What should an ecommerce site cost?

Should an ecommerce site cost  £1,000, £10,000 or £50,000? There is a big disconnect between our development costs for an ecommerce site and many prospect's expectations. 

• With a designer, we developed an ecommerce brand and web site (www.dotcomstore.co.uk) and used Pay Per Click (PPC) to get clients. I think it is a great site, and it seemed like a pretty good wheeze at the time.  We spent a couple of grand on PPC before we gave up.  Why?  We got lots of leads but every one of them expected to spend £2,000 or less on their web site. 
• www.firstratedirectory.co.uk is a pretty smart lead generation web site.  They publish requests for information (RFI) aimed at people in our businesss.  I don't believe I have seen an ecommerce RFI on the site ever with a budget of more than £5k.  Mostly budgets are in the low four figures.  I recall someone who wanted an ecommerce site for an auto parts business with a full bill of materials breakdown and a budget of £2k. 

There is no way we can produce a web site for this sort of number.  A good design by a competent designer is in the £3-6k range and even a very vanilla site will take us a week to set up.  So our starting budget is around £6k and we go up to £50k for the bigger more custom sites. 

Are we doing something wrong? One of our clients decided for very good reasons to use a (very) serious hosted US-based service called Netsuite.  Sadly there was some confusion about who was going to skin the various functions and we ended up doing it.  My programmer was on this for about four days with the HTML already cut.  This stuff can be complicated.

The only way these expectations can be met as far as I can see is:

1. Use a pre-existing template married to a vanilla ecommerce system.  Literally plug and go.
2. Off-shore the work.

There is obviously a market for the first approach.  I have had a few leads myself from people with low level operations and a vanilla requirement who I have pointed at a templated system.  However the leads I am discussing above are for people who's business model justifies something better than this, indeed won't get anywhere without a substantial marketing budget.

The other issue is that project management time is often disproportionately high for low-budget projects, simply because small clients don't have the internal disciplines of larger companies.  A £2k budget could easily be eaten up in client relations.

The £64,000 question is what happens to these guys? Do they get a serious site built in India for £2,000 and then go on to make their first million or do all these enterprises disappear? 

Should I be developing a templated £2k solution or stick to my knitting?   If I do, how do I control project management costs?