Join our newsletter

#noSpamWePromise
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
cross-icon
Subscribe

Run your data operations on a single, unified platform.

  • Easy setup, no data storage required
  • Free forever for core features
  • Simple expansion with additional credits
cross-icon
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Download the file

#getsmarter
Oops! Something went wrong while submitting the form.

3 practical examples of data-driven eCommerce

Customer Stories: How 3 different businesses are making data-driven decisions in eCommerce.

Use Cases
May 12, 2022
3 practical examples of data-driven eCommerce
Customer Stories: How 3 different businesses are making data-driven decisions in eCommerce.

Ignorance is bliss. But not so much if it’s costing you money, right? If you are running an eCommerce business and are still not leveraging data to make data-informed decisions, you are missing out on crucial insights that can maximize your sales and minimize costs.

One of the common misconceptions is that becoming a more data-driven company is complicated, difficult, and will cost a lot of money. This can be true, but there is also a better way.

Keboola, a data stack as a service platform, allows you to utilize the power of data hassle-free, and to put our money where our mouth is we invited three of our eCommerce customers to share their journey, and how they use data to make their sales & marketing decisions. 

Without further ado, let’s see how it's done and what you can do too.

1. From passive spreadsheets to active data-management

Ascend Climbing is a relatively new and small company that offers climbing facilities with yoga and fitness departments and a variety of classes. During the first few years of their operations, they were riding the wave of rock climbing being a trendy sport, but unfortunately, the pandemic took that momentum and it was time for an eCommerce pivot. They created a Shopify store to sell gift cards and merchandise as well as utilizing Union to host virtual fitness and yoga classes. 

Their data journey began in spreadsheets and comfortably lived there for quite some time.

‘’The pre-pandemic momentum made it really easy to rely on our gut instead of data when it came to marketing and sales activities.’’

The vast majority of their analytics occurred within their own siloed platforms: Website, email, and Facebook Ads were not connected with each other and almost all of their BI lived in the Google Sheets. At the end of the day, they found that they were spending valuable time manually gathering data and creating dashboards that were never used. 

Things got even more complicated when they introduced Shopify and Union, and they found themselves communicating information back and forth regarding members' information, inventory, etc. 

That’s when they realized it was high time for them to integrate all the sources, and since their data team consisted of one analyst, they knew immediately they will not be able to build something for themselves. Besides that, their budget was also minimal, so they also wanted to utilize free tools whenever possible. 

When they discovered Keboola, the first thing they loved was that it was a cost-effective tool to integrate different data sources and overall had exactly what they were looking for. After implementing, they found themselves not just passively ingesting data but starting to actively manage it. 

‘’We've been able to automate a large number of our manual data tasks, which for me frees up hours to do other important things.’’

They were also able to implement email validation, which helped them increase their delivery rates and email reputation.

In just 5 weeks, Keboola helped them set foundations so they can now make big strides with their data. 

Their next goal is to build a deeper customer profile that will drive their marketing efforts. They also want to connect online and in-store behaviour, in order to get a complete understanding of their customer journey. 

2. Controlling the weather (effects) with a prediction model 

Sportisimo is a CEE-based sports retailer with 190 stores in 5 different countries that also operates a very popular eCommerce store selling sports shoes, clothes, equipment, and accessories. 

Because of the nature of their industry, their sales are highly dependent on different seasons and weather conditions. That is why it was important for them to be able to predict sales peaks, their shape, if they are expected to be high or low and how long they will last.

One way to make these predictions is with a crystal ball, but a more realistic one is to create a prediction model. They decided for the latter and partnered up with Revolt BI to prepare a model that would answer the question of how many pieces of each product are going to be sold at each store and each shop in the upcoming two weeks every day.

The biggest challenge of this project was the scope because their products ranged from all year-long such as T-shirts and socks to highly seasonal equipment such as skis and bikes, which are very dependent on the weather. 

The second challenge were the many locations across different countries and weather climates. 

Essentially, what Sportisimo needed were 12 million predictions recalculated every day and information to be provided to anyone who needs it - from sales managers to the warehouse operators.

How did they make this possible?

They started with top-quality data about the most recent sales to capture data signals and also included data from older sales to get the long-term and seasonal information. 

Obviously, they also needed a weather forecast for which they used Keboola’s integrated dark sky API and very robust product categorization. All this was stored in Keboola: both the informal data and the SQL or Python transformation. 

The result of the implementation was an increase in seasonal sales such as skis and bobsleds as they were able to predict the winter season one week in advance. 

”The model performance was improved by 10 to 40%. With respect to our benchmark, obviously, the model accuracy is much better one to three days ahead than 15 days, simply due to the better precision of the model forecast.”

Besides the prediction model, they also use data for:

  • Stock position optimization
  • Long-term sales predictions to optimize their purchase process in order to optimize stock in the long term.
  • Analyzing the risk factors. Some products are in extremely high demand at the start of the season and they have to prepare enough stock to fulfill the demand. 

3. Data has to empower, not overwhelm

Dáme jídlo, the biggest food delivery company in the Czech Republic has been using Keboola since it was founded in 2021. 

They cover more than 170 cities, almost 60 % of the population of the Czech Republic, and they deliver over 2.500.000 meals per month. 

“All these transactions and data are stored in Keboola. And basically, thanks to Keboola, we are able to provide insights to different departments, we have good data for reporting and all the data is easily accessible.”

One way they are making data-driven decisions is by utilizing clever maps - visualizing the data within a map. 

“Because it's one thing to have data in a table or in a chart, but the owners of the restaurants, when they can see these maps, and see the demand for  pizza, or burgers, or for Asian food, they're much better able to set up delivery zones of the food, optimize their business and so on.” 

Having Keboola and clever maps is one of their competitive advantages because they are the only ones providing this kind of insight.

For Dáme jídlo it’s not only crucial to have the data, but also to translate it to the business, so it can be easily accessible to everyone. 

They also noticed that the sales & marketing team can get overwhelmed by the amounts of reports they have. For example, one account manager can be responsible for more than 200 restaurants, so on daily basis, it’s practically impossible to check every restaurant and see what happened, if it’s doing well, etc.

That’s why they are now working with Keboola’s consultants on anomaly detection, which in practice means that account managers receive a list of restaurants where something happened the previous day, so they can react in a timely matter and manage the business much better. 

Create a proof-of-concept for your business (risk and money free)

The next goals for all these companies are to automate even more processes and to provide the best possible customer experience. From the begging of their customer journey with more relevant targeting with dynamic ads to personalized communication, and down to better customer care.

And definitely, Keboola will continue to be a key partner on their data journey. 

This article is an overview of a live webinar: Data-driven eCommerce - success stories from our customers. Listen to it here

Check out Keboola’s free tier to connect your data sources and start making data-driven decisions today.

Recomended Articles