Business data need love also

Every Business is different, and the data quality is one of the most unforeseen challenge you have to face with your automated marketing projet. As such you need to know about your structure data.

Most companies ERP are complex, trying to suit multiple purpose, as such the product data, is often modelled or reproduces multiple time in different configuration.

When you are using a Cloud solution (AWS / BQ / Azure), an API connection to your different product data sources will be a key to your success. There’s no real limit in number of fields allowed however you need to apply some logic into it.

1. What is Structured data ?

Structured data is a standardized format for providing information about a page and classifying the page content; for example, on a recipe page, what are the ingredients, the cooking time and temperature, the calories, and so on.

looking at Google logic regarding structured data markup for a page, you can understand the core requierement for a product level attribute.

https://developers.google.com/search/docs/appearance/structured-data/intro-structured-data#:~:text=Structured%20data%20is%20a%20standardized,the%20calories%2C%20and%20so%20on.

Case Exemple

Let’s take a moment and go by an exemple. the Shoe exemple will do well.

Here are the values Found in your different sources for a Women Red Shoes Size 36

Source 1 (SAP / Salesforce / other) : sent by a product provider

• Size = ’36eu / 38eu / 39eu ‘
• Product = ‘shoes W Hhl’
• Color : ‘red or black’

Source 2 : Magento

• Size eu = ’36 / 38 / 39″
• Product = ‘High Heels Woman Shoes!”
• Color : ‘Red / Black’

In this current state we cannot use the data without massive cleaning.

Using Database to structure the data.

Using database language we can call functions to segment and order each attribute separately. This can be used to create our Campaign naming, ad group naming, ad headlines, keyword combinations and so on.

Tip : Choose a unique attribute as an anchor for the rest of the data : Ie. product sku

Product name	type	brand	Color 1	Color 2	Size 1	Size 2	Size 3
High heels	women shoe	x	red	Black	36	38	39

The level of segmentatin will define the accuracy futur template.

The solution is to use 1 row to match one 1 product id with all the attributes linked to it.

Product id	Product name	type	brand	Color 1	Color 2	Size 1	Size 2	Size 3
1	High heels	Women shoes	x	red	Black	36

SA360 : inventory management logic

From then, use the multiplier for ‘Product ID’ to request the template system for the creation of values.

ie. Create keyword based on your attributes for each product id.

Now it’s simple SQL math, no punctuation, no separators :

Keyword template 1 : {type} {product name} {Color 1}
Keyword template 1 : {type} {product name} {Color 2}
Keyword template 1 : {type} {product name} {Size 1}

Securing the data quality in advance will give you an edge on the templating later on. Depending the products you want to advertise, sure your data fondation is strong first.