Mobile devices drive one-third of global eCommerce sales and that share is growing. As consumers purchase more and more from the smartphones in their pockets, eCommerce and retail companies have to match their businesses with the digital habits of today’s consumers – otherwise, they risk losing their time, attention, and dollars. As a result, many top retailers rate improving mobile experience as a top business priority in 2017.
That includes Vente-Privee, the French eCommerce site that reinvented the flash sale in Europe. But the product team there knew that “improving mobile experience” would quickly devolve into iterating based on intuition if they didn’t first develop a strong sense of customer behavior.
In order to own their category in a competitive industry like eCommerce, Vente-Privee needed end-to-end analytics to understand how users behave on web and mobile platforms and what it takes to actually take them to the final sale. With product analytics, Vente-Privee built a data-driven company at scale for six million loyal and engaged shoppers and purchase funnels that drove three billion in annual revenue.
Driving engagement from browsing to buying
Vente-Privee had mechanisms for retention built in from the get-go. In France, clearance sales only happen twice a year. So, by giving its members daily access to discounted designer brands in every product category, Vente-Privee offers shoppers an entirely new, economical way to buy. Still, the product team at Vente-Privee doesn’t take its members for granted – not a single one of the 25 million of them.
From product’s perspective, continually driving sales depends on more than the size of Vente-Privee’s existing customer base or the mere fact of its web and mobile presence. According to Vente-Privee’s Chief Marketing Officer, Catherine Spindler, “Online shopping is now the new normal for today’s smartphone user. At Vente-Privee, it’s mission-critical for us to understand the entire journey of a customer, from browsing to buying.”
At Vente-Privee, that understanding surfaced after a dive into product data. As Catherine reports, “Today, product analytics continues to provide us with the insights we need to optimize purchase funnels and build out a fashion ecosystem that our customers love and designer brands want to sell to.”
Acquiring (and retaining) six million loyal users
Before Wanda Saint-Paul joined Vente-Privee as its Web Analytics Manager, teams often had to make decisions without the analytics to back them up. This proved particularly problematic for Vente-Privee’s product teams in France, which are organized around different features.
“For example, one team focuses on user-engagement: how to turn site visitors into members into buyers,” she explains. Wanda’s team serves each feature group by implementing analytics tools, monitoring data accuracy, maintaining product performance – with particular focus on newly released beta functionalities – and producing reports and creating dashboards.
The first analytics solution that Vente-Privee implemented didn’t allow feature teams to easily measure how their changes impacted user experience or behavior. “We couldn’t view data in real-time, merge navigation data with member data, measure the downstream effects of notifications on user retention, A/B test features and UX, or model multi-touch attribution.” The tool didn’t last.
Then, last year, the features teams started looking for a self-service analytics solution so that teams could finally feel confident in their decisions and know the impact they were making.
“We chose Mixpanel because it allowed us to track our fifteen KPIs across different platforms like web, Android, and iOS. In addition, we were able to run A/B tests to tweak the product and optimize user behavior flows that created repeat buyers.
“On top of that, Mixpanel gives every account a unique ID, which gives us the ability to track users across devices.” This was particularly important to Vente-Privee, because its customers engage on both web and mobile – often within the same day and sometimes even the same hour. “Having People properties tied to user behavior really differentiated Mixpanel from other solutions.”
Especially for mobile shoppers, distinct IDs were critical in sending targeted and timely Messages to customers to encourage a specific action. With Mixpanel Messages, the team had the advantage of sending push notifications to certain users in a particular geolocation when there was a flash sale. Then, with Core Reports, they were able to monitor the downstream effects of creating an ultimate feedback loop with timely nudges.
“Frequent communication with our customers is really important given our business model. We don’t have physical stores, so we have to regularly encourage customers to come back on the site and app through email and push notifications.”
Because Vente-Privee offers new sales and brands every day, customers tend to come to the site more frequently. “In fact, we have customers who come to our mobile app every morning to browse.” At first, this pattern concerned the product team. “But then we realized that while people might go on Amazon or e-Bay with the intention of buying something specific in that moment – a black skirt for example – our customers go on our app to check regularly if there’s a new sale or brand available.”
After using People properties in Mixpanel to map customer behavior across devices, Wanda and her team learned that despite frequent visits, these mobile customers eventually end up buying on the site. “Since sales open at 7 a.m., people would go on the mobile app at in the morning – maybe on their commute to work – to see what new items were available. Then, they would log back on later in the day on the desktop and purchase the items they had viewed on mobile.”
Two distinct advantages give Vente-Privee the power to see and understand customer journeys: “First, everything that our customers do on the site is logged on Mixpanel. There aren’t any actions the user can take without first signing into their account. Second, five months into our implementation of Mixpanel, we are able to understand cross-platform behavior – drilling down to the user level by device, country, and segment to improve re-engagement across our five different customer profiles. So, for us, it’s not just about developing a mobile app for its own sake. The mobile app (and the analytics) is key to building and understanding our loyal, long-term customers who buy from us, again and again.”
