Estimating and Project budgets

Question

I'd be interested to hear how you manage dealing with project budgets when having to provide a fixed cost up front for the customer ahead of starting the work?

For context, and without giving too much away, I work as a PM developing online 'software' type builds for a company where we have to provide costs up front for developments based on signed off specs, but often the development takes much longer than estimated - leading to the budget going out of the window. Obviously that's not sustainable long term.

It'd be great to hear how this is dealt with in other organisations...

Answer 1

Assuming you've done the work before on similar engagements, you and your team's experience should be able to arrive at a probabilistic estimate for the work ahead in a range from best to worst case. Since the work is fixed price, you would then select for your planning / target values--against which a price would be set--closer to the worst case side of your estimate. Said a different way, you would load a ton of contingency to cover all of the unknowns.

For customers that like fixed price, they essentially remove the risk of cost growth for the project but in doing so they pay a premium for it.

In addition, you would be very explicit to what is assumed, what is included, and what is excluded. The language in your contract needs to be very clear so that, if the project bangs up against one of those criteria, then you can submit a change request and add both cost and schedule to your fixed price.

If your project is new to you or even a first of its kind, then avoid a fixed price contract if you're able. If you're unable, then agree to a fixed price for portions or stages of the project and get very very pessimistic with your targets based on your estimate curve, i.e., choose the worst case scenario and then some. Of course, that might make your price uncompetitive but unless you can survive losing money let someone else take the loss.

Answer 2

There are two crucial parts for making the estimation that will be met in the future.

1. First, the estimation should be correct (the amount of work to be done predicted without mistakes).
2. Secondly, the scope will likely change. So new tasks should be evaluated in addition to the cost and time you've agreed before.

You only asked for the first part, so here is what I think based on my experience.

1. Ask the person (or the team) that will actually perform the job to make estimations.
2. Compare the estimation they made to the time the similar tasks took in the past.
3. Make sure that the project is decomposed to a number of small tasks of 4 to 8 hours each max.

If the estimation differs a lot from a similar task's estimation from your prior experience, this means you should either find out the reason, or expect the task to take longer (like it already happened before). If the estimated tasks are expected to be done in more than 4-8 hours, it means they may be too big to be estimated properly, so you may want to ask to decompose them further.

You could also outsource some tasks. In that case you may want to have the contractor that has already proven its/her/his ability to do the proper job for you, or start with small not important tasks.

Finally, you should analyse the possible risks and increase time estimation accordingly, so that you promise the date that is further in the future than what you was promised by your team.

---

And one more thing, that is also important. Once you notice that somebody falls behind the schedule, you should act right now, don't let the delay to increase.