The Cone of Uncertainty

There is an ongoing discussion happening over in the LinkedIn Drupal group regarding how much a particular website might cost. Search for "I am interested in finding out the price range that would be charged for a website redesign brought into Drupal. Approximately 30 page it needs a blog, picture gallery, and video, PR added." in the "groups" section.

Many web developers chimed in with the different ways they estimate a cost of a website, but the most interesting post to me related to the agile method of development without a fixed bid; instead, the company creates projects and bills in 10-hour block increments at a time, with the idea that as the project comes to fruition, the client has more leeway to decide priorities and to flesh out the detailed specifications as needed. With more frequent demonstrations, the client can also visualize and engage with the site better -- and thus lead to a more satisfying end product.

Since our inception, we've offered a fixed-bid contract for a fixed set of deliverables, and we'd consider ourselves in the camp of web design/development companies who often end up "eating" some of the cost dues to "feature creep" and our clients realizing they need additional functionality mid-way through development. While some change orders may take less than 15 minutes to install, other items require heavy customization, but customization within the web development framework is the name of the game. Why build a site that looks like everyone else's? You want your website to stand out from the crowd.

All building professions, including ours, are dictated by the "iron triangle" of quality, time, and cost. I've blogged about this before, but basically, you can only pick two of the three: high quality, fast turnaround, and low price. By removing the fixed-bid aspect of the engagement process, the development relationship then moves towards more of a partnership model with both parties providing feedback on the direction of the site and what should the next priority be within that next "block" of time.

I'm interested in your feedback about fixed-bid with rigidly-defined scope versus incremental charges and a more expanded potential for definition. During our website development process, we spend the first phase modifying, parsing, and understanding the full functionality of the website. I wonder if that time could be better spent doing wireframes, starting up some basic pages, and giving you -- our customers -- more to think about.

Do you prefer a fixed fee for a fixed set of deliverables or do you desire more flexibility? Of course, some budgets are firm, so then it becomes a matter of prioritization: where do we want to allocate the time available? Towards specifications, design, functionality, user workflows, or _______ ?

The "cone of uncertainty" refers to the unknowns that arise at the beginning of a website project. The cone gets smaller and smaller as more features become defined, but the whole process relies on defining, in almost infinite detail, what the website is "supposed to do" when it goes live.

Our projects come to a satisfactory completion -- see our testimonials -- but I'm still interested in what your thoughts are about converting from a fixed-fee process to an incremental fee process.