I just signed up to thread.com and had a consultation with a personal stylist. A pretty eminent one too. She asked me some questions about what I wear, and the brands I like, and she looked at some pictures of me (poor woman). Then she recommended some outfits.
The result: I liked most of what she plumped for. In fact I already owned a couple of very similar items. Clearly I have a better sense of style than I thought…
The site already had my sizes. I handed over my credit card details and picked my favourite items. I’ll try them on at home, and if I like them, keep them. The site — and my stylist — take a margin on the transactions.
It’s a very personal service: even mediated via webchat you know you’re dealing with a real person. It feels premium, yet it costs you nothing. But most important are the recommendations.
Amazon has reportedly spent millions of dollars building its recommendation engine — the software that drives the suggested purchases at the bottom of the page. Yet I find I am much more likely to find something I want to buy on a walk around Waterstones.
This is a problem both for the online retailer and the high street.
The online retailer would like to replicate the feel of a premium high street store where an attentive shop assistant helps you find what you want and recommends suitable purchases. Why? Because you will spend more.
Doing this economically usually means automating the recommendations. This is not something computers are necessarily good at — hence why Amazon has spent so much.
The high street has a problem because more and more of us are discovering what we want on the high street and then buying it — usually cheaper — online. So-called ‘showrooming’.
Thread.com has avoided the multi-million-dollar development costs by giving stylists a way to earn a few extra bucks. Work as a stylist — I can believe — is not 9–5, Monday-Friday. It is intermittent: frantic weeks followed by frantic worry. With a few minutes here and there, and an internet connection you can earn some cash just by doing what you do from the comfort of your home, your bed, coffee shop or local bar.
This is a prime example of the sort of secondary income that I believe many of us will need and have in the future. But it is also evidence of the current limitations of personalisation technology and recommendation engines.
One of our long-term goals with CANDDi was to recreate the local shop experience, where the retailer knew you just well enough to create a personal, engaging shopping experience. We can do a lot of great stuff with the software, but we can’t do that — yet.
Even when we can, personal recommendations will retain a premium value. Dealing with a real human being will be something we pay for, either directly or via a margin as with thread.com. High street stores may start to offer virtualised personal shoppers — real people but communicating with us remotely — as a way of providing more support in-store but with a lower cost (and potentially more knowledgeable service) than a body in-store.
Whether they are recommending boots or books, gadgets or grub, human recommendation engines are here to stay.