With four million customers and operations in six countries, Birchbox has been at the forefront of innovation in the beauty sector.
Though it might be less well-known, Turkey’s leading ecommerce site Trendyol has been similarly innovative in understanding the fashion needs of consumers.
What do these two companies have in common? A shrewd grasp of data – and two female CEO’s at the helm.
I recently heard both Katie Beauchamp of Birchbox and Demet Mutlu of Trendyol give a panel discussion at Web Summit in Lisbon.
Here are a few key takeaways from the talk.
Digital-first companies have a head start on data
Unlike bricks and mortar businesses that have moved into the digital realm, Birchbox’s starting position as online-only meant that it had the edge from the get-go.
According to Katie, this allowed the brand to collect and use data natively – as opposed to looking back at it in retrospect.
The company was also founded on the belief that the negatives of being online-only, such as its ‘2D’ nature and lack of textured storytelling, would be outweighed by the positives. Namely, the ability to capitalise on data to serve a better experience for customers.
Demet also touched on this topic, relating it to Trendyol’s ability to scale up at a rapid pace.
Instead of relying on physical stores, it uses data to interpret its customer’s needs in real time. Now moving into selling private labels, the ecommerce site has been able to bypass constraints that might limit traditional retailers.
Using data to speed up the supply chain
Speaking of time – another point raised during the discussion was how both companies use data in order to speed up internal processes.
By using data to determine in advance how a product might perform, it allows them to reduce or increase inventory in response.
So instead of buying now and hoping for sales – maybe even resorting to discounts later on – both rely on data to tell them how much to invest.
Trendyol in particular appears to be focused on this, using it to feed into every part of the business.
Demet explained how data not only determines what goes on the site, but it also informs other factors such as shipping algorithms.
Generating demand that didn’t previously exist
While both Birchbox and Trendyol use data to personalise for consumers, a lot of the discussion revolved around why it is vital to avoid venturing into ‘creepy’ territory.
For Birchbox, the aim is not to meet the expectations of the consumer – let’s say sending them an email about hand cream during winter, for example.
Instead, it is to generate demand that did not previously exist. Or more to the point, that the consumer thought didn’t exist.
This means using data to determine a picture of the consumer and to deliver a relevant message – but to also introduce them to something new.
According to Katie, this type of personalisation is the difference between satisfying a customer and turning them into a long-term and loyal one.
Trendyol also focuses on this notion of customer loyalty, emphasising the importance of long-term strategy – even if it means going against the grain. Demet cited that fact that 30% of the company’s sales derive from its loyalty program.
Likewise, instead of being afraid of product returns, the company has realised how an easy returns experience for a customer can be far more valuable – and this experience can in fact lead to more purchases in future.
Test and test again
Finally, both Katie and Demet cited the importance of testing to be able to truly innovote through data.
Trendyol in particular runs hundreds of tests at any one time, using it to inform the continuous changes being made on the site.
What’s more, data runs through the company’s DNA. Instead of analysis it in silo, it is a thread that runs through all teams.
Finally, Birchbox touched on the fact that data allows it to be a voice of authority for other brands.
Armed with the knowledge of what their customers are loving or hating from one month to the next (which, as Katie pointed out, can change like the wind) – it has found that merchants and stores are more than willing to listen.
(For more on testing, read ‘A beginner’s guide to A/B testing‘)