What If Amazon.com Actually Is a Horrible Website?
buzzfeednews.com
buzzfeednews.com
We are left to consider that perhaps UX, particularly to the mass market, really doesn't matter. Or maybe that Amazon has such an entrenched lead, mindshare or other, it can weather some horrible, horrible experiences.
Considering that AWS is almost identical on the cloud front, I'm not sure really what to think here.
I'm a programmer. I use cloud services everyday. I have actively stayed away from AWS in the past few years b/c it has such a bad UX/DX and yet....it grows.
Azure is better. GCP is better. AWS, when compared to those two objectively, is downright terrible. Their services are disjoint. Their command lines don't work together. Their web console? Hooooooollllllly mother of god.....
I guess I don't know what to make of Amazon. Developers seem to fawn over things like Heroku, Zeit, GitHub and they are hugely successful in their own right....but still AWS is used by developers and, IMO, it shouldn't be anymore. It's 2018 and AWS is very clearly stuck in the 1970s/80s DX. They just don't get it, or maybe they do? And DX doesn't matter to most developers?
If we are honest with ourselves, AWS should be in third place in 2018. And it shouldn't even be close. Their services are comparable to the point of no real differentiation to their competition, and they have worse billing, worse experience, worse DX, worse support and overall worse nearly everything.
And yet they grow.
So yes, Amazon.com and AWS are terrible. And for some reason it doesn't seem to matter. I would like to live in a world where it does, particularly for developer tools where I hope that developers have more taste and sense than the choice of AWS shows.
I'm a developer and don't know what that is.
Sometimes referred to as "dev x".
Admittedly not the most easily understood acronym, and I'm sort of guessing.
UX DX
you will find lots of hits about.
That still doesn't explain AWS v Azure or GCP which have similar (or better) services, more than competitive pricing and in Azure's case, waaaaay better support.
In fact, everything about AWS is typically worse. Perhaps AWS entered the 'no one ever got fired for buying AWS territory' even when there are a ton of better options out there. If that's the case, kinda sucks for devs who are stuck using their crappy services when so many others exist.
Agreed that the console's UX violates almost every UX principle in terms of its use of signifiers, etc.
We literally have AWS technical generalists on-site once a week and they bring in subject matter experts on specific topics/services as-needed, sometimes on-site, sometimes via video.
A lot of people are probably using something like puppet/chef/ansible/terraform and don’t encounter the web console very often.
I think CloudFormation is difficult to use and could be much easier, but ops rant about the AWS UI being unfriendly makes it sound like he’s not really the kind of customer AWS is really after.
I started a chat with AWS after giving them the ARN of the template. He looked through my CloudTrail logs and the documentation and couldn’t figure it out either. He said he would get back to me....
He created a simplified CF template that reproduced the problem, talked to an CF developer, he said the developer looked through the code of the API call that CF was attempting and they found that the property was case sensitive. The guys at Hashicorp couldnt do that.
Don’t get me wrong, I’m a Hashicorp fanboy when it comes to onpremise implementations. I developed a system around Consul, Nomad, and Vault with a helping of Fabio.
I do think people in this discussion are looking at this from a developer or designer point of view. I read some article a while ago about how Amazon saw a increase in sales from millisecond level improvement in page loads. I bet that’s what they optimize for, and less so on how pretty the UX is.
That all being said, I think Amazon is turning into EBay with all the garbage ads I see on the website, just trying to buy replacement toothbrush heads, but that’s an entirely different discussion.
I imagine a company like Amazon has the resources to do both.
> I do think people in this discussion are looking at this from a developer or designer point of view. I read some article a while ago about how Amazon saw a increase in sales from millisecond level improvement in page loads. I bet that’s what they optimize for, and less so on how pretty the UX is.
Same point. These features, performance and an attractive user interface, aren't mutually exclusive. I'd argue that in fact both are tightly coupled and are part of what makes the design and user experience.
Amazon is successful because it has mostly everything in stock and can get most of it to you in 2 days. Amazon is logistics and trust. The trust is eroding and the logistics is worse because they're delivering things themselves, but still that is why it is.
