My start-up experience (as an employee, not a founder) was absolutely terrifying.
Shortly after starting I found we had 6 employees (total, not including founders and funders). 2 of them were classified as HR. Turns out both founders were from a large corporate background and insisted on running the company as if it were IBM or Microsoft - not a small company. From reading some posts on this site I see that is not at all uncommon.
Our product was more geared towards the corporate market rather than public access or consumer. I can see how presentation makes a difference but it surely wasn't worth all the effort, time and money devoted to it (often more than was devoted to product development).
In the time I was there we released three major revisions of the product. At the end of each revision our workforce would triple (the additional people being sales people) who would traipse around the world (often with a developer in tow to answer technical questions) to tell everyone who would listen about our feature list for the next version of the product. Strangely enough most potential customers decided they would wait for the next version to come out before they would commit to buying anything. On the upside I got to visit Malaysia, the US and the UK on the company dollar during these periods.
The development team realised we could scale down our product to run on consumer level hardware and spent a lot of time ensuring a common code base could be compiled to run on a range of hardware (with features excluded or scaled down to cope with limited capability depending on the target platform). This was at the start of the whole Web 2.0 phenomenon but we still included support for XML/RPC calls back to a central server for long term data storage and common configuration.
The proposal we gave to the founders was that we should license this out to the firmware manufacturers for the consumer equipment and license the server side support to the major ISP's in our markets. The first approaches failed (possibly because we didn't have a 'name' for doing such things but - I personally believe - because of the huge prices that the company wanted to charge as well, per megabyte transfer on firmware licenses. Yes - you read that correctly - to use our firmware they had to pay a fixed price per megabyte of data that was routed through the device).
After this initial failure the solution was not to reduce our price (or at least reduce prices for our initial market to get the product out there) but to make a plan for global domination of all consumer networking equipment. Yes, we were going to buy hardware from Taiwan, load our own firmware and run all other manufacturers out of the market. In less than 12 months. After that we could charge what we wanted for the ISP side of things which would be the real money maker according to our CEO. This was in the business plan and anticipated growth charts.
I (and what remained of the development team) left within 3 months of that plan (it took a while to realise it wasn't actually a joke). We were replaced and within 6 months the company was defunct.
1/ Do not over estimate the value of your service. It may seem irreplaceable to you (and your friends) but most people will put a value of zero on its functionality until they have used it. If the entry cost is prohibitive people will not use it and your market will not develop let alone grow.
2/ A start-up is not a corporation (even though it may be listed as such). Things like HR & OHS do not require a separate staff - they require a damn good (and damn patient) PA until staff numbers grow.
3/ If you previously worked in a high profile position for a large corporate - perhaps a start-up position is not for you. If the only thing that got you attention previously was 'from XX corp' when you said 'This is Joe Blogg from XX corp' perhaps you are not as good as you thought you were.
4/ Start-ups are technology sensitive at the start - you need to get the technology out there and in use before you can capitalise on the business opportunities. This is not hard to do with minimal cost these days.
5/ Make sure your sales people (or sales information if you do it all online) concentrate on what your product does now (and perhaps compare it against what your product used to do to show how it is improving) rather than concentrating on what it will do. If I was trying to sell you this years version of the Prius would I concentrate on the differences between it and next years Prius (well, next year the brakes will really work - all the time)?
6/ Don't muck around your developers. Start-ups are technology sensitive (see point #3) - if your developers make suggestions about who that technology is good for don't ignore it if it fails the first time (hey - we go through several iterations to get your requirements implemented - can't you do the same for us?)
There are a million stories in this city - that's mine (Hope you are a Dragnet fan :P). I spent 4 years working between 60 and 100 hours a week on four continents developing a product I really believed in. At the end of it all I probably wound up earning less than I would have for a strictly 9 to 5 government job with no stress and spent 12 months or more recovering from the burnout. On the upside the work was always challenging, I made friendships with some really (really) good coders and learnt a lot in the process.
Statistics say your experience is going to be more like mine than that of the first Google employees - don't let that stop you though, the experience itself is worth going through at least once.