“Look”ing for problems in all the wrong places

Amazon Echo Look: Solving Problems That Aren’t Problems

I just received an email from Amazon about the Echo Look. My first reaction was: what is this? After reading more about it, I asked, why would I want this?

Echo Look invitation email.

Amazon Echo by itself has great potential as a chatbot for a smart home. Though I would argue that Amazon Echo is not great at natural language processing (NLP), it is great at connecting a diverse array of connected home products. You know what isn’t a real problem in the connected home? You needing anyone, machine or human, telling you how to dress on a regular basis.

As a consumer and Amazon Echo and Dot owner, here are the challenges that I doubt Amazon has figured out the nuance of given the number of things Alexa still can’t do or understand:

  • Outfit context: what is an appropriate suggestion for work vs non-work
  • Picking out new clothes: Assumption is they will want to suggest clothes for consumers to buy. The thing about machine learning is that the computer uses existing data. If all your current clothes are black, grey, and white, it will continue recommending those for you, even if you want to branch out.

Additionally, I’ve tried services that recommend clothes before, and after 5+ iterations of recommendations, I’ve bought zero of the suggestions.

As a consumer, I don’t see enough benefit of having someone recommend clothes I already have and recommend more clothes I probably don’t need to pay $200 for something I could text an image to a friend.

This is such a specific, niche problem to solve, I see this going the way of Google Glass. Several early adopters will jump in and share their experiences using it, then purchase of the Echo Look will fall off, as well as usage of the ones already purchased.

This product gives Amazon more benefit than it gives the end user.

Data. Imagine having data on what everyone is wearing, and what they seem to like. What would retailers pay for that information? Imagine how Amazon could use that data to alter its algorithm for what clothes it recommends to you. How much more could Amazon charge for ads for its fashion business if it claims it knows your wardrobe so well it can increase the likelihood you’ll click on that ad?

For the amount of money Amazon can make from that data, $200 seems like a steep price for the consumer to give away the data in their closet. Solving problems that aren’t problems is a good indicator of fad technology. I can think of a number of applications for Echo Look, and style consultant is one of the most useless.