Depop Is Using Machine Learning to Boost Its Mobile Retail Platform
Artificial intelligence (AI) and machine learning are providing a great advantage to digital brands.
Once the domain of science fiction writers and filmmakers, artificial intelligence is now finding its way into real-life industry. While we are yet to see fully sentient thinking robots such as those from our favourite stories, we now have machines which can learn as they work, refining and streamlining operations to make ecommerce and other industries more productive and cost-effective than ever before.
AI also generates enormous volumes of data which can be analysed to gain a greater understanding of customers and their wants, needs, and experiences.
Now, Depop is using machine learning to make its mobile platform even more powerful.
Founded just a few years ago, Depop was born from a desire to create an entirely mobile retail platform - no desktop presence at all.
"I came up with the idea of integrating social networks, geolocation and push notifications in a single platform - all features that today are taken for granted," said founder and CEO at Depop, Simon Beckerman. "In the beginning, it worked like eBay. My intuition was to introduce a chat paired with a smart notifications system to replace the emails: this approach lowered the waiting time down to zero, bringing 'instantaneousness' to the purchasing process. We were the first ones to build such a marketplace with a specific focus on young users, and we were relying on Instagram, which boomed shortly after our launch, as our marketing platform."
Since then, the platform has grown in popularity. Depop is now considering adding a consignment option for users. In the past, an item would be listed and the seller would wait to receive offers from interested parties. However, the new service would have the item given off consignment to a third-party site which processes images and guarantees its authenticity.
Taking cues from other providers of digital services, Depop is now looking towards artificial intelligence -- machine learning, more specifically -- to improve the way it delivers personalised recommendations.
"The second next big thing is the impact of Machine Learning - we can finally train machines to have taste and recognize what we like or not," said Beckerman. "But we're only at the beginning of this revolution. There's still a lot of work needed in order to create integrated systems with AI and humans working in deep synergy. One of the best examples of this is Apple Music, where a human editorial team preselects songs for users and AI generates a playlist of songs they might like based on that first curated handpick."
The important thing to note is Depop doesn't believe in 100% automation, but rather thinks AI can be used to augment and strengthen the work done by human employees. It sees a desirable balance as something like 60% automation with the remaining 40% being carried out by flesh-and-blood humans.
Entirely automated recommendations can often miss the mark, as the AI draws relationships between products which may not be appropriate. However, if a human begins the process, the AI can then build recommendations by following their example.
"But there's more," said Beckerman. "The more information available online, the more difficult it will be for the user to find what they're looking for. We want to simplify the discovery process. And we will do this through curation, which we will deliver both in traditional and unexpected ways. This might depend on my age and on the romantic vision I have on this, or maybe it's society itself that's taken a step back and is looking for the human touch after the tech frenzy of the last decade."
Other innovations stemming from AI technology include voice search and chatbots. There's nothing specific in the pipeline right now, but the company has stated that, should it ever expand into Asia, it would certainly be eying up the WeChat platform to help penetrate that market.
AI and machine learning are helping Depop become an even more powerful mobile retailer than ever before. With possible plans including chatbots, voice search, and maybe an expansion into the lucrative Asian market, the future looks bright for this innovative brand.
Download the agenda today for more information and insights.