Optimizing purchase funnels to generate more than three billion dollars a year
Before Mixpanel, team members at Vente-Privee couldn’t easily monitor the activity of customers. This gap was unsustainable, given Vente-Privee’s primary, company-wide KPI: conversion rate. “Our collective goal has always been to bring on more customers who buy on their first visit and then come back and buy more.” In order to drive customer conversion, team members needed to be able to track customer purchasing patterns and user flows without having to ask for analytics resources.
With Mixpanel funnels, Vente-Privee can see where their customers dropped off in an event sequence. “We did customer analysis that revealed that many people add products to their cart in mobile and then buy them five minutes later on desktop. Well, it got to the point where we saw so many customers following that purchasing pattern that we actually decided to dedicate resources to improving cart user experience on the app. There would have been no way for us to find an insight like that before.”
Especially for an eCommerce site like Vente-Privee, it’s important to understand all the different paths that lead to a purchase. With this goal in mind, Vente-Privee’s developers created a custom-built, multi-touch linear attribution model with JQL, making it possible to analyze large volumes of traffic every day.
On the less technical side, marketing teams measured the effectiveness of their email marketing and advertising campaigns. Through Mixpanel, they’re able to assess the ROI of their marketing budget and feel confident in how they allocated resources in order to influence customer conversion.
From understanding the purchase funnel to improving it, Vente-Privee’s team also used data to improve customer service post-sale. Quality customer care is a key part of Vente-Privee’s brand, and now they can standardize and improve each experience with data.
“Before, we couldn’t match customer care data with navigation data. Now, we can find the specific users and track the actions they took on the site before and after contacting customer care,” Wanda explains.
Using that data, Vente-Privee’s customer care team initiated a preventative and proactive approach to service. “We look at the data in Mixpanel to see if our response was effective – that is, if the customer came back. Then, we determine and solve the problem that caused the customer to contact customer care in the first place, so that future users that take the same path become customers, too.
Importance of data quality
According to Wanda, teams like customer care wouldn’t be acting on the insights they discover if Vente-Privee hadn’t invested in analytics that instills data trust. For eCommerce companies who want every team to act more on metrics and less on instinct, Wanda has this advice, “Focus on data quality. In order to get people to make decisions informed by data, teams have to trust in its reliability. Otherwise, they’ll ignore it.
Building trust in the data takes time. “We spent five months implementing Mixpanel. As a result, every bit of information we gather from it feels like a victory because everyone is confident that it’s accurate.”
A developer at Vente-Privee also built an internal collector to cross-check Mixpanel data against. “Again, we want to feel certain in our data and we’re not afraid to invest.”
Wanda has also found that proving the reliability of Vente-Privee’s data source has convinced more teams that it’s worth their while to engage with Mixpanel on their own. “It’s important for everyone to go into Mixpanel and generate their own reports. People tend to ignore reports that get delivered to their inbox every morning. At Vente-Privee, we limit automated reporting to the executive team. No one else receives daily metrics from our data scientists.” And that’s intentional. “The most effective way to make data a priority, company-wide, is to encourage everyone to check the data themselves.”
Because they’re not tasked with producing analyses for the business teams, Vente-Privee’s web analytics team has taken to building custom reports with JQL to help developers monitor and test the functionality of new features and changes to the product. “Mixpanel is not just used to track KPIs and other metrics on the front-end, but also as an operational tool that helps Vente-Privee improve and iterate on the product on the back-end.”
Investing in data quality is foundational to the success of any data-driven team – startup to the enterprise. “Don’t skip out on ensuring data integrity,” she warns. “Companies, even small ones, need to make the time (and find the right people) to own the implementation of analytics tools.” Implementing spends operational resources, but the alternative is not a risk one should ever take.
“A high-quality data source prevents problems down the road and allows teams to start relying on the data (and learning from its insights) the minute the implementation is complete.”
Vente-Privee’s data-driven horizon
When Wanda started working in analytics, she lacked the kind of high-quality data sources and analytics tools that make a data-driven culture company possible. Instead, she worked with complicated, clunky tools that boxed most teams out. For that reason, she thinks the make-or-break capability of any product analytics tool is its ability to make data accessible across teams. With these tools available, Wanda is confident that building data-driven products and cultures will eventually become table stakes.
But Vente-Privee is ahead of the curve when it comes to optimizing its platforms and channels to improve conversion and revenue goals. As they continue to grow, teams at Vente-Privee will look to product analytics not only to understand how shoppers’ behaviors can help them improve products and expand their customer base, but also to find anomalies in the data and identify new user trends.
Catherine, Vente-Privee’s CMO, agrees. “With Mixpanel, we can learn and iterate our platforms to improve the shopping experience for our customers. However, what’s going to keep us ahead of the competition is crunching large volumes of data to uncover the trends we couldn’t see by ourselves. That’s why Mixpanel’s predictive analytics is key to constantly improving shopper experience.”
“Predictive analytics is very important for large eCommerce companies who have to anticipate customer behavior within a base of hundreds of millions,” says Wanda. “Still, it’s never just about making analytics the best it can be for data analysts, but also the best they can be for the entire company.”