How We Learned To Predict Increasing Demand — What the First Corona Symptoms on idealo Have Taught Us
idealo’s data lake has enabled us to combine information on the performance side with inventory data. A study of the first effects of the Coronavirus pandemic on idealo shows us what we can learn from it and how we can predict increasing customer’s demand with inventory data.
In its function as price comparison, idealo is serving both eCommerce buyers and merchants. On the buyer’s side we enable our customers to find the best offer for the product that they are looking for by making the offers from our merchants comparable. This way our customer can make an informed purchase decision. Historically idealo has been growing by building a larger and larger user community. This was possible because we were listening to our customers. We tracked their interactions on idealo and could improve the customer’s journey as well as our service and built trust.
On the seller’s side we enable our merchants to list their products on idealo and they get access to millions of customers. The eCommerce market is very dynamic, and merchants need to manage their inventory every day. These changes in prices and availability are valuable information for us and they are transmitted to us via a feed. What if we could understand what the merchants are saying to us?
For a long time, it was not technically possible to put the historic inventory information into a data warehouse structure. A comparison between the size of traffic data and inventory data illustrates this fact. In May we had five million sessions on idealo per day. On the other hand, we have 500 million inventory updates each day. On the bright side cloud computing and idealo’s new Data Lake allows us to embrace this challenge. Now we can listen to what our merchants are telling us through their inventory data.
In February 2020 idealo felt the first symptoms of Corona. Within 5 days we witnessed that a significant share of traffic was going to products that had a reduced stock of offers. A quick analysis showed that this development could be attributed to an increasing demand in ‘House & Garden’ and ‘Health & Pharmacy’. Within these two root categories we could identify two categories that caused the shortage of offers: ‘Disinfection’ and ‘Occupational Safety’ where the protective masks can be found. Under normal circumstances both categories do not have a significant impact, but within this time window the traffic share of both categories peaked at 10 % of the entire idealo traffic.
During this time, this sudden customer interests seemed to have popped out of nowhere. The urgent need for disinfectants and protective masks caught us by surprise.
If we would have listened to our merchants, we could have known.
The amount of offers in both categories had been decreasing before the exploding interest on the customer’s side. The supply of disinfectants had been steadily decreasing since the beginning of February. There are trends like this in our inventory that can anticipate customer behavior. This gives us a time advantage that we can use to be better prepared for situations like that.
This learning was possible because idealo was wise enough to make the inventory data of our merchants accessible in the Data Lake. Thanks to the help of many idealos, especially in Inventory Analytics and Data Platform, we are empowered to listen to both sides of eCommerce.
What would be possible if we connected what the customers and merchants are saying? This is not out of scope anymore. Our Data Lake is a reality now. It is up to us.
Do you want to be part of the idealo success story? Have a look at our vacancies.
