One issue that has been mentioned multiple times on Hacker News is that, in some categories, it is almost impossible to find good products: https://www.vox.com/2019/2/28/18168354/amazon-marketplace-pr...

I think their biggest weakness is their relationship with sellers. Sellers hate Amazon because their livelihood is dependent of Amazon's ever changing policies and bans. This is why most of them have a Shopify store and hope to create a brand on their own. But, they haven't succeeded yet. If you can provide a credible alternative for these sellers, there will be a stampede of sellers out of Amazon.

BigCo problems: slower decision speed, dehumanized support, limited social presence, volume requirements w/ supplier side, draconian terms, data collection and privacy issues...

Marketplace problems: limited control over 3rd party suppliers, fake products, lack of accountability for indirect damages...

Volume problems: fake reviews, hard to find specific niche products, lack of product details, standardized photos...

The same weaknesses that come up when you make any David/Goliath comparison. They are big, and slow to react, so you can hit them with speed and agility. Think early 90's when MicroSoft was slow to move to the web while tiny startups ate their lunch. Amazon will be hesitant to make major changes because they don't want to kill the goose that lays the golden egg.

How about something specific to Amazon?

Obviously those are all good, re David/Goliath. But in my experience investors dont think along those lines as much.

You can hit Amazon with niche categories. For example they have shoes and clothing, but there is no "Amazon for nurses scrubs or school uniforms" or other niche categories. They also do not serve many countries (e.g. Philippines). A startup that served these needs would not have Amazon as a competitor at all.

Out of curiosity, I just searched Amazon for "school uniforms" and there's a lot of stuff there. (Makes sense, since there are millions of kids in private and parochial schools - it's a big market.)

Even if you manage to find a retail category that Amazon is not currently involved in, you might wake up tomorrow to find that they've suddenly become your competitor.

You have found a category that Anazon serves, but does not serve well. I buy nearly everything on Amazon but I still have to buy her uniforms in person (each school has a specific color scheme, plaid pattern, collar style, etc.). My wife still buys her nurse scrubs from a guy who comes around the hospital and only takes cash. These are things you just can't get easily on Amazon, even though it looks like they have that category covered.

Also, you could compete with Amazon on any category just by having a bigger selection of same day delivery in some niche.

AWS is pretty expensive. Might be possible for a tech driven company to cut out all the MBA types and undercut them on price.

I would be very hesitant to use a new hosting service for anything business critical or sensitive. I don’t know how well they will operate nor the security of my data. AWS/GCP/Azure/etc. do a lot to physically secure your data.

Have you looked to see all AWS offers? If you just want VM hosting, their are plenty of cheaper offerings - including from AWS - Lightsail.

Lack of curation. I would personally be happy to pay extra for a marketplace with a limited set of products that have been tested & vouched for, instead of Amazon's Wild West full of Chinese no-name fly-by-night brands.

The biggest issues I see Amazon's e-commerce operation having are discoverability of products, and quality of products (namely counterfeits and used products being sold as new via FBA)

Amazon's weakness is that they're big. Big like an Elephant, do you know why Elephants are scared of mouse? Because they are small and fast.

Read "Innovator's Dilemma"

No dark mode on their ecommerce website, or in the AWS console. Seems like a critical weakness.

Sorry what is "dark mode" and why is this important?

They're referring to a (default?) dark color scheme on the web products, to make it better on the eye with better contrast.

Not really what I'm asking but OK... I doubt a lacking of a color scheme is going to add much to your competitive analysis.

Your humor detector got turned off, I think.

In lieu of any actual progress dark mode got hyped in many products recently as an important new feature.

Of course you doubt it Jeff, that’s why you don’t see it coming.

I think a wood grain texture would be the right choice