When personalisation goes wrong: an ecommerce story.

"A story of dried mangoes and split yokes"

A key investment theme for me around ecommerce is personalisation.  I have a new post coming on this but a quick bit of background on the sector by way of anecdote. Two examples from recent personal experience:

 

Graze have a beautiful box and bodenesque (innocentesque ?) branding for middle class brits. They also have a very comprehensive online form to fill out so you spend 15 minutes telling them exactly what you do and don’t like so they can tailor it for you.  However my first box arrived and I contained something I hate: Macadamia Nuts.  I cancelled it.  I am sure it was my fault.  I am sure I missed a tick box somewhere.  But the point is that if you raise expectations for a perfect product (which by definition a personalized product is), then too much choice can be negative.  

 

TailorStore solves my problem of over-paying for tailored clothing.  Which is a big problem for me. After a similarly convoluted online form-filling process (but this time involving a measuring tape, my wife, an old shirt and a mirror) I got a shirt ordered.  However when the shirt came, the shirt did not fit.  A bit of forensic investigation revealed an error in one of their pictures on how to measure for a shirt.  This was their fault and after an awful customer service episode I have a credit note, but the point here is that there is a now a useless shirt on my shelf that has pissed me off and destroyed the margins of Tailorstore.

 

As an investor this is another area of caution for personalisation businesses: Returns can completely alter the profitability of a standard ecommerce operation but when you can’t sell the returns on…

Comments

[Dave Stone]

Couldn't you put these down to UX/UI problems?

I must say, of all the startups I've met in London recently at various events, a good 1% have good UX/UI — the rest resemble a train wreck. Sure, underneath the hood they might be functionally sound, but a while to go before they're end user friendly.

However, I've used graze & loved the UI/UX there — found it very solid, got what I wanted & recommend them all the time.

[leafar]

So true. Personalisation is a kind of mixed blessing. I have been experiencing it with www.ulike.net
To deliver something really personalized it takes time and one mistake ruins every thing. It's even more tricky when it's physicaly consume like the shirt or a box.

About to much choice i recommend that you like at the paradox of choice here :
http://www.ted.com/talks/barry_schwartz_on_the_paradox_of_choice.html

Sincerely.

[PaulSweeney]

Great "personal experience post". The Tailored clothes issue has one that has been attempted over and over since 1999, and it's never been truly cracked, perhaps for the very reasons you cite, "human error". Even if it starts bad, an automation solution with the ability to iterate is probably (only probably) a better business model. On a personal level, if you do find a good service, let me know. :)

I would really love to see a good iphone-camera/video solution, mashed with a motion sensor software for this. How hard could it be? Point the camera at your spouse (putting their head in box a la like.com and off you go.....).....

Finally, the job of "handling the exceptions" is the bit that breaks the margin as you point out. If you can "get into that final 1-5% that are error/failure" you have a real chance of operating profitably. How companies "catch the error" is key. You can "wait until it happens and then react", which is good, at least you are dealing with it, or you can "profile use cases/people" and get pro-active. I suggest that in both cases you were a first time user? or first 5 attempts? Critical to get in there and pro-actively check that everything was done correctly.

[custom term papers]

Good and informative article. I like this.
Thank you very much.