AWS has more than Azure and GCP.
Appearances and experience aren't the whole of a product. Overfocus on them as a producer or consumer is a problem.
Obviously Amazon.com is different and it's mostly about logistics, inventory and reduced buyer friction among some other things.
It was probably much more important in the early days than it is now.
There is no reason to use AWS for green-field software development. But who's even doing that, these days, anyway?
I mean, lots of people are, right? I'm not myself, so I guess I don't really know, but surely lots of people are still making green-field software (if not, I think our industry is in some trouble...) and presumably a reasonably high proportion of those people are using the current #1 cloud hosting provider. Which part of this is wrong?
I'd argue that if they were building towards one, big, consistent UX (a UX they'd prefer to use the APIs and command-line tools directly anyway, rather than the GUI), that they wouldn't have the amount of functionality, breadth-of-coverage or marketshare they have today.
Out of all the things I care about, the user interface is not one.
I might run through the console to learn how something works, but in general I'm interacting with AWS through a terminal, and doing so that way is generally quite enjoyable.
I agree that GCP at least is a better experience for me as a developer, but I'm not sure if that translates directly into better anything for a customer of mine.
It may not be good UX as measured by some aesthetic standard or set of best practices, but the site is accomplishing its goal for Amazon and arguably the consumers who use it.
Maybe they should A/B test this team and go with the other one.
They miss the big changes like 'shall we show all the products in a 3d immersive VR gallery' or 'make a site which downloads all data and works entirely offline'.
Those big changes might usually fail, but if you never try any of them, you'll end up with a site that looks and behaves like it's from the 1990's, while the rest of the world has moved on.
The consumers who use it have no goal other than getting stuff they think they need, faster.
1) You can A/B test yourself into a corner. They are definitely doing it right, whether it is the right thing to do is a different question.
2) Having some process insight into the operations / logistics side I can confirm that everything thing is connected to the webside one way or the other. That is what gives Amazon its huge operational advantage. Downside is, with everything developed in-house, this approach can end up as a motely collection of legacy systems. And that makes changing things difficult and had the potential to become a major pain in the ass long term.
just because they have a process doesn't mean it works. but yea, like someone said, i guess the general user doesn't care. i know it has lessened my use of the site.
So, it ugly and bad as hell, but everybody is trained and accustomed to it. Why risking confusing customers without need?
To me it's rather simple and clean, and I would expect that they have a vast team of UX experts testing their site to make sure it meets their overall product goals.
Besides, what you might find aesthetically pleasing is not necessarily pleasing to a global population of users with varied backgrounds/diverse age groups -- e.g. older generations might prefer the super simple fonts and general design standards that were more prevalent one or two decades ago.
Especially for electronics or computer parts, if you don't know _exactly_ what you're looking for down to the part/model number, you're not going to get any help on amazon.com. I get the feeling they don't do any actual categorization (like say newegg seems to do) and instead you're actually getting just full-text search on product descriptions.
If you really want the part soon, you'd be fine with either "Prime" or "Prime FREE One-Day", right? But you can't search for both at the same time, for some reason they are mutually exclusive. So you duplicate the browser tab and search for "Prime" options in one and "Free One-Day" options in the other, like a caveman banging two tabs^W rocks together. I just noticed there's yet another option, a checkbox under "Delivery day" reading "Get It by Tomorrow".
If you _do_ know exactly what you want, good luck getting it because of the whole similar products from different sources getting intermingled in inventory.
I'm frankly blown away by how user-hostile the whole thing is. Prime shipping buys a lot of goodwill, it seems.
It's actually a terrible site to "browse" if you're looking for something but not really sure what, but if you know roughly what you want it works fine.
I would love to see developers prioritizing themselves better. But they don't, themselves, do this.
There's not really any news here, other than the fact that managing such a huge _and extremely effective_ commercial website is difficult even for one of the most well-resourced tech companies in the world.
