Set a Baseline Before You Start
Send the primary field team out to note store layouts, signage and other key features prior to the commencement of a data collection project. This will act as a reference against which subsequent dates’ figures can be compared in order to identify anomalies.
Record Context That Explains Anomalies
Anomalies in your data such as unexpected dips or spikes can often be explained by a simple field visit to the store in question. For example, roadworks reducing footfall, a store having moved to a different location, a temporary promotional activity blocking off a display – all of these and more can be logged against the date and store of the observation and then accounted for in your analysis.
Standardise Your Recording Forms
Field data must be comparable across sites and visits. Thus, a data analysis company designing a field research program needs to treat the design of the observation form for the field operative to complete the structured visit data for visit after visit as seriously as it treats the analysis of the collected data.
Cross-Reference Against Automated Feeds
By comparing field findings with third party data or even automated data analysis (such as from retail partners or from automated shopping basket tracking systems), you can soon spot any large gaps in data. Data Analysis Company is covered in more detail at https://shepper.com/.
Time Visits to Validate Model Outputs
Scheduled field visits are typically required at very specific decision points where models have produced outputs which require action. These visits validate the model’s ‘picture’ of reality before any decisions are made or actions taken as a result of the model’s output.
It’s also worth noting that field programs collecting personal information or commercially sensitive information will need to adhere to the guidelines set out by UK law in respect of data collection and handling.
Plan your visits and they will sharpen your data analysis.