How do you estimate a Salesforce.com implementation? PART I

It's been a while now that I have had a chance to post something on my blog for a long time! Lot has changed, as I have now moved to Singapore from India, but still in the Salesforce.com space, trying to grow as this space grows :)

Well, after having dirtied by hands on both project deliveries and pre-sales for Salesforce.com engagements including sales, service, marketing, custom, integration, data migration, training, global/multi-country rollout etc, there is still one question that pops-up every time I look at a new set of requirements or RFP - How do you estimate or size up a Salesforce.com project??

If you look up on Google, you will find tools and frameworks for estimating an SAP or an Oracle implementation, but unfortunately nothing comes up for sizing a Salesforce.com project. I understand that the usual, conventional ways of estimations fail when it comes to SFDC and its largely driven by experience, which, well, may not be accurate but at least gives you a good sense of what to expect in terms of the effort, duration and of course, cost!

In my opinion, ONE, it more or less depends upon the depth of insight which have into the requirements and objectives and to what extent of detailing as been done into functional requirements and TWO, how intelligently you write your assumptions. 

For example:

Customer says - I need to implement Opportunity Management. That's it! That's all you have!! 

So, how do you estimate the effort and cost for implementing Opportunity Management in Salesforce.com? I mean, this could just be the TIP of the ICEBERG! Opportunity management could mean a lot more than just creating opportunities in Salesforce.com. 

You could either go plain vanilla saying

1. I do 3 sales processes (maybe you were able to get something from the customer) - which will entail 3 record types, 3 sales processes and possibly 3 page layouts with custom fields (say 20) and validation rules (say 5) with maybe a few workflows (say another 5)

Key here to note would be the capping that you do on each of the function point i.e. record types, layouts, validations, workflows etc and that these are clearly articulated in your assumptions. Other important aspect would be to break down your function points into 3 levels of complexity - LOW, MEDIUM and HIGH - and based upon your experience, you assign a fixed number of hours to each of the complexity level for a given function point

Would love to write some examples here but running short on time. But will be back soon with more thoughts and experience.

Meanwhile, I would love to hear from my fellow community members on their experience in doing estimations for a Salesforce.com/Force.com project. 

You never know, maybe we might come up with our own tool or framework that could help the other out there!