If the purpose of an website-that-sells-things is to have carefully tuned UX for every product page, Amazon.com has some small (although hardly 'horrible') UX problems. If the purpose is to sell stuff, the article makes it very clear that for an arbitrarily specific product (bassoon straps), you're extremely likely to be able to buy one quickly and easily from Amazon.
What if BuzzFeedNews.com is actually... trolling its audience?
Couldn't agree more. Buzzfeed always makes me cringe. Even when an article sounds genuinely interesting and less click baity, I can't bring myself to visit the site.
The toothbrushes I'm buying on Amazon might be $4 with free shipping but I immediately question its authenticity. I've bought enough junk that's turned out to be a scam[0] that it's just easier to buy the same thing in a store. At least 15% of my one-off purchases have turned out to be regrettable. Hell, I even poked my head into one of the Facebook groups of sellers and "buyers" gaming the algorithm to see what it's about, and couldn't be more astounded to find that _everything_ on Amazon is being gamed. Everything you can buy is being replaced with knock-off junk.
Amazon's greatest asset, its convenience, has become outweighed by its low quality bar for me. It's a real disappointment. The massive influx of utter garbage has made everyday purchases risky, and it's an easy risk to avoid.
[0] I once bought a shirt that was physically impossible to put on for any human being with a head diameter greater than 10cm.
Also I went through two years of personal and business Amazon purchases and literally couldn't remember a time when I had an issue. Especially after the last HN discussion made about this I realized just how much I still buy in brick and mortar.
Ignore for the moment the counterfeit issues, because those are not a "site" problem.
I still routinely fail to get what I search for. And then find it i. The "also bought" of whatever they do send me. Shipping times either lie more often than not, or are "true" after a seemingly arbitrary number of days between now and when it ships.
Anything that is deeply categorized (prime day. Prime now, Restaurants, etc) loses your filtering all the time. This past prime day I had to search through their pages of deals by hand because their search just failed.
This might well sell a lot of stuff...but like a news article that 8s broken up into multiple pages to seve me more ads, it is not good at what I want. Other forces (network effect) keep me there.
Any given item in Amazon's database (where 'item' == 'something that has its own page') can have multiple variants on color, size, and other varieties. Each variant may have its own price. So if you want to sort items by price, which price do you use?
E.g. if I search for a computer it should pick the variation with the cheapest cpu, not the one with bring-your-own-cpu
I occasionally dealt with the Amazon catalogue. It's a great example of 'garbage in, garbage out'.
This is also the reason I never bought computer parts on Amazon, even when I got a staff discount.
Use the lowest price, but include filtering options to exclude items based on the criteria that makes them different (e.g. show only Small shirts that are Pink or Yellow), and use the lowest price of the items matching the filters.
Even for a single SKU, Amazon will often stock the same item as new/used/refurb, from several sellers, at a whole galaxy of prices. This means there is no lone price to sort by, because you can't start with asking the user "Are you okay with a used one? Will you trust Acme Reseller Ltd as a merchant?".
Search for a product, find 156 results. Sort by rating or price and it's now often something absurd like 6 results. Where did the rest go? No manager or dev ever noticed this when it's now impossible to shop there and not notice this? This seems like a great whiteboard interview question in our strange alternate timeline "how would you sort a list without losing well over 90% of the list?"
Go back to the original search, tediously wade through the crap and realise none at all of these are the actual item you tried to find. Randomly click around a few of the shitty products shown. Find the actual product, often from the actual manufacturer with thousands of reviews, in the "customers who viewed this also viewed" or "customers eventually bought" lists. Question your sanity. Go back and check search. Nope, the actual item isn't there in the primary search at all. Just the shit.
Fortunately for me at least, there aren't enough other forces to keep me there and I now mostly shop elsewhere. Sometimes it would be annoyingly easy to just Amazon it, but they're not even cheapest or fastest any more.
It's not just suboptimal search and categorization, which is a difficult problem across a product range as large and changing as Amazon's (but would be considered intolerably bad on virtually any other ecommerce platform, except possibly eBay). It's also deliberately making choices other websites could never get away with, like sticking banner ads for a third party website right above its additional product recommendations. Someone will have done the calculations on that and figured they make more from the clicks than they lose from people not coming back to Amazon though. Probably the same goes for any product page refresh/reorganization clean enough to make users wonder if they're still on Amazon...
