I would guess as: 1. For a lot of companies that have sites running the site is not their primary business. E.g. bank's business is not their site. Surely, the

I would guess as: 1. For a lot of companies that have sites running the site is not their primary business. E.g. bank's business is not their site. Surely, their customers (maybe majority of them even) use the site, but that's not what they came to the bank for, that's something they are forced to endure to get their loan, access to their money, their credit card account, etc. Consequently, no motivation to improve UX. 2. A lot of companies that do run their business via the site, they are big enough so that losing any particular customer is below rounding error for them. If you're mad at Amazon for scammy products, what you're going to do, use... what exactly? Not many sites have as wide catalog as Amazon, and can seriously threaten to take their customers. Consequently, no incentive to improve. Also, automatic problem handling tool cost much less than live support (that would demand benefits, complain about working conditions, and generally cause headache) - and if that costs 1% of customers wrongly banned, the cost is acceptable. In any case, not a lot of people would abandon Amazon because there's 1% chance they'd be banned by mistake - and in fact, nobody even knows how big the chance is anyway. 3. A lot of seemingly independent sites are actually owned by the same company (travel industry is notorious for that). So, if you don't like the UX on the site A, and take your business to the site B, you're didn't take your business anywhere in reality, you're just moved to the next isle at the same store. Same consequences. 4. Scammers are much more motivated to scam than services are motivated to fight them. If a scammer scams Google, they get $$$BIG BUCKS$$$ (or at least make a living). If Google catches a scammer, practically nothing changes in anybody's bottom line. Consequently, you get scammy search results. 5. Not changing anything and just keeping everything worked as it worked before rarely gets people promoted. Reworking the whole site and successfully implementing a buzzword technology and completing a project with 100 people involved in time and under budget - just might. Consequently, unnecessary redesigns all around.