Increased Campaign Performance for Notino
Customer segmentation and market basket analysis for better targeting of the sampling campaigns and increased cross-sells for a major, international perfume and beauty online retailer
How to improve targeting for sampling campaigns
Notino has been offering samples of perfumes and other beauty products to their customers as a gift with their purchases. The goals of such sampling campaigns are:
Upsell the customer e.g., from a mass-market brand to a luxury brand, or
Cross-sell the customer so that they buy goods from a different category.
Notino had a purchase history of 1,011 sampling campaigns and wanted to increase the conversion rates of these campaigns and to support cross-sells. Therefore, Notino engaged with Simplity to examine customers and products offering, and to solve these challenges:
How to leverage their transaction history to target the right customer segment. This would enable the personalized approach expected by customers nowadays.
How to discover relations among the product items and categories.
What would be the most appropriate sample to add to the shopping cart of a customer for the best chance of the purchase of the full product, so upselling and cross-selling.
Project in numbers
samples of beauty products
conversion rate cross category
Improve cross-category expansion in the e-shop
Notino was looking for ways to improve the campaigns by analyzing the purchase behavior of the customers, their affinity to buy, and what products are purchased together. This would enable Notino to offer the most suitable sample to a customer who would, in the end, buy the full product or a different product of the same brand. This improved targeting of sampling campaigns would improve customer experience and increase customer satisfaction plus cross-category expansion for the retailer and hence, also increase revenues.
As a first step, Simplity conducted data preparation of the transaction data from Notino to evaluate the current state of it. There were almost 12 million customers, 6,896 products, and 1,011 past sampling campaigns analyzed.
We decided to take a two-step approach to the data, firstly conducting the customer segmentation to see WHO to target and then, secondly, market basket analysis to understand WHAT to offer.
For customer segmentation, we performed RFM analysis to determine quantitatively which customers are the best ones by examining how recently a customer has purchased (recency), how often they purchase (frequency), and how much the customer spends (monetary). Based on this we got 32 RFM groups of customers. For better usability of this segmentation, we assigned scores to RFM and based on this scoring got 4 tiers of customers for campaigns and communications based on the RFM segments in which they appear. These distinctive groups were:
Champions – the most valuable customers buying often, recently and with the biggest spend.
Rising stars – new customers that have a potential to become Champions soon.
Churned good customers – those who used to spend well but have not submitted any order recently. (This would be a good segment to target with an email or postal campaign with discount coupons rather than samples.)
The rest – this was the biggest group of customers but whose sales volume was not significant.
Then we conducted market basket analysis over the four customer tiers, getting the separate list of association rules for each of the tier. We got more precise results for a respective customer tier without skewing the results with the irrelevant purchase history of other customers. The goal of the market basket analysis is to get insights into what products customers purchase together. Due to the nature of the industry Notino operates in, where shopping carts usually have a low number of items with each purchase and the rather low frequency of purchases (compared to food for instance), we decided to conduct the MBA over the purchase history of a customer i.e., in this case, the “basket” is the whole history of the purchases of one customer.
The results of market basket analysis were showing that the association rules are more common within the brand i.e., from one Product Type to another within the same brand e.g., Body care by Hugo Boss with EDT Hugo Boss, EDT Armani with EDP Armani. Also, Champions and Rising stars had more and stronger rules, so we recommended Notino to focus their sampling campaigns on these two specific tiers.
As a last step, Notino used the outcomes of the market basket analysis for pilot campaigns to verify the results. Notino tried both to offer samples within the brand as suggested by our results and test it against the hypothesis of affinity between different brands in different categories e.g., hair-care routine by Kérastase offered to those who had in the cart skincare from Lancôme, YSL, Estee Lauder. After 120 days of running this pilot, it clearly confirmed the brand loyalty and showed that the cross-brand offering was less successful in terms of conversions. The ultimate results of these pilot campaigns will be evaluated after 12 months again due to lower frequency of purchases for some categories (such as perfumes). The case study will then be updated accordingly.
The market basket analysis carried out for the Notino project helped Notino to better understand the relations among product categories and what products are being purchased together. This will facilitate the improved targeting of samples of the products, so what should be offered when certain product is added to the shopping cart. This will lead to improved conversion rates of their sampling campaigns.
The project delivery spanned just 40 person-days and enabled Notino to benefit from data science insights and test hypothesis for affinity targeting, with the aim of category expansion. Yet the aspiration to offer cross-brand samples didn’t prove to be valid and, as such, could be eliminated from future targeting campaigns.
Customers nowadays expect personalized offers tailored to their interests. E-shops that are able to keep up with this customer requirement will win the battle of loyal customer and improve their performance.
How can we help you?
Are you curious about how Simplity can grow your e-shop performance using data science insights?GET IN TOUCH