Cosmetics firm Avon Products has implemented a suite of business intelligence (BI) solutions to better understand the way its business works, and work out how best to tune its sales offers to customer demand.
But these customers are not the end users of its products, with whom Avon has no direct contact. Instead these customers are Avon representatives. Avon supplies its representatives, who order online or over the phone, with its products, and the reps act as franchises and sell to the end users personally.
The firm manufactures, boxes and ships its products to its 6.5 million representatives every two weeks.
Danny Siegel, director, enterprise information delivery at Avon, told delegates at the Gartner Business Intelligence Summit that this equates to far more than merely shipping 6.5 million items each time.
"Whereas people buy one or two things at a time from Amazon, our typical rep buys 50 things in a visit. So that's a massive amount of units every two weeks," said Siegel.
Avon's business, which operates and sells product in 140 countries across the world, is complicated by the localisation of its systems – its data has not historically been directly comparable between countries.
This is unhelpful when trying to take an overall view of the business and its performance, said Siegel.
Typically, large organisations will have one enterprise resource planning (ERP) system running its entire value chain. Siegel explained that this would be unworkable for Avon.
"We have several ERP systems because of the differences within our business. One ERP vendor may be great for one aspect of the business, but not for another, so we end up with system sprawl.
"And our systems and even our master data have local flavours, so the way I define my business in Argentina is different from the way I define it in Chile, for example."
This system, which grew organically, was far from ideal for understanding the business. Avon needed to pull its data together into logical warehouses, where it could be rationally sorted and analysed.
This would enable it to compare its performance across countries and regions.
It did that by dividing it up by business function, such as sales or marketing, and by identifying linkage points, where data from one country or system could logically be directly compared and sorted with data from another.
"We built our first set of functional warehouses, for example a sales warehouse, supply chain warehouse and so on. We identified business owners for the data, then separated it by subject area.
By eliminating high entry costs for big data analysis, you can convert more raw data into valuable business insight.
A discussion of the "risk perception gap", its implications and how it can be closed