How to compete on a scarce spec'd project to avoid team death-march

I'm time+cost estimating a semi-complex software solution, that hasn't got specific requirements in about 75% of features. I would still like to make as good estimate as possible, by getting additional data from the client. There will still be parts that may end up not being able to develop, since there's too many dependencies with other products/technologies and lack of definition. I also have a very tight schedule to produce this estimate.

There will also be other contenders on this project. Client expects a price+duration (and probably also features) and I know everyone will be off. I know that's impossible, but tell that to marketing people. Another problem is that I'm talking to middle-man and not directly to the client. I can get confidence with middle-man only, but not with the deciding client. Which is a different problem altogether.

What disclaimer/info can I put in my price plan/contract not to kill team with this project, so when project starts slipping (in terms of cost/time/features) we will be covered with some sort of payment. I would of course like to be paid by sprints or releases with relation to time, but I doubt client could be convinced in this. I'm sure we can finish this product before deadline and also create a great product, but how can I convice client to believe me?

Question

What can I do to get this project and avoid death march situation at the same time?

Any suggestion welcome!

EDIT: Outcome

In the end we (me and my co-worker) convinced the client we need at least a week to evaluate the product. So we did. We also pushed (and got) a slot for a few hours long meeting with the client to clarify any outstanding requirements' questions. So we did. Meeting was done after we made the first estimate draft, so we were sure we have all the questions to point out specifics that were either completely misunderstood or too vague to estimate. I hope we get the project, because it would mean 8 months of full time work for us, plus a reasonable pay. We'll know in about a week and a half.

Of course I also pointed out that the way we'll be delivering this product will get them exactly where they want to be with a product that will actually be what they wanted. And also that we only commit to price and time, but not functionality, because it is and will be subject to change. I think we made a good enough impression.

Welcome to the world of fixed priced development services :-)

Techniques for to win this project and avoid death march situation at the same time:

• Don't underbid a project. Bid for what you think the project will take and add some percentage for things likely to go wrong.
• If you are missing 75% of the detail, odds are the project will be significantly different than you currently expect. Document some reasonable detail assumptions within the outline of the defined work. When the project actually starts and the details don't match the assumption, you have an opportunity to negotiate the costs for the changes. At that time, you may also be in a better position to know how much you are over/under and attempt to compensate with this quote.
• Your goal in an SOW (statement of work) should be to define enough details so that it gives you an opportunity to renegotiate the cost of changes when you know more about the project. Write these as positives, as much as possible. Note, it is unlikely that people that actually understand the project will read or understand the SOW...I base this on the point that you are given few details to quote. This means it isn't a consultative sale and neither party is really focused on building the 'right' solution.
• If you can get a contract as T&M (time and materials) great. I doubt you'll get it or unable to get it without some restrictions that essentially defeat the purpose of a T&M. Your potential customers look at this as them accepting all of the risks around your abilities.
• Hopefully, you aren't the first at your company to do this. Find out, historically, how projects have been and the typical result rates. Many software development groups charge an hourly rate that is significantly higher than cost...but their quotes tend to be lower and not actual hours. Customers often will argue more about the hours/days than the actual quote. Enterprises tend to be used to paying high hourly rates.
• Figure out your department's expected margin (profit you need to gain from the job). This may help you to understand how much of a 'death march' you may face when your project slips.
• In the SOW specify the level of detail that will be required in a specification before you begin work. While Agile and other customer focused processes take an approach that oriented at finding the best solution, they aren't designed to keep costs under control in a fixed bid environment. You will need to take a waterfall approach to requirements and then build in an agile fashion so that you can adjust along the way. The specification, like the SOW, will give you an opportunity to bill for changes. While the customer won't like this, it will put the burden and risks associated with requirements on them and not your team.

Note, to be successful with these negotiations, you needs a supportive management, sales and project management team. If you don't have that, you are bound to always be on 'death marches.' Even if you forgo quality, process, testing and other items, you'll find there's never enough time for a project.