The vicious cycle of eCommerce stores and how to break it?
A company is measured by its profits, not revenue
Everyone is talking about personalization in eCommerce and about giving the customer what they are looking for. The goal is clear - optimize the store’s conversion rate and boost sales. It might increase the revenues, but is that the bottom line of a business?
Personalization alone is not enough.
The vicious cycle
eCommerce stores usually have a classic Pareto of 20% of the products (in a good situation, for some it's less) are generating 80% of the revenue. By promoting only the best sellers, and by giving the customers ONLY what is most likely to sell, stores create a vicious cycle for themselves. They are mostly selling these 20% products that are selling easily, and then also order more inventory of those products. The other 80% of products are being neglected and create non-moving inventory. Then, in fashion brands, for example, by the end of the season, they create a promotion and sell these products with a huge discount and minimum/no margin. The products that did not sell there, are moved to the outlet, donated, or thrown away.
These products’ inventory costs a lot! The cost could end up being up to 50% over the actual cost of the product. There are many hidden costs, such as: storage, handling, physical deterioration, and lost margins.
These costs end up eating away around 20% of the company’s profits!
How can a store change that, optimize its merchandising and increase profits?
By enhancing its merchandising abilities and integrating it with personalization.
Break the vicious cycle and optimize profits
So, how can a store avoid this vicious cycle and balance between boosting revenue, reducing non-moving inventory and maximizing margins?
After working with 100s of eCommerce companies from all shapes and sizes, I realized that the best way to go is integrating personalization with smart merchandising. Taking all the required parameters into account and by that, create a balance between providing a great, personalized, shopping experience for the customers and meeting the business needs of the company.
One of the most important aspects of a store’s merchandising activity is collection/category sorting.
Around 75% of clicks on products are done on the 1st results page. Therefore, for a store with hundreds or thousands of products, the collection/category sorting is crucial.
To avoid that vicious cycle, a store’s manager needs to form a sorting strategy for each collection/category that will take into consideration multiple parameters of the products:
- Conversion rate
- Days to finish inventory
- Number of variants with inventory
- Margin – the real one after all the discounts and promotions
- Reviews – product’s rating & number of reviews
- Days in store
- Market demand
- And many more…
The manager should consider the relevant parameters (personalization is one of them and not the only one) for each collection/category based on the business needs.
Of course, not every parameter has the same importance for each collection/category. In your SALE collection/category the amount of inventory and days to finish inventory would have more importance than in your BEST SELLERS collection/category. Therefore, you need to choose the relevant parameters and decide on the importance and influence it should have on the sorting strategy.
Does it work?
After 4 weeks of using smart collections of Kimonix, Swarovski IL has achieved an 8% increase in revenue, an 18% increase in conversion rate, a 29% reduction of non-moving inventory, and a 72% increase in the daily sales rate of top products by inventory volume.
Kimonix allows you to set a sorting strategy using AI personalization & advanced analytics. It enables taking multiple parameters into account, setting the influence of each on the products' order and by that, creating the required balance between boosting sales and reducing non-moving inventory.
Check out how Kimonix allows to create and sort AI personalized collections using all these parameters and setting the influence of each one on the sorting algorithm.