Cost = -b +/- sqrt[(b*b - 4ac)/2a]

A few years ago, I was asked by an account manager whether there was a formula he could apply to calculate the cost of “a website”. He was looking for a spreadsheet he could take to client meetings to provide costs on the fly without getting project managers and developers involved. You can imagine the sort of thing – a picking list with functionality and prices next to each one.

I remember thinking it was a reasonable enough request. But very naive.

I’m sure many an agency has tried to do similar and I’d be interested to know with how much success. In my experience, there is no substitute for solid understanding of the technical side of what we do. Building websites is as complicated as building a house – and often takes longer! Counting how many bricks you need is the easy bit; the money is earned working out how to deal with different terrains, the drainage, the exposure, the electrics, building regulations. Not to mention the countless unforeseen issues that will inevitably arise.

To coin a phrase, the devil is in the detail. And that detail is hard to capture with a checkbox…

Conservatively, I reckon I’ve been heavily involved in the launch of 150-200 distinct websites over around 10 years. If I were so inclined, I could probably group those into categories: e-commerce sites, content managed sites, Flash sites, custom sites, integration projects etc. No question, if somebody spoke to us about building “an e-commerce site”, we could probably establish a ballpark budget range fairly quickly. This would be based on years of experience and, to be honest, gut feel. Maybe you’re checking boxes in your head, but there’s usually an accompanying story with each tick.

If you’re not convinced and are still in search of the holy grail of spreadsheets, here’s some things to consider:

• Design is the most unpredictable element to estimate. It’s easy enough to demonstrate whether you’ve delivered a piece of functionality, but it’s far less clear cut with something as subjective as design. Not only that, but it tends to be the element most personal to a client.
• The number of pages or products (in an e-commerce store) is rarely significant. Generally, we’re buidling dynamic, database driven sites. If you have 10 pages or 100, the main bearing on development time is where you have distinctly different functionality. For instance, a site which sells 10,000 different CDs may be less complicated than one that sells 20 CDs and 20 tailored shirts.
• Integration is the most likely thing to throw a spanner in the works. Getting one site with relatively complex functionality up and running is a challenge; get two or more sites to share data successfully and the costs/complexity go up exponentially. If anyone tells you otherwise, they’re kidding themselves!
• Developers and designers underestimate (a generalisation with some foundation) to differing degrees – read my thoughts on estimation here.
• The biggest variable is the client – try factoring that into your spreadsheet!

P.S. In the spirit of Web 2.0, please share your spreadsheet with us if you have one.