of potential elements (b) Identify effort to automate data collection	Develop approach to calculate CFI	Included w/ DATAssurance
3	30-Store Validation Audit	(a) Collect actual vs. electronic counts (b) Compare predicted CFI to observed data	(a) Validate CFI (b) Refine BCF	~$55,000

Counting Technologies will use the field research to create the various Correction Factors required for more accurate data. The Factor Modeling will be store-specific, and can include company-wide, regional, store-design, store-type, or other attributes to determine the most appropriate factor for each store.

Phase 1: 15-Store Manual Count

Description

Using a similar but refined field methodology as Phase 1, manual counting would be conducted at fifteen (15) locations. Data would also be collected to support the creation of an algorithm to calculate a Correction Factor Index. The primary objectives are to:

Further quantify the approximate Base Correction Factor.
Collect actual data from each of the stores that are theorized to influence the CFI.
Calculate a proposed Correction Factor Index.

Costs

Typical Phase 3 studies run approximately $30,000 - $35,000. A significant cost savings is possible by using the client's staff after a short training by Counting Technologies field managers.

Phase 2: Identify CFI-related data

Description

Meet with IT, data management, and other staff to create a complete list of what data streams may be available to use for future construction of a Correction Factor Index. The primary objectives are to:

Identify a list of likely elements that can influence the accuracy of infrared counting.
Prepare a list of what data is collected by the client and the likely effort to retrieve those data for all stores.
Propose the algorithms for calculating a CFI.

Costs

The cost to perform the study is included with the QA Service Plan.

Phase 3: 30-Store Validation Audit

Description

Using a similar field methodology as Phase 3, manual counting would be conducted at thirty (30) locations. The primary objectives are to:

Validate and refine the accuracy of the Correction Factor Index.
Further quantify the approximate Base Correction Factor.
Optional: Initiate streaming data sources to automate the CFI calculation for all stores.

Costs

Typical Phase 3 studies run approximately $50,000 - $55,000. A significant cost savings is possible by using the client's staff after a short training by Counting Technologies field managers.