mindis/adm-adstock

Adstock Model with Time Series Analysis in Salesforce – Sales attributing to different promotional campaigns – TV, Radios

---

Case:

Marketa analytics wants to optimize their promotions campaign. They have weekly historical data and have built a model that shows sales a function of temperature, tv ad spending, radio spending and adstock effects.

Adstock Analysis

• CodeLab Link for Assignment: https://codelabs-preview.appspot.com/?file_id=1PSsKcbe53Y1PKxCQOKHN5wUrM666ZdMuP4gBZ7oyng0#7

Objective

• To analyze the data and understand the influence of ad spending on sales including the adstock effects.

Tools

1. Einstetin Analytics - Salesforce
2. Google Colaboratory
3. Pandas for Data Analysis

Future Enhancements

1. Validating Adstock Model to understand how Adstock Value is generated.
2. Manipulating the Adstock Model 1 and 2 to understand the decay effects of TV and Radio Spend.

How to work with csv files

Steps to be followed to Analyze data using Einsten Analytics

1. Data Wrangling CSV - Change the Sr.no format to "Text" and create the schema (final.jason) for the CSV file.

2. Upload the file with Schema mentioned in Step 1, setting "Week" as Dimension and other attributes as "Measures"

3. Create the App and Lenses with associated logic as:

   • Sales Evaluation • Effect of TV Spend On Sales • Quantify_TVspend_v1 • AdStock Effects on TV Spending • AdStock Effects on Radio Spending • Effect of Radio Spend On Sales • Quantify_RadioVspend_v1

4. Derive the insights from the above lenses.

---