no-line commerce in action
When tasked with innovation, it is all too easy to lose oneself in the bulls**t world of buzzwords,
gadgets and novelty for novelty's sake.
Instead, we opted for a problem-driven approach and consequently began by identifying and targeting a
few of the issues that offline fashion retail (OFR) suffers from:
- building brand loyalty is comparatively hard in OFR, even more so in the budget sector. There's no
technological eco system, no emerging harmony between products that encourages sticking with a brand. On
the contrary, customers often enjoy combining items from a wide variety of brands to create their own,
Consequently, brand loyalty has to be generated and sustained with ultra-expensive marketing campaigns
and even then a brand's popularity is rarely more than a fleeting phenomenon.
- although not limited to fashion, offline retail in general suffers from a lack of information about
its customers shopping behaviors. There's no tracking of items "in the basket" versus items bought, no
metric of how often something is tried on and by what target group.
- for the past decades, the apparel shopping experience has remained largely unchanged. There is a
distinct lack of fresh ideas despite the fact that competition has become increasingly fierce. To gain a
true edge over
its competitors, a brand needs to implement radically new concepts in order to distinguish itself from
- while online stores alleviate some of those problems, they struggle to recreate the offline shopping
For many customers, apparel needs to be seen, felt, tried on and walked in, before the buy-decision can
be made. This has led to enormous return costs and a contributed to the sustained popularity of OFR - in
sharp contrast to most other retail sectors. Fashion retail therefore urgently needs to bridge the
divide between online and offline shopping.
We found that we could address all those issues by reimagining a central aspect of OFR, unique to the
industry: the dressing room. To that end, over the course of three iterations, we created a system that
allows for easy enhancement of existing rooms.
The first iteration comprised a 360° multi-camera setup as well as a tablet and a human-sized digital
'mirror', that enables the customer to view herself from any desired angle as well as to take pictures
that can consequently be shared via a wide range of social media.
Great emphasis was placed on ease-of-use in order to provide a seamless enhancement of the shopping
experience instead of a cumbersome burden. This required that any and all of the features provided were
entirely opt-in for the customer, so the traditional, analog experience could still be attained without
any action required on her part.
In the second iteration we focused on creating direct value for the retailer by tracking what articles
the customer brought into the dressing room and what she ultimately purchased. To this end, we
incorporated RFC tracking technology and provided the data gathered to the most common ERP systems via their
respective APIs. The entirety of the data is stored securely with AES 128-bit encryption and transmitted
to the ERP system in anonymized form.
For the third and last iteration we decided to up the ante and enhance the customers experience by
providing meaningful suggestions on what articles would complement their current choice of apparel.
This was accomplished by employing computer vision techniques, providing information about the users
current choice of clothing in regards to coloring and a machine learning algorithm, that utilized color
theory and ERP systems clothing databases to find suitable additions while incorporating factors such
as the current season and the customers gender. While heated debates about the conformity of the
algorithm with current fashion trends have been raging within our female workforce ever since, we found
the resulting suggestions to be superior to the seemingly random results (in regards to fashion
compatibility) of comparable, less specialized recommendation engines.
Easy and cost-efficient integration into existing retail-infrastructure is facilitated on the hardware
side by keeping the setup ultra-compact and flexible in regards to the dimensions of the room as well
as the number of cameras used (from 1 up to 32).
On the software-side making use of retailers already established online user account systems as well
as common ERP system's APIs means that changes to existing IT systems are kept to a minimum.