How Vinted nearly doubled conversion to buyer - by using the data they already had
Head of Communications
Posted Nov 11, 2014
We’ve been talking about pulling real-time marketing data into your app for the last week. Early adopters of this tech were fashion startup star Vinted, who help more than one and a half million users every month trade and swap fashion and clothing on their mobile devices. The app was launched in June 2013, and today commands a global rank position in the top 6 % with 75 % highly positive reviews.
"We plugged our source data from existing campaigns right into the app, seamlessly carrying on a conversation from the ad campaign down to the first purchase. Within the first, initial tests, the conversion to buyer jumped up by 77 %. Let's say that's enough to make sure we'll be building this out for heavy production use."
Tautvydas Gylys, Product Manager, Growth at Vinted
I pinged Tautvydas Gylys, Product Manager of Experiments, over at Vinted earlier in September. “We’re trying out some stuff to dynamically match content in the onboarding process to the original message that we showed in the adverts,” he told me then, “so when a user hits adverts for a certain brand or a certain item in our catalog, we can then tailor their on-board views based on those preferences.”
This is a core use case of the real-time marketing data. Like many other mobile startups, Vinted’s monitors the success of its user acquisition campaigns very carefully, tracking dozens of sources down to each creative and ad wording. This data is filtered through to their adjust dashboards, where performance and user engagement is reported per each ad source. With a simple hook, that data can be brought back into the app post-crunching.
Custom tracking data sent to the SDK
Vinted installed a delegate callback function in their SDK that, upon every tracked event, received and handled up-to-date tracking parameters from the adjust servers. By configuring their tracker structure, the Vinted app received a number of parameters they controlled at click-time - including data on the product and the creative that had been displayed to the user when they were acquired.
In their app, the method can then check for instructions or parameters in the tracking data to use for the customized onboarding. Vinted used this feature to bring up a certain product category or brand in a search in the app - meaning they could implement a single method and re-use it for every campaign with similar tracking parameters.
Testing with category-specific messaging
With the method and the data conduits set up, Vinted started a small-scale test to collect some data on the effectiveness. In this case, their social campaigns advertised a category-specific messaging with content available out of their app. Using the adjust APIs, Vinted could then pull the advert names and generate the onboarding process within a second of the first app open.
After a few days of testing, the results were astonishing. Many ecommerce outfits settle on conversion to buyer as their core UA metric - an approach we often recommend, in order to keep the noise down in reporting and the signal high. In this case, even with Vinted’s small-scale tests, conversion to buyer jumped by 77 % above the campaign average at the time.
Next week we’ll talk more about using in-app data for improving referrals and couponing, and we’ll have more data and cases lined up for your enjoyment.