30 Days of E-commerce Testing Day 24: High Profile Public Failure

Day 24 of 30 days of ecommerce testing is:

Who’s had a high profile public ecommerce failure that you’ve read about in the news? Share it on The Club.

So I cheated a bit on this one, I have a podcast for today!

https://testingpodcast.com/notd-international-edition-for-black-friday-2017/

1 Like

That’s about failure of Indian E-commerce giant (now acquired by Amazon) on it’s big billion day- something like black Friday you can say … but it was about performance and when it is India, you can expect the amount of people :wink:

3 Likes

I like these 2016 failures from blazemeter
https://www.blazemeter.com/blog/biggest-web-failures-2016-and-2017-resolutions
Interesting examples of failures after big everts and possible resolutions
‘ASOS website and app crash after Brexit’
‘Canadian Immigration website crashes after US elections’

4 Likes

What I could read on the news was a strictly local thing, so the articles are in Polish.
But browsing the web, I’ve found the story of an on-line apparel shopping boo.com (http://web.archive.org/web/20010927092825/http://boo.com/ ) a very interesting read, for quite a few reasons:

  • it was an e-commerce pioneer ('98-'00)
  • it was founded by people who had prior (successful) experience with e-commerce (bookstore)
  • it crashed rather loudly ($135 million invested in its 18-months-lifespan)

The reasons for its failure are many and various, but test-wise a couple of them has caught my attention:

  • They used a lot of Flash/JS for rendering pseudo-3d views of products - so what we sometimes use today, for flag products, on separate pages (en masse its not for obvious reasons). Combined with the fact that the size of some of their pages was serious (like home page : ) - it killed the performance for many/most users (fun fact: they warned their users that “this site is designed for 56K modems and above”)
  • They used complex category navigation for luxury items - meaning that there wasn’t enough products in each sub-sub…-category to make the division worthwhile (+ sometimes the only products there were out of stock : )
  • there were many (incoherent) standards of navigating aroundthe page - now an obvious thing, but reminds me why this breadcrumb must (almost) always be there, as an absolute frame of (navigational) reference for the user (Where am I? Where was I before? How to get out?)

2 Likes

There are the obvious data leaks every few months and performance issues on major shopping occasions.

But there are also:

None of these were really huge but I found them inspiring as a tester. Or just funny :wink:

2 Likes

Twitter peeps:

This was all i could google and find out:

Happy reading!!

1 Like

Alexa trolling is classic, South Park did anepisode where Cartman was able to add items to viewers shopping carts if their Alexa was close to their TV

3 Likes

Here’s my Day 24 https://wp.me/p9EXXo-5V
Love this story, lessons on failures in testing and in PR

1 Like

I had a hard time on this one at first, because typically I associate high profile failures to data breeches, not necessarily e-commerce site failures.

I asked in our slack channel for guidance, and got a two word hint:

“Black Friday”

Although it’s somewhat aging news, apparently Lowes’ website crashed last black friday, as well as British game store GAME. Lowes’ website came back quickly enough and GAME still sold thousands of products on it’s online store as well ask brick-and-mortar stores, but the public outrage was palpable.

Being int the U.S., I didn’t hear about GAME, but I didn’t hear about Lowe’s either, probably, thankfully, due to a quick recovery time.

-Dave K

2 Likes