One big thing people don't ever talk about is how much more expensive it was to get a site running back then. I was hitched to one of Utah's brightest dot-coms in 1998, and building a basic web portal was a LOT of work.
Consider our building blocks: Oracle or MySQL and Java with servlets.
Everything else was written in house, because the open source alternatives were all currently being written in-house elsewhere (or were too immature to use). It took phenomenal amounts of work to get the most basic sites working.
To make matters worse, horizontal scaling barely existed at the data tier, so the only answer was stupidly-expensive Sun boxes with even more-stupidly expensive Oracle licenses.
For hardware, you had little choice but to buy boxes and drop them in your ISP's kind-of-datacenter. Boxes purchased up-front, please.
We had about 15 engineers accomplishing about the same amount of work as the two engineers at my current company. The difference in productivity astounds me every time I think about it.
Now, all that said, companies were still stupid with their money. We had as many product managers as engineers at one point, because everybody knows it's as hard to think up new features as it is to build them (I'm a product manager now and understand the real work involved in good product management, and I still think it was stupid). We made multi-million dollar deals that failed to work out that a leaner company would never have touched.
When I look at the current environment versus back then, I see some very interesting differences: first, the productivity enhancement means value can be created for far less money. As a result, many, many more companies are getting a piece of the action. That is potentially good for investors because it broadens their portfolios. That is probably good for entrepreneurs.
The other problem is there are very few perceived mega-winners. This means that investors aren't getting many large returns and, when somebody smells "jackpot", there is a frenzy. That is good for successful entrepreneurs.
Consider our building blocks: Oracle or MySQL and Java with servlets.
Everything else was written in house, because the open source alternatives were all currently being written in-house elsewhere (or were too immature to use). It took phenomenal amounts of work to get the most basic sites working.
To make matters worse, horizontal scaling barely existed at the data tier, so the only answer was stupidly-expensive Sun boxes with even more-stupidly expensive Oracle licenses.
For hardware, you had little choice but to buy boxes and drop them in your ISP's kind-of-datacenter. Boxes purchased up-front, please.
We had about 15 engineers accomplishing about the same amount of work as the two engineers at my current company. The difference in productivity astounds me every time I think about it.
Now, all that said, companies were still stupid with their money. We had as many product managers as engineers at one point, because everybody knows it's as hard to think up new features as it is to build them (I'm a product manager now and understand the real work involved in good product management, and I still think it was stupid). We made multi-million dollar deals that failed to work out that a leaner company would never have touched.
When I look at the current environment versus back then, I see some very interesting differences: first, the productivity enhancement means value can be created for far less money. As a result, many, many more companies are getting a piece of the action. That is potentially good for investors because it broadens their portfolios. That is probably good for entrepreneurs.
The other problem is there are very few perceived mega-winners. This means that investors aren't getting many large returns and, when somebody smells "jackpot", there is a frenzy. That is good for successful entrepreneurs.
But I'm not sure I would call it a bubble.