1. You couldn't just lease hardware by the hour like AWS. So you had to put a lot of money buying hardware from companies like Sun. Also for the same dollar amount you would get a far smaller box in '90s compared to today.
2. No good quality free software, you had to buy everything from Oracle/Microsoft/Netscape to get a decent website up and going. Licenses cost a lot.
3. No lean startup model everything was a big bang approach which meant a lot of cash needed to get version 1. out.
4. No stackoverflow/google etc. so it just seemed to take longer for engineers to build something. Debugging took longer , the programming languages were less mature etc.
Put all together it is about a 100x reduction in getting something in front of users today.
Having started companies in both 2000 and 2012, I just don't think this is true. It was easy to put something in front of users in 2000 for less than $1000. We started with a $20/month shared host, PHP and MySql, and easily scaled to hundreds of thousands of users just by renting some dedicated machines. Google, Yahoo, ViaWeb, Facebook, and Ebay all started the same way. Yes, there were crazy $2M funded startups that bought Sun hardware and Oracle software, but this was absolutely not required.
Many parts of startups today are harder. Users' expectations for design and functionality is much higher. The internet was also much smaller in 2000, which made startups like Google possible.
The explosion in seed deals seems much more driven by the funding landscape, the rise of mobile and the social web, and cultural recognition of startups like Facebook.
MySQL and PHP both released in 1995 or so, there was a lot of movement in the "late 90's" so it really depends on where in there they are actually talking about.
Depending on what you mean by the "late '90s" I don't think it was quite that bad, but the budget was still up there. Apache was good enough for most if not all purposes by then, if you really knew what you were doing and your database people collaborated really well you could save 1/2 by using IBM's DB2 (very little independent documentation and lock vs. MVCC concurrency). Of course assuming your VCs bought off on all that.
Another problem is elasticity; do your architecture correctly and AWS will allow you to quickly scale. Otherwise you tend to buy too much or too little hardware, either of which can kill or damage you.
By then the web was quite useful for finding technical answers; obviously not as good as today, but still very useful.
2. No good quality free software, you had to buy everything from Oracle/Microsoft/Netscape to get a decent website up and going. Licenses cost a lot.
3. No lean startup model everything was a big bang approach which meant a lot of cash needed to get version 1. out.
4. No stackoverflow/google etc. so it just seemed to take longer for engineers to build something. Debugging took longer , the programming languages were less mature etc.
Put all together it is about a 100x reduction in getting something in front of users today.