I'm still a loyal Audible user, I might end up cancelling Prime soon, since I no longer order from Amazon enough to justify it. I've been a Prime user since 2009, since they were giving for free to students around that time. Now I'll just walk into Target or Walmart, or if I'm buying a PC/game/console getting it straight from the OEM such as Microsoft.
1. Most of the changes/extra features were added piecemeal over the decades, with the basic site look not really changing all that much in the meantime.
So a lot of the awkward/poorly thought out/non integrated features are there because that's where it was practical to add them.
2. Obviously it's a large brand, so people are more forgiving of the design issues compared to a lesser known competitor. That's why copying the market leader may not always be the best idea, since the leader can do a lot of questionable things/make a lot of questionable design/content/whatever decisions and still have millions of people using their services, whereas a small company/site with these frustrations will just go ignored/unused.
3. They A/B test everything, and what the data says is most effective is not always what a designer would think looks/feels best. Look at some of the best performing landing pages; many look outright hideous yet convert really well.
As a customer, I find that Amazon has some of the worst usability of a shopping site that I have run across in a while. I think amazon is successful in /spite/ of their design, certainly not /because/ of it.
If their design is actually "bad", then my only takeaway is that design isn't nearly as important as people claim (at least, for this particular industry).
95% of the time, everything works as I expect it to, which is about as good as it gets for me.
My gripe was mainly about the density of the UI elements - it always felt like there was a lot of text/images crammed into a small space, so I had to inspect every single option before I found the one I wanted.
People have made similar comments about the Azure portal, though, but I never felt the same, so YMMV. Like I said, personal preference.
- I think that Javascript is widely used needlessly forcing lots of unnecessary reflows and repaints; this slows down searching and using the site.
- I find the searches are broken to the point where it is easier for finding stuff using another search engine which indexes amazon.com's webpages.
- I don't like collapsing portions of the webpage (such as long reviews). I'd rather download simple static markup and scroll through reading what I want presented to me in the order I've chosen.
- I don't want to see JS-based image/movie previews; inline images and using a modern browser's built-in video is much better as those can be styled in the way the user wants and obey other user-centric preferences as the web ought to do.
So to me saying 'it works' is putting commercial interests above my user interface interests and interests I deem more important such as how Amazon treats its workers. At best I'd say this puts into perspective how overvalued website development really is for commercial interests and how wildly off the mark website development is at identifying and building on something simple and easy to use. I've also found that Amazon purchases are often misleading (selling used merchandise as new, selling USB keys that hold less than the advertised storage size, and more) which means more time and money spent not really knowing what's on offer and higher chance of returning bad merchandise. All of this taken together means I shop elsewhere.
We had a lot of non web developers writing JavaScript. No idea if it's like that any more, but I can certainly say that that was responsible for some serious performance sins!
Modern browsers ... hahahahahahaha. I suspect parts of the website still fully support IE6.
Amazon has picked exactly one low-hanging fruit for recommendations: it frequently recommends to me that I should buy an item from my wishlist. That was a good idea on Amazon's part.
But it also frequently recommends books to me based on a book that I purchased, read, and left a one-star review for. That seems like a pretty obvious mistake.
The reasoning might be that the bad-review signal means that you found the book to be a low-quality instance of its type, not that you discovered that you disliked all of that type.
Unless you're being recommended books by the same author, I suppose.
I don't get that either all I get is the 'Sponsored Products..' one, maybe I'm in some A/B test
The Roku software is top notch. It’s amazing what you all have done to optimize the software to run on really low end hardware. I’ve also heard good things about the SDK. I listened to a podcast interview of your CEO and he said that using low end hardware was a business decision to keep the cost down for the OEMs where the margins were already razor thin.
