Beijing-based Lenovo is a global leader in PCs and computer hardware. The company serves customers across Asia, North America, Latin America, and EMEA with a thriving ecommerce hub — and that hub depends on up-to-the-moment analytics.
Lack of ability to drive the analytics is one of the major pain points of any data-driven organization. To make decisions for the future, we need to make these things available in real time.
Ranadip Dutta, Solutions Architect Manager for the ecommerce hub at Lenovo, and his team were initially tasked with creating a forecasting model based on data in 3 countries - since expanded to more than 180 countries. "There are 40+ projects that are coming on ecommerce right now,” says Dutta. “Digital marketing, telesales, inventory price optimization, product loyalty, attribution modeling, customer lifetime value review — there are so many.”
Dutta understood that each and every country, region, and subregion had unique needs and purchasing habits. In order to create demand in the market, Lenovo would need a way to read the minds of individual consumers making purchase choices in every area.
To do that, the company was going to need robust forecasting. And that was going to take a lot of data.
The data forecasting journey: predictive to prescriptive to cognitive
As early as 2016, the team started using its wealth of data to build prediction models. The team had no trouble producing a functional forecasting model, but they weren't giving executives the caliber of insight they needed.
The team moved to prescriptive analytics, giving them the ability to anticipate a solution to a problem. The model looks at nearly 60 request summaries and scrapes user reviews across multiple sites, including Amazon.com. They apply this authentic user feedback to make key decisions about specific product offerings on the ecommerce hub.
But prescriptive analytics was not the end of the journey. To perfect the ecommerce experience — and to make decisions for the future — Lenovo would need a different kind of analytics.
Cognitive analytics applies near-human intelligent decision making to data to adapt to consumer behavior as it happens. “Real data. That is where cognition is the rule,” explains Dutta. “The raw data comes in at near-real or real time. The neural networks and algorithms are adapting automatically to understand the market gaps and ultimately giving us solutions out of it.”
With cognitive analytics, Lenovo is able to detect fraudulent transactions in real time, which is good for Lenovo, for the bank, and for the legitimate account holder. And none of that is possible without Talend.
A perfect partner for data management
When it comes to cognitive analytics, access to data is key. Lenovo had been using Talend for years, so the ecommerce team already knew that they could rely on the platform to smoothly deliver the trusted data they needed. “Talend has a lot of generators on the data,” says Dutta. “I get to tap into the most granular data — and even manipulate and work on that data — which is very tough in other limited tools. There are other solutions that can do it, but Talend does it better.”
The team was also impressed with Talend’s ease of use. Because it does not require extra training or special skills, anyone can apply all the right business rules and start working directly with the data.
“Every single thing is based on Talend,” Dutta continues. “The data that we get, the major heavy lifting, the business rules, the guidance — all of it is built on Talend.”
Delivering ecommerce with a difference
In any ecommerce business, the objective is not just to close a deal, but to get repeat sales. With their new cognitive analytics engine powering the ecommerce hub, Lenovo was able to increase sales and drive higher lifetime value from their customers.
The moment they enter the shopping cart, consumers see recommendations based on their current preferences and the choices they are actively making. That kind of relevant, customized experience increases both individual sales and likelihood to return.
For Lenovo, this means that they were able to hit their ecommerce global needs rate target, and are now heading for a bigger target rate in the future. “With the growth we’re seeing, we’re on target to hit $4 billion this quarter. By 2026, our target is $8 billion across all the sectors — but particularly the ecommerce market. And that's a significant growth in sales revenue,” says Dutta.
Did you know that a surveyed 36% of executives don't base the majority of their decisions on data? Talend can help.
In this article:
Customer & Partner Spotlights