As an auto dealer, you are always looking for any information that may have a direct impact on the success of your store. In the first Driven Data Science blog, we looked at the impact of weather and day of the week has on the service department. Now we are going to take those same inputs and apply them to how it affects new dealership leads. The weather data is provided by Wunder Ground and the lead records are extracted from the CRM of two Driven Data clients. We limited the time range of this research to the years of 2015 and 2016. We used more than 400,000 leads records from the Chicago, IL and Charlotte, NC areas to build our analysis. Since dealer stores of different scales have different amounts of new leads per day, we have normalized the lead data by following two steps:
- Remove the outliers of raw data and exclude data of Sunday when dealerships are usually closed. List out the number of total leads, the number of Internet leads and the number of showroom leads of each day in 2015 and 2016.
- Calculate z-scores for those stores that have normally distributed lead data (almost all of them)
After normalization, the value of lead data actually becomes a statistic which describes the level of new leads on a given day. 0 = an average level, >0 = a high level and <0 = a low level.
Now the normalization process is complete, we have some interesting findings that we elicited from this research: [Read more…]