That being said, I hate Roku’s business model. From the same interview he said that Roku isn’t in the hardware business or for that matter even in the software business. It’s business model is collecting user data to sell to advertising.
Everything about that business model is irritating to me - from the remote with hard coded buttons to services that no one would choose - like CBS News and the defunct Rdio service to the ad that takes up half the screen.
I have a 4K AppleTV connected to the one TV we use most frequently and I replaced the Roku stick I had in my home gym with another ATV 4K.
The Roku business model is everything I hate about the consumer electronics business model outside of Apple. I like simple transactions - I give the company enough money to profitably stay in business and they give me stuff.
That being said, if my other two non Roku TVs in my house ever need replacing, I would replace them with RokuTVs. The interface is great, the TCL Roku TVs are dirt cheap, the mobile apps are great and Roku has a history of keeping their software up to date. I couldn’t recommend that most people buy an $199 4K AppleTV box when you can buy a $399 55 inch Roku TV from Amazon.
I've been fighting with seller support for about two weeks to get this fixed. First they said they need a link to the manufacturer's website for the product (!?) Then when I gave them the link they said I need to upload the new pictures myself (!?) So I did, I downloaded the pictures from the manufacturer website and uploaded the black pictures on the black version and the white pictures on the white version. Now they've just been sending me messages every few days saying "we are still working on this issue." This is for a real simple fix, not some sort of complicated issue, but it's a big issue if a customer doesn't know what they are getting, this should be priority one.
How would you know? The result is "simple error", but the cause could be deep down in their backend.
As a random dev in a different dept, I even got the first iteration of the Amazon app's notification ads pulled temporarily. They were badly implemented (they advertised a product, but linked to something different, and there was no obvious way to turn them off), I felt they reflected badly on the company, so I filed a ticket and it actually got listened to.
All that's gone out the window now. Every time I see the Prime clickthroughs, credit card deals, 'lightning deals' which are only about 20c off the normal discount, etc, it makes me grind my teeth.
Maybe it would be different now that I know my way around these things, but the only really usable cloud services I've seen were DO & AWS (DO was the simplest).
edit - I wonder why my experience has felt so different from everyone else. I use AWS in part now because its so easy to work with (besides the billing page)
I often find a random docu-series on Amazon and completely forget I was enjoying it or what the name was when I want to go re-find it.
I often wonder if I made a good desicion after purchasing something. That doesn't happen with Amazon, instead I'm just happy that I don't have to use that shitty page more.
I was visiting amazon.de (not .com) yesterday and 50% of what I saw had article names/descriptions like eBay 10 years ago. "desk light white awesome whatever keyword buy". Then I looked at the Audio CDs and sorted by price, and everything was 0 - XX €, yeah - because I was logged in apparently some of the music I can stream for free. Totally interesting to me if I deliberately visit the AUDIO CD section. Also not sure if it still sorted by price or not, because for sure the first 3 pages were all 0.
Anyone know to what extent Amazon.com still uses Oracle DB? I know they said they would be completely off Oracle by 2020. Is Oracle really that good for mission critical applications that no open source DB can match?
The problems were mostly associated with Oracle specific syntax or implementations, nothing else
In the past year, I've almost bought:
- a USB-A cable instead of a USB-C cable
- a Pixel 2 XL case instead of a Pixel 2 case
- a sink strainer instead of a sink stopper
http://digg.com/2017/amazon-vs-walmart-size
The pace and scale is even more dramatic than I thought.
Rising interest rates and a big jump in expiring retail sector bonds could make 2019 an especially difficult year:
https://www.bloomberg.com/graphics/2017-retail-debt/
The potential impacts are hard to fathom.
> The job losses in retail could have unexpected social and political consequences, as huge numbers of low-wage retail employees become economically unhinged, just as manufacturing workers did in recent decades. About one out of every 10 Americans works in retail.
https://www.nytimes.com/2017/04/15/business/retail-industry....
That article makes it sound like all of suburbia is one giant manufacturing town, and the plant is about to close. The first link makes it look like we somehow decided to ditch the entire sector and just give it over entirely to Amazon and